BOTANICAL MUSEUM
LEAFLETS
HARVARD UNIVERSITY

PRINTED AND PUBLISHED AT THE
BOTANICAL MUSEUM
CAMBRIDGE, MASSACHUSETTS

BOTANICAL MUSEUM LEAFLETS
HARVARD UNIVERSITY

VOLUME XXIV

BOTANICAL MUSEUM
CAMBRIDGE, MASSACHUSETTS
1974-1976

TABLE OF CONTENTS

NUMBER | (July 31, 1974)

Archeological Evidence for Snuffing in
Prehispanic Mexico
By Pati PONS coyeenenndecnacna den pes teks i

NUMBER 2 (October 25, 1974)

A Study of Systematic Wood Anatomy in Cannabis
By LoRAN C. ANDERSON .........0cc cece ceeeeces 29

Two New Brazilian Species of Brunfelsia
By TIMOTHY C. PLOWMAN ...............-0eeee ee OU

NUMBER 3 (January 10, 1975)

Archeological Maize from Northern Chile
By PauL C. MANGELSDORF AND
CORDON Foil AW een sa ong te acaanaes 49

NuMBER 4 (June 16, 1975)

The Contribution of Linnaeus to Orchidology
By WILLIAM T SPEARN 9 co-aton4 nd bteedegeus ideas 65

NUMBER 5 (June 27, 1975)

Ethnobotany of the Western Tarahumara of Chihuahua,
Mexico, I. Notes on the Genus Agave
By ROBERT A. BYE, JR., DON BURGESS, AND
ALBING MARES TRIAS. vcuw ores bah5 or tah dad vex aon 85

NuMBER 6 (October 31, 1975)

Nutritional Value of Coca
By JAMES A. DUKE, DaviD AULIK AND
TIMOTHY PLOWMAN ...........cccccccccccccesves LES

De Plantis Toxicariis e Mundo Novo Tropicale
Commentationes XIII. Notes on Poisonous or
Medicinal Malpighiaceous Species of the Amazon
By RICHARD EVANS SCHULTES ........- +e ee eeeees 121

NUMBER 7 (December 12, 1975)

Studies in the Genus Oncidium. I

DSU Fe SEAGY. tavuep su iaveeeseereeke napa ee 133

NuMBER 8 (April 9, 1976)

Plantae Colombianae XIX. E Partibus Amazonicis
Witotorum Plantae Fructuariae Sativae Novae
By RICHARD EVANS SCHULTES «2. ...00. 0008s eeees 193

Pachytene Chromosome Morphology and
Its Bearing on Interspecific and Intergeneric
Relationships of Coix
By J. VENKATESWARLU, R.S.K. CHAGANTI AND
PANUCANTI NN. RAG: beeee se iisdeeias ees be esees 205

New Chichuahuan Umbelliferae
By LINCOLN CONSTANCE AND ROBERT A. BYE, JR... 229

Chromosome Numbers in Chihuahuan Umbelliferae
By LINCOLN CONSTANCE, TSAN [ANG CHUANG AND
ROPER T Ba PI, ga sc chara ed pee yes Chee Ses 24]

NuMBER 9 (June 30, 1976)

The Genus Oeceoclades Lindl.
By LESLIE A. GARAY AND PETER TAYLOR ........- 249

NuMBER LO (November 30, 1976)
Manicaria saccifera and Its Cultural Significance
Among the Warao Indians of Venezuela
By JOHANNES WILBERT sxi 544s ccee Ses sqm wees 2795

[ vi |

INDEX OF ILLUSTRATIONS

PLATE
Agave americana var. expansa ............0.000- XVII
Agave bovicornuta .............cccccceecceeees XVIII
Agave multifilifera ..............0.. 0 ccc eee ee XXII
AGAVE DACINCS. oni uGale tied eae ed bbe de heeds e ae nns XIX
Agave polianthiflora ............ 0... cee eee eee XXII
Agave SATEVEl cad iiwine ii ers ¥e% beeen beannei XX
Agave Vilmoriniana ................c cece eeeeeee XXIII
Agave WOCOMIAML kana s iss oyeecusatiah cae codunews XXI
Brimrersia Dil0Ga. s+ peeve nso el ecek een vor8 XIII, XIV
Brunfelsia Martiana ............0cc cece ee eees XI, XII
Cannabis wood sections .......... 0.0.0. ccc cee cence X
Coix
Ideograms of pachytene chromosomes .......... LX
Five pachytene chromosomes of C. aquatica ... LVI
Pachytene nucleus of C. aquatica .............. LV
Pachytene nucleus of microsporocyte
GE CG. ZiPanted .4scad bc Sods eeu Veedei ces edcana LIX
Pachytene nucleus of C. Lachryma-Jobi ...... LVIII
Pachytene nucleus of microsporocyte of
C. Lachryma-Jobi ............ 0. cee cee euees LVII
Corn cobs, late prehispanic ...............0 00000. XV
Corn cobs, dated at ca. A.D. 100 and 1250 ........ XV
Donnellsmithia silvicola .............. 002.0000 LXI
PYyneiiin Gentry) s+ swe cenieed owihyie ee cees LXIV
Heteropteris macrostachya 444. é.s ised ivaeia sae MLV
Pie@terapreris Fiala coh<eseehecie cay en se cchanncelVy
PRinA@a APADOTOUSIS +. ie n:kst nau ee dak ad os od
[niga @norerana.- 4.66222 <40y dx cic cde wedavnwacadvs LIV
Macoubea guianensis ......... 00.0 cece eee eee ase LII
Macoubea witotorum ............c00 cece eee XLIX, L
Macoubea wititorum, fruits of ..............0-0 000s Li
Manicaria saccifera
CAOWINe ti 1Otest 64 ees 4aeabaidted sein escent’ LXVI
Details of inflorescence ...... LEAVE, bas VL LA
Thatch on Warao houses ....... LAX, DAS, LAXH

Leaves used for sails on canoes

LXIII

PER any eee See ee ae,

AUIS saqd2au Swneee ter eres ea ogee he aes LXXIV
Hats made from spathes .... LXXV, LXXVI, LXAXVII
Fruiting spathe and infructescence ....... LXNXVITI
TOP WRU Cd OI SOR, ne siyss oto eved Soho ass LXXIX

Steps in the processing of sago

Map of Karaparana-Igaraparana region of

Colombian Amazon
Mascagnia glandulifera
Mauritia flexuosa trough
Mochica IV terracotta vessel

Oncidium
Oncidium
Oncidium
Oncidium

Pome nr me SAN er eee Sar ar ee ee

Ce

LXNXXIT, LXXXITI
ea ee es eee bene s IX
... XXXVI, XLVIII

aequinoctiale (in color)

alticola (in color) ......XXXV, XLV, XLVII
ANCIPERUGM fo naae en ce siadeneesis ges oa
APUAVENSE dndeneans chawdedn ee ieee

Oncidium chimborazoénse (in color) .XXXVII, XLVII
Oncidium cucullatum (in color) ......XXITX, XXXVITI
Oncidium cucullatum var. dolabratum ..........XXDX
Oncidium cucullatum var. macrochilum ........ XIX
Oncidium Dayanum (in color) .... XXXI, XLII, XLIV
Oncidium erosilabium (in color) ......XXXVI, XLVII
Oncidium Kennedyi (in color) ................ XXXIX
Oncidium mimeticum (in color) XXXII, XL, XLI, XLII
Oncidium nubigenum (in color) ........ XXXIV, XLVI
Oncidium olivaceum (in color) ..... XXVIII, XXXVITI
Oncidium olivaceum var. giganteum ......... XXVITI
Oncidium Phalaenopsis (in color) .............. XLII
Oncidium rhodostictum .............. XXX, XXXVITI
Oncidium sanguinolentum (in color) ..... XXXII, XLI
Oncidium spathulatum (in color) ........ XXXII, XLIV
Oncidium tarquiense (in color) ..........XXXV, XLV
Oncidium tripterygium (in color) ............... XLII
Oncidium tunguraguense (in color) .. XXXVIT, XLVIII
UTIs. “nech ees ac oecae ewe eens eae bad oe a I - VIII
TASCA OCG 2nbese cian ceeeee eee has es owes LXII
TP etisC ie PAPAIN. enact ees erred Lae eae teas LXIII
Petraptens Muctonat®: s425004 441-44 50544es rene VEL
Tetrapteris styloptera .......-.-.-+-000++0+.+. MXVII
Triorchis sera (Orchis morio) .............002+++. V1
Umbellifereae, karyotypes of Chihuahuan ....... LXV

[ vill |

AECEOCLADES 253

AEROBION
maculatum 266

AGAVE 85
americana

var. expansa,

bovicornuta 94
multifilifera 97
nomenclature of Western
Tarahumara LOL
pacifica 95
polianthiflora 97
Shrevei 96
Vilmoriniana 97
species of Western

93

Tarahumara LOO
Wocomahi 96
ahue 91
AMANITA
muscaria 22
ANADENANTHERA
colubrina 22
ANGRAECUM
maculatum 253,254,266
monophyllum 268
ARRACACIA 245
atonali 88
balila 89
BANISTERIOPSIS
Caapi 3,122
Martiniana
var. laevis 123
bee-ra-ree-a-ma_ 130
BRUNFELSIA
pilosa 42
Martiana
CANNABIS
indica 29
sativa 29
wood anatomy 29,31
chahui 93
cobisi 89

37

INDEX

[ioe 4

coca 113
117

insecticidal residues 117

alkaloids in leaves

nutritional comparison 119
nutritional value
COIX
aquatica 207
chromosome numbers 205
gigantea 210
Lachryma-Jobi 210
COPERNICIA
cerifera 277
curanderos 3
CYMBIDIUM
calcaratum 261
DONNELLSMITHIA 241
silvicola 226
eé-taw-gaw 126
emukohoko 300
EPIDENDRUM
connivens 267
ERYNGIUM 245
Gentryi 235
ERYTHROXYLUM
coca 113
EULOPHIA 254
alismatophylla 258
ambongensis 268
atrovirens 260
Bierleri 271
calcarata 261
cordylinophylla 261
Decaryana 262
dissimilis 264
Elliotii 271
gracillima 263
Ledienii 266
lanceata 263
latifolia 264
lokobensis 261
lonchophylla 264
Lubbersiana 265

ies

Mackenii 265
maculata 266
Milbraedii 270
pandurata 268
paniculata 261
petiolata 269
quadriloba 269
Saundersiana 270
Schlechteri 258

sclerophylla 271
seychellarum 272
spathulifera 272

tainioides 264
ugandae 273
,ULOPHIDIUM 253
alismatophyllum 258
ambongense 258
analamerense 259
analavelense 259
angustifolium 259
ssp. diphyllum 259
boinense 260
cordylinophyllum 261
Decaryanum 262
dissimile 264
gracillimum 263
latifolium 264
Ledienii 266
lobense 262
lonchophyllum 264
Lubbersianum 265
Mackenii 265
maculatum 266
var. pterocarpum
monophyllum 268
nyassanum 267
paniculatum 261
panduratum 268
Perrierit 268
petiolatum 269
quadrilobum 269
Rauhii 270
roseovariegatum 270
Saundersianum 270
sclerophyllum 271

267

seychellarum 272
spathuliferum 272
tainioides 264
Warneckeanum 266
zanzibaricum 273
galime 92
GEODORUM
pictum 266
GRAMINEAE 205
GRAPHORCHIS
260
maculata 266
Saundersiana 270
vue huali 89,90
gusime 92

atrovirens

hallucinogens 1
henequén 85
HETEROPTERIS
macrostachya 124
riparia 123
suberosa 124
HIRAEA
apaporiensis 124
Schultesii 124
ho-ree-a-mee-see

234

130

huve

ILEX
Guayusa 3
paraguariensis 3

ime 92

ime juice 89

INGA

chorrerana 202

JUSTICIA 4

LEOCHILUS
sanguinolentus 152
LIGUSTICUM 245
LIMODORUM
maculatum 266
LISSOCHILUS 254
ambongensis 268
analamerensis 259
analavelensis 259
barombensis 271

boinensis 260

cordylinophyllus 261
Decaryanus 262
Elliotii 271
gracillimus 263
Hebdingianus 263
lokobensis 262
lonchophyllus 264
panduratus 268
paniculatus 261
petiolatus 269
quadrilobus 269
Schlechteri 258
spathulifer 272

litos 5

LOPHOPHORA
Williamsii 3

MACOUBEA
guianensis 198
witotorum 193

maguey 85

maize
archeological 49
Capio Chico 61
Chutucuna Chico 61
cobs from site A.D. 105 51
Confite Iqueno 53
Confite Morocho 57
Confite Puntiagudo 57
Polulo 53
Rabo de Zorro 57
sting cobs 57

Malpighiaceous Amazonian
species
medicinal 121
poisonous 121

MANICARIA

saccifera 275
composition of starch 319
ethnobotanical lore of 319
ethnomedicine of 303

MASCAGNIA
glandulifera 126

masiavari 230

maté 3

mee-see-gaw 123

mesagoli 89

mescal bread 89

MEZIA
includens 126—

NICOTIANA 2

no-na-mee-koo-ma_ 130

OECEOCLADES 249,253
alismatophylla 258
ambongensis 258
analamerensis 259
analavelensis 259
angustifolia 259
atrovirens 260
boinensis 260
calcarata 261
cordylinophylla 261
Decaryana 262
gracillima 262
Hebdingiana 263
key to species 254
lanceata 263
latifolia 264
lonchophylla 264
Lubbersiana 265
Mackenii 265
maculata 266

var. pterocarpa 267

monophylla 267
pandurata 268
Perrieri 268
petiolata 269
quadriloba 269
Rauhii 270
roseovariegata 270
Saundersiana 270
sclerophylla 271
seychellarum
spathulifera 2
ugandae 273
zanzibarica 273

ojcome 92

ONCIDIUM 133
aequinoctiale 162
alticola 159
andigenum 151

a oe |
~l
aw

azuayense 156
chimborazoénse 164
cucullatum 137,144
subvar. andigenum 151
subvar. spathulatum 148
var. andigenum 151
var. Chestertoni 148
var. Dayanum 150
dolabratum 144
flavidum 156
maculosum 146
microchilum 146
nubigenum 158
olivaceum 14]
Phalaenopsis 150
purpurascens 146
sanguinolentum 152
spathulatum 148
Dayanum 150
erosilabium 162
Kennedyi 142
mimeticum 152
var. flavidum 156
nubigenum 158
var. spathulatum 148
olivaceum 141
var. andigenum 15]
var. flavidum 156
var. giganteum 141
var. Lawrenceanum
var. maculosum 146
var. nubigenum 158
var. Phalaenopsis 150
var. sanguinolentum 152
Phalaenopsis 150

var.
var.
var.
var.
var.
var,
cdr.
var.
vdr.
cdr.

142

var. Brandtiae 150

var. excellens 148
rhodostictum 146
sanguinolentum 152
spathulatum 148
tarquiense 160
tripterygium 151]
tunguraguense 166

Orchids of Linnaeus 65
o tosa 92

[ xii |

palm, Temiche 286
pevote 3
pipes, nose 14
Piptadenia 3
pod corn 57
popcorn, Peruvian 57
pottery artifacts, Oaxaca 16
PRIONOSCIADIUM 245
PSEUDOCYMOPTERUS 245
PSYCHOTRIA 4
pueblo duck mythology 18
RIVEA
corymbosa 3
rolobusi (Bovardia sp.) 87
SACCOLABIUM 253
sapogenin 91
sa puli 92
sarabatucu. 124
sepe 235
shamans and shamanism
sisal 85
smilagenin 91
snuff 1
hallucinogenic 4
snuffers and snuffing 1,6
Ixtlan 10
Virola 4
SOPHORA
secundiflora 24
SORGHUM
intrans 218
purpureo-sericeum 218
starch, palm 275
stupifving fish 91
sugui (tesguino) 88
SYAGRUS
Romanzoffianum 277
takwatsi 22
tamatsi Maxa Kwaxi
tarahumaras 89
TAUSCHIA 245
bicolor 229
tarahumara 232
TETRAPTERIS
mucronata 128

90

ae

20)

silvatica 128 wee-po-awk 130
styloptera 128 willka 22
tortillas 90 Xochipala art 12,14
UMBELLIFERAE 225 vahuhi aru 279
chromosome numbers 241 yvawihi aukwaha 300
usabi (Prunus Gentryi) 87 yaje 123
vilca 22 ve-aing 124

[ xiii |

a i ae

oe

~ BOTANICAL MUSEUM LEAFLETS
“a HARVARD UNIVERSITY

CampripGr, Massacnuserrs, Jury 31, 1974 Vor. 24, No. 1

ARCHAEOLOGICAL EVIDENCE
FOR SNUFFING IN PREHISPANIC MEXICO
BY
Perer T. Furst

The various hallucinogenic or psychoactive plants the
extensive religious and divinatory use of which by Mexi-
‘an Indians both fascinated and appalled the Spanish
colonial clergy of the sixteenth and seventeenth centu-
ries were smoked, chewed, sucked, brewed or macerated
into beverages and otherwise ingested in liquid, solid or
incinerated form. ‘Tobacco, most commonly smoked,
was also ground into a fine green powder that was not
taken internally but was rather applied externally to the
patient's body in shamanistie curing practices.

All these customary uses of ‘‘mind-altering’’ sub-
stances are well described for both prehispanic and post-
Conquest Mexico by such writers as Sahagin, Duran,
Hernandez and, somewhat later, Jacinto de la Serna and
Ruiz de Alarcon. In contrast, there is no mention what-
ever of hallucinogenic snuffs, taken through tubes or
through ‘‘nose pipes’’, a common practice in the West
Indies and in Central and South America. Since these
and other chroniclers of indigenous practices and beliefs
were usually careful observers, and since the Church was
engaged in a vigorous—albeit ultimately unsuccessful—
campaign to discover and suppress the indigenous use of
intoxicants of all forms, we must assume that, trade and

[1]

ECONOMIC BOTANY * \RY.

las Fa
Le

other contacts with Caribbean and Central American
snufE~using cultures notwithstanding, the Indians of
Late Post-Classic Mexico seem not to have assimilated
these practices into their own extensive complex of
ritual intoxicants.

Nevertheless, there is a growing corpus of data, in the
form of archaeological art, to suggest that snuffing was
once known and. practiced in several parts of Meso-
america as early as 1500-1200 B.C. and at least as late
as the first centuries A.D. Before we examine some of
the evidence, we need briefly to consider the problem of
potential indigenous Mexican sources for hallucinogenic
snuffs.

Thanks to prodigious research in the field, the labora-
tory and the historical sources, especially by such inves-
tigators as Richard Evans Schultes, S. Henry Wassén,
Siri von Reis Altschul, and Bo Holmstedt, the various
kinds of South American snuffS are botanically and
chemically rather well understood. For Mexico, how-
ever, the botanical data are inadequate, and chemical in-
formation is wholly or largely lacking. Nevertheless,
there are indications of the direction that future research
might fruitfully take.

First, one cannot rule out one or more species of Nico-
tiana. These native tobaccos have a much greater nicotine
content than do the hybrid species from which cigarette
or pipe tobaccos are made. Wilbert (1972: 55-73) re-
cently documented the use of tobacco as the sole psycho-
tomimetic employed by the shamans of the Warao In-
dians of Venezuela, who smoke themselves into ecstatic
trance states that are phenomenologically indistinguish-
able from those elsewhere triggered with such botanical
hallucinogens as ayahuasca, the sacred mushrooms,
morning-glory seeds, or cfnadenanthera and Virola
snuffs. Several species of Nicotiana are in tact employed

[2]

for psychotomimetic snuff in South America, either
alone or in combination with other psychoactive plants.
The shamans of some indigenous cultures—e.g., the
‘Tacana of Bolivia—use pulverized tobacco as a magical
repellent against hostile demons (Hissink and Hahn,
1961). Furthermore, it is not impossible that the use of
tobacco powder as a magical external medicine by Mexi-
‘an curanderos had its ultimate origin in an earlier use
of powdered tobacco as snuff.

Secondly, there appears to be no reason why some of
the better known Mesoamerican plant hallucinogens
should not be as psychotomimetically effective when
taken as snuff through the nasal membranes as when as-
similated through the stomach. By way of illustration,
I am informed by Dr. Schultes that, in South America,
the bark of Banisteriopsis Caapi, which is usually macer-
ated or brewed into the potent hallucinogenic beverage
known by such names as yajé, ayahuasca, etc., is re-
putedly sometimes pulverized and inhaled as snuff. Even
Ilex Guayusa, a caffeine-rich holly widely utilized as a
stimulating tea, along with its sister species, e.g. mateé,
Ilex paraguariensis), has served as snuff, at least in pre-
hispanic highland Bolivia, where Ilex snuff and snufhing
paraphernalia were recently discovered in a Tihuanacoid
shaman’s grave, dated ca. A.D. 500. The shaman’s kit
also included clysters, suggesting that the same plant
might even have been employed for stimulating enemas
(Schultes, 1972b).

Whether or not peyote (Lophophora Williamsi), olo-
liuhqui (Rivea corymbosa) or other hallucinogens native
to Mesoamerica were ever used in the form of snuff,
there exist extensive Mexican populations of shrubs and
trees of the Leguminosae that should be investigated for
possible psychoactive properties. Included are two Mexi-
can species of Piptadenia, a genus closely related to the

[ 3 ]

psychoactive genus Anadenanthera of South America,
which may well possess the same or similar psychotomi-
metic constituents. Two of these Mexican species are
Piptadenia flava, found also in Central America and Co-
lombia, and Piptadenia constricta. Both are found along
the Pacific coast, from Sinaloa and Jalisco in the north
to Guerrero in the south. To my knowledge, neither
have been tested for hallucinogenic alkaloids. There are
in addition more than sixty species each of the allied
genera Mimosa and Acacia in Mexico, and some of these
may, like certain South American species, contain psy-
choactive principles. This might be found to apply es-
pecially to those species credited with sacred, magical,
or ‘“‘dangerous qualities by local Indians or rural mez-
tizos and should be chemically studied from the point of
view of possible hallucinogenic alkaloids.

Finally, there appear two other possible candidates as
potential sources of hallucinogenic snuff in southeastern
Mexico, both with significant South American. ties.
These are species of Psychotria and Justicia. The former
is a well known additive in hallucinogenic potions pre-
pared basically from the Banisteriopsis Caapt vine in
Ecuador, Colombia, Peru and Brazil, while the latter is
added to /rola snuff or is said even to be employed
alone as a source of psychotomimetic snuff (Schultes,
1972a: 45-46, 52). These possibilities emerge from a
comment by Wassén (1972: 37-88) on a suggestion of
mine (1968: 160-164) that snuffing might have been
practiced by the Gulf Coast Olmec. In support of this
comment, Wassén cites the following excerpt from a let-
ter to him by Schultes, dated February, 1969:

We are finding so many plants with tryptamines—the active
principle of many of the snuffs of South America— that it is very
possible that in the Mexican Gulf Coast area the Indians could
have found a plant which, prepared in the form of a snuff, could
intoxicate as does the snuff of the Waikas. One of these is Psy-

7

chotria, a species of which in South America has now been found
to have N,N-dimethyltryptamine.

Psychotria occurs up as far as Vera Cruz and it is possible that
other species have this principle. Furthermore, Holmstedt be-
lieves that he has found this same chemical in our species of
Justicia which is added to Viro/a snuff by the Waikas. Other
species of Justicia occur as far north as Vera Cruz and may pos-
sibly also have this chemical constituent.

While the botanical sources for hallucinogenic snuff
in Mesoamerica must, for the present, remain conjectur-
al, the evidence for snuffing in archaeological art is, as
we shall see, beyond question. Moreover, on the earliest
level of the proposed Mesoamerican snuffiing complex—
that is, the Early to’ Middle Formative—the evidence
points persuasively southward, at least to Central Ameri-
‘a, if not actually to northwestern South America.

We owe much of our knowledge of Central and South
American snufting paraphernalia—prehistoric as well as
recent—to Wassén’s several studies, and I would here
like to acknowledge my own debt to our Swedish col-
league in this area of research. It was a paper by Was-
sén, published in 1967, that first set me on the track of
a possible snuffing complex in Mexico. Specifically, my
attention was drawn to the so called Brazilian “tos, small
effigy stone carvings, usually bird-like, with carved, shal-
low, oval or circular depressions that made them appear
like receptacles. A number of these were found in the
last century in the shell middens of Santa Catarina,
Brazil. Wassén thought it likely that these bird-efhgy
litos might have served as tablets for hallucinogenic
snuff, rather like the archaeological wooden snuff tablets
found in the Chilean and Peruvian desert, or more re-
cent snuff tablets from Amazonia.

Subsequently, I raised the question of the use of hal-
lucinogens by the Olmec, suggesting that the well known
jade artifacts called ‘*spoons’’, might, like the Brazilian

[ 5 |

litos, have served as snuff tablets (Furst, 1968: 162). At

least some of the Olmec ‘‘spoons’’ seemed to represent
long-tailed birds in flight, seen in profile. In any event,
like some South American snuff tablets, certain Olmec
jade spoons are decorated with bird-jaguar motifs, a com-
mon symbolic theme in South American ritual intoxi-
‘ation.

At the time, this was still highly speculative. No
direct evidence existed to show that the Olmec had used
snuff or other hallucinogens; for that matter, there was
no proof that the ritual use of psychoactive substances
in Mesoamerica was any older than the oldest of the so-
‘alled mushroom stones, 1.e., from the end of the Mid-
dle to the Late Formative. All that could be said was
that it would be surprising if the Olmec had used no
hallucinogens, considering what was already known of
the antiquity and wide distribution of the hallucinogenic
phenomenon in the New World. ‘To mention only snuff
ing: the earliest known archaeological snuffing imple-
ments are a whalebone tablet and associated birdbone
snuffing tube which Junius Bird of the American Mu-
seum of Natural History excavated at Huaca Prieta,
Peru. These are dated at ca. 1500-1700 B.C. The evi-
dence was thus conclusive fora time depth of some 8,500
years for the use of, hallucinogenic snuffs in South
America.

The first evidence that snuffing was in fact known at
one time also in Mesoamerica came to my attention in
the form of a hollow, redware effigy figurine from Co-
lima, representing a seated man with a horn on his head
and a small, gourd-shaped nose pipe held to one nostril
(Plate 1). Subsequently, [ was to come across a second,
considerably larger, Colima effigy (Plate I), of burnished
brown clay, sculpturally far more sophisticated, again
depicting a manin the act of snuffing from a gourd-

[6 ]

PLATE I

Small, hollow terracotta efigy of seated man, holding a gourd-

shaped snufling pipe to his nose. Colima, shaft-and-chamber tomb
~ly

phase, ca. 100 B.C.-—A.D. 200. Anon. private collection. H. 75

PLATE II

Effigy figurine of burnished brown clay depicted in the act of in-
haling snuff from a bottle gourd-shaped nose pipe. Colima, shaft-
and-chamber tomb phase, ca. 100 B.C.—-A.D. 200. Kurt Stavenhagen
Collection, Mexico City. H. 11”.

shaped nose pipe. Both of these figurines belong to the
larger West Mexican shaft-and-chamber tomb art com-
plex and can, therefore, be dated between 100 B.C. and
A.D. 100-200.

In addition to the effigies, we were able to identify,
in several collections, a number of pottery snuffers or
nose pipes from West Mexico that closely resemble the
well known Costa Rican snuffing pipes illustrated by
Weassén in several publications. [specially interesting is
a red-slipped snuffer with bifurcated stems, one for each
nostril, from the Ixtlin del Rio area of southern Nayarit
(Plate IIL). The Ixtlin snuffer is actually a convention-
alized bird effigy, with nubs at the side of the bowl to
indicate wings, and a projection at the front for the head
or beak. Such abbreviated bird symbolism is common on
Costa Rican pottery snuffers as well. That this is hardly
fortuitous was recognized by Wassén: birds and_ bird
spirits are widely connected with the ecstatic trance ex-
perience and with shamanism.

For atime, these West Mexican specimens seemed to
be all that there was. Snuffing, therefore, appeared to be
an isolated phenomenon in time and space, associated with
the shaft-and-chamber tomb cultures of the West Coast.
Their funerary art indicates that these same cultures also
employed the peyote cactus and, possibly, mushrooms.
The close similarity of the West Coast pottery snuffers
to those of Central America, and their restricted distri-
bution close to the Pacific coast, suggested a somewhat
short-lived trait, introduced possibly from a southerly
source, that eventually failed to take hold alongside es-
tablished cults involving such well known indigenous
Mexican hallucinogens as peyote and the sacred mush-
rooms.

However, West Coast snuffing was not to remain the
isolated and short-lived phenomenon that it appeared at

/ 10 |

Puatre Ill

Bifurcated bird effigy snuffer from Ixtlan del Rio, Nayarit, shaft-
and-chamber tomb phase, ca. 100 B.C.-A.D. 200-300. Coll. Mr.
and Mrs. William Kaplan, New York. L. 24”.

first. New evidence has come to light in the form of effigy
and undecorated snuffing pipes from the Karly to Middle
Formative, from Nochipala, Guerrero. The dating of
XNochipala is still somewhat uncertain: there have been
suggestions that it represents the very ‘“‘origin’’ of Olmec
art, predating the San Lorenzo Phase in Veracruz (Gay
1972). However, the reported and confirmed associations
of the extraordinarily sophisticated and sometimes as-
tonishingly naturalistic figurines from Nochipala with
typically Olmec incised bowls and other Olmee artifacts
characteristic of the Late Karly to Karly Middle Forma-

6

tive, including *“*spoons’’ and beads of blue-green trans-
lucent jade, suggests dates equivalent to San Lorenzo
and contemporaneous sites in Morelos and elsewhere in
Central Mexico—i.e., between 1800 and 1000 B.C. On
the other hand, a recent series of thermoluminescence
tests tend to support an earlier date at least for the be-
ginning of realistic Nochipala art, possibly as early as
1500-1600 B.C. (Robert Stroessner, pers. comm. ).
The uncontrolled looting of Nochipala, with its re-
markable assemblage of some of the finest Formative
ceramic figurines to be found anywhere in the New
World, is a scientific tragedy of major proportions. We
‘an only guess at the evidence that has forever been lost ;
nevertheless, it has been possible to study and verify the
authenticity of a number of Nochipala pieces now in
private hands or museum collections (interestingly
enough, although the accidental discovery of the Nochi-
pala site by local farmers dates from the mid-nineteen
sixties, a typical Nochipala figurine has been in the ex-
tensive pre-Columbian collection of the Peabody Mu-
seum of Harvard University for more than seventy
years). Among recently discovered Nochipala artifacts
are several unmistakable snuflfing instruments or ‘‘nose
pipes’, dating far earlier than those of the West Coast

[ 12 |

Piatt LV

Terracotta bowl snuffer, Xochipala, Guerrero, Early Formative,
ca. 1500-1200 B.C. Anon. private collection. L. 24”.

[ 13 |

shatt-and-chamber tomb phase and approaching in an-
tiquity the earliest South American paraphernalia found
on the Peruvian coast.

The first of these to be examined and identified as a
nose pipe was a small, round, undecorated bowl with a
horizontal perforated stem (Plate IV). If one compares
this pottery snuffer with examples from Central America,
it is clear that, except for its characteristic local paste
and the lime encrustation typical of ceramics from the
XNochipala burials, the little Mexican snuffing pipe is
virtually identical to similar instruments from Guana-
‘caste or Linea Vieja, Costa Rica (Wassén, 1965: 25).
As in the case of the Nayarit snuffer, it is difficult not
to postulate a genetic connection between them, al-
though the known Central American pottery snuffers,
and also that from Nayarit, are appreciably later than
this Early Formative pipe.

A second Nochipala nose pipe which | was able to
study in detail is much more complex (Plate V). It is
an effigy pipe, measuring 4?” in length, representing a
human figure on its back, with knees drawn up—a posi-
tion somewhat resembling the post-Classic *“*Chacmool”’
stone sculptures. On the basis of the wrap-around loin-
cloth, the figure can be identified as male. In a recent
museum catalogue, the piece was erroneously described
as an efigy bow] in the form of a kneeling person (Gay
1972). But that would place the nosepiece at the top
and the bowl opening facing vertically toward the front,
which seems hardly likely. Once the piece is recognized
for what it is—a nose pipe used for snuffing—the location
and inclination of the nosepiece alone dictate a supine
position for the figurine, as does the opening of the bow]
itself. Indeed, in handling the piece, its real purpose
suggests itself almost spontaneously.

While it is certainly the finest example known to me,

[ 14 J

[ ¢r ]

Ethgy snuffer representing a man lying on his back. The burnished nosepiece emerges
hornlike from the topof his head. From Xochipala, Early Formative, ca. 1500-1200 B.C.
Anon. private collection. L. 4}.

ALWIG

A

this effigy pipe is by no means unique in form. Several
similar nose pipes have come to light, including one
(Plate VI) of aman with animal characteristics lying on
his stomach, with the bowl in the back and a horn-like
nosepiece on top of the head. Here again the sex of the
efligy is male; indeed, there are indications of a phallus
on the underside.

On the basis of the evidence, then, we can postulate a
snufling complex of appreciable duration and antiquity
along the west coast of Mexico, with the earliest evi-
dence dating to 1500-1200 B.C., and the latest approxi-
mately to the beginning of the Christian era.

A recent re-examination of early pottery artifacts from
Oaxaca as well as from Central Mexico shows, however,
that divine inebriation with psychotropic snuff was not
limited to the Guerrero Formative or the shaft-and-
chamber tomb phase of coastal northwestern Meso-
america. | have only just begun checking through col-
lections and the literature on Monte Alban ceramics, but
already it appears that the evidence for snufling from the
Late Formative at least into the Karly Classic is sub-
stantial. Thus far, we have been able to identify more
than a dozen spouted ‘‘miniature effigy vessels’*, includ-
ing a group in the Museo Frissell de Arte Zapoteca in
Mitla, Oaxaca, as probable nose pipes, dating from Monte
Alban I and II. Some of these appear very similar in
construction, if not in style and paste, to those from
XNochipala. In addition, I have located at least one
probable Early to Middle Formative animal effigy nose
pipe, in the form of a turtle, from Tlatilco, in the col-
lections of the Museum of Ethnology in Vienna, Austria.
This relates stylistically to black effigy ceramics of Olmec
derivation or origin from Tlatilco and Las Bocas, Puebla.

One interesting little polished black snuffing pipe,
possibly transitional from Monte Alban I to IT (i.e. ca.

[ 16 |

PLATE VI

Effigy bowl snuffer in the form of a person with human head and
animal-like body, from Xochipala, Guerrero, Karly Formative, 1500-—
1200 B.C. Asin Plate V, the nosepiece is on top of the head. Anon.
private collection. L. 4’.

200 B.C.), appears to symbolize transformation, in this
instance from human into a duck-like bird with rounded
body and flipper-like feet (Plate VII). The frontal half
is human, with hands held palms together to the chin;
the rest of the body is that of a duck. A conical per-
forated horn on the head forms the nosepiece, as in some
of the effigy pipes from Nochipala.

Space limitations preclude detailed discussion of duck
symbolism, but it should be noted that chimereal or
anthropomorphic ducks are not uncommon in_prehis-
panic art, especially en the west coast. Ducks are present
also in the art of Tlatilco and other Karly to Middle
Formative sites. A study of Pueblo duck mythology and
behets about the duck as supernatural among the Cor:
and Huichol of West Mexico may throw some light on
the problem; a Duck Person is a prominent figure in
Huichol origin myths, as it is also among the Zuni and
other Southwestern Indians; ducks seem to be messen-
gers of the gods or else a form that the gods assume when
they travel. Also, ducks are associated with shamanism,
perhaps because, as wide-ranging water birds, they ap-
pear to inhabit several planes at once.

Of even greater interest isa Monte Alban effigy snuff
ing pipe of grey clay, representing a deer resting on its
stomach, with legs drawn up and head turned to the
right (Plate VIII). Cloven hooves leave no doubt about
the zoological identification. What makes this piece
especially fascinating is that it holds an unmistakable
peyote cactus in its mouth (I am greatly indebted to
Miss Julie Jones of the Museum of Primitive Art,
who recognized the significance of the deer-peyote asso-
ciation here in relation to the Huichol conceptualization
of peyote as deer, and vice versa, and who, on that ac-
count, drew my attention to the artifact).

While anthromorphic pipes from Oaxaca and Nochi-

[ 18 J

Puiate VII

Black-slipped, burnished terracotta efigy snuffer in the form of a
duck-bodied man, from Monte Alban, Oaxaca, Late Formative, Monte
Alban I[A-II (2), ca. 300-100 B.C. L. 33/, Anon. private collection.
A number of similar snuffers combining head and animal characteris-
tics, with the nose piece either in the tail or forming a horn on the
head, are in the Museo del Arte Zapoteca, Mitla, Oaxaca.

pala have the nosepiece in the form ofa horn on the head,
it is the tail that forms the nosepiece in the deer efligy
snuffer from Oaxaca. Such choices on the part of the
prehistoric pipe makers cannot be considered to be arbi-
trary. If one may venture some guesses, the horn atop
the head as nosepiece may relate to the well known and
widespread concept of horns—both single and double—
as asymbol and even a source of shamanic or supernatu-
ral power (I*urst 1965). Single horns on the forehead are
a characteristic especially ofa certain class of Colima figu-
rines, but they also occur elsewhere in Mesoamerica (e. g.
at Tlatileo, Tlapacoya, Chalcatzingo, Nochipala, Monte
Alban, ete.) and even in Peru, especially in Nazea art.
As for the nosepiece of the deer effigy pipe, this might
have to do with the concept of the deertail as magical
power object in some North American shamantic prac-
tices and beliefs. Among the Huichol, for example, the
deertail is an important element in the shaman’s equip-
ment, as it is in Papago shamanism. Likewise, it is
hardly insignificant that the name of one of the principal
Huichol supernaturals is Tamatst Mava Kwaat, Elder
Brother Deer ‘Tail.

The association of deer, divine inebriant and shaman
which we perceive archaeologically in the Monte Alban
snuffing pipe and ethnographically in Huichol and Cora
religion, is itself an important culture-historical problem
that remains to be seriously explored. Andean art dating
to the fifth or sixth century A.1). suggests that there
was something very like these Mesoamerican associations
also in Peru. A common theme on Moche IV painted
ceramics is a ritual deer hunt, in which the hunter is
clearly not meant to be an ordinary man but a god, cul-
ture hero, or great shaman (Plate IX). Moche painters
consistently depict the deer in association with a shrub
or tree which, though to some degree conventionalized,

[ 20 |

PuLaTE VIII

Effigy snuffing pipe of burnished grey clay in the form of a deer
holding a peyote cactus in his mouth, from Monte Alban, Oaxaca,
Late Formative, Monte Alban I-II, ca. 300-100 B.C. The erect tail
of the animal forms the nosepiece. Anon. private collection. L. 43’.

[ 21 |

is identifiable botanically as cfmadenanthera colubrina,
with the long, bean-like seed pods characteristic of the
hallucinogenically rich family Leguminosae. The seeds
of this tree, called vilea or willka in the Andes, are made
into a potent psychotomimetic snuff; they are also in-
gested in a beverage and, in some highland Quechua
villages, play an important role in the making of //ampu,
a sacred substance used in cattle increase ceremonies and
other rituals the origins of which lie far back in Andean
prehistory (Billie Jean Isbell, personal communication).

In any event, the deer is often a semi-divine celestial
animal for American Indians, connected with Sun, Fire,
sky beings, and shamans. Among the Warao of the
Orinoco Delta, its flesh is still strictly taboo for shamans,
suggesting at least a former sacred relationship (Johannes
Wilbert, personal communication). Among the Huichol,
it is the shaman’s spirit helper and companion: a pair of
feathered ceremonial arrows that he wears on his head
in certain ritual contexts symbolize deer antlers; the
oblong basket of shamanic power objects (takwatsi) 1s
identified with the divine Deer Person, MKauyumarie;
certain deities are deer and vice versa; the deer is mount,
guardian and guide on the shaman’s celestial quests and
fights, especially on the peyote hunt; the ‘*Principal
Deer’, Elder Brother Waiwatsari, is peyote, and vice
versa, etc.

Such concepts remind one at once of the role of the
deer in Paleo- Asiatic or Siberian shamanism. In Siberia,
too, the deer is the celestial mount that carries the sha-
man to the Upperworld and its spirit rulers. In parts of
Siberia, moreover, there is direct association between
deer and the divine inebriant
used by shamans to attain the ecstatic trance states in
which they embark on their supernatural journeys—the

in this case the reindeer

Amanita musearia, or fly agaric mushroom, for which

[22]

PuatE [LX

Supernatural deer hunting scene on a Mochica IV terracotta vessel,
northern Peru, ca. A.D. 500. Drawn from the original by Alan
Sawyer. Here, as in similar Mochica deer hunting scenes, the animal]

is shown in association with a tree or shrub almost certainly to be
identified by its characteristic seed pods as Anadenanthera colubrina,
whose seeds were, and still are, ground into a potent psychotomimetic
snuff, widely known as vi/ea or willka. In some Andean communities
vilca seeds are also used in an intoxicating ceremonial drink; else-
where, as in Chile, the pods may be burned and the smoke inhaled
to achieve ritual intoxication.

the reindeer is said to have an inordinate predeliction
(Wasson 1972:204) and which some scholars regard as
the Paleolithic or Mesolithic prototype for the Meso-
american mushroom cults.

It is difficult to escape the conclusion that the esteem,
not to say veneration, with which some American In-
dians regarded the deer represents a survival from an
ancient, archaic, shamanistic substratum—a substratum
that forms the underlying basis of American Indian
ideology, including that of Mesoamerican civilization,
and the ultimate roots of which he in the religion of
Eurasian Paleolithic and Mesolithic hunting and gather-
ing culture. ‘The curious association of deer as celestial
mount on the shaman’s ecstatic Journeys and the sacred
hallucinogens that are employed as aids in such mystical
quests in parts of northern Asia as well as in America
might well be a part of this very ancient belief system.

In this connection, a new series of radiocarbon dates
from rock shelter sites in Trans-Pecos Texas and north-
ern Mexico is of special significance. These dates, for
which IL am indebted to J. M. Adovisio of the University
of Pittsburgh‘, confirm a time depth of over ten thousand
years for the use of the potent hallucinogenic red seeds
of the Sophora secundifiora shrub by Desert Culture
hunters and food collectors as well as historic tribes in
the same area. More than that, one important and well
studied Texas site, known as Bonfire Shelter, yielded
Sophora secundiflora from the lowest occupational stra-
tum—Bone Bed II, witha C" age of 8440 to 8120 B.C.,
in direct association with Folsom and Plainview pro-
jectile points and the bones of extinct bison. The same

'A4 short paper on the topic by J. M. Adovisio and G. F. Fry was
presented at the 71st Annual Meeting of the American Anthropol-
ogical Association, Toronto, Canada, November 1972. A fuller treat-

ment by the authors is in preparation.

[24]

seeds also occurred in the topmost occupational level,
dated A.1). 420 to 1040, and in all of the intervening
cultural deposits. A related rock shelter site in northern
Mexico, Frightful Cave, similarly yielded Sophora se-
cundifiora beans from its lowest level, dated at 7500 B.C.,
through all subsequent cultural deposits. Interestingly
enough, here, as in many other Desert Culture rock
shelter sites, Sophora secundiflora was invariably asso-
ciated with Ungnadia speciosa, in contexts strongly sug-
gesting, according to Adovisio, ritual use.

It was Weston La Barre who suggested on several
occasions (e.g. 1972: 270-278) that the origins of the
American Indian hallucinogenic complex had to be
sought ultimately in ecstatic, vision-seeking Paleoasiatic
shamanism, the fundamental religion of the pre-
agricultural Paleolithic and Mesolithic hunting peoples
of Siberia who presumably constituted the ancestral pool
from which flowed the Late Pleistocene migrations into
North America. La Barre’s contention, which came to
be increasingly shared by some of his colleagues in the
study of aboriginal American religion and the botany and
anthropology of hallucinogens, thus appears to be con-
firmed: the historic shamanistic ‘‘red bean”’ cult of the
Southern Plains is at least as old as the big-game hunt-
ing phase of the terminal Pleistocene and thus appears
to reach back toward a time when the peopling of North
America across the Bering land bridge might still have
been in progress. If knowledge of ritual or divine in-
ebriation with plant hallucinogens was part and parcel
of the intellectual baggage of these early migrants, it
implies at least equal antiquity, if not a much greater
one, for such practices in Eurasia.

Just when or where in prehistoric antiquity the tech-
nology and chemistry of snufling might have arisen is
unknown. The concentration of these variants on the

[ 25 |

common theme of divine inebriation especially in the
Amazon suggest that area as its ultimate source. If it
did indeed diffuse from the tropical lowlands, its age
would have to be at least three and a half thousand years,
since, as mentioned above, the earliest known snufhing
implements date to the mid-second millenium B.C.
The fact that in addition to its numerous other psycho-
tomimetics prehispanic Mesoamerica can now be shown
to have shared in a wider pan-.American complex involv-
ing the use of snuff opens up a host of new possibilities
for culture-historical and ethnobotanical research. Ob-
viously, multidisciplinary study of what now appears to
have been a Mesoamerican snufling complex of substan-
tial distribution and duration is important in and of itself.
At the same time, it might provide answers to questions
of external relationships, especially with South America,
at roughly that crucial moment in time when Neolithic-
type farming communities became transformed, by some
as yet little understood processes, into the first great
Mesoamerican civilization—that of the Olmec.

REFERENCES

Adovisio, J.M. and G.F. Fry. 1972. “‘Prehistoric Psychotropic Drug
Use in Northeastern Mexico and Trans-Pecos Texas.”’
pared for the 71st Annual Meeting of the American Anthropological]
Association, Toronto, Canada, 1972.

Paper pre-

Furst, Peter T. 1965. “West Mexican Tomb Sculpture as Evidence
for Shamanism in Prehispanic Mesoamerica.’” Anthropologica 15:
29-60. Caracas.

——, 1968. ‘The Olmec Were-Jaguar Motif in the Light of Ethno-
graphic Reality.’’ In: Dumbarton Oaks Conference on the Olmec,
Elizabeth P. Benson, ed., pp. 143-178. Washington, D.C.: Dum-
barton Oaks Research Library and Collection. Trustees for Har-
vard University.

——, 1972. “Ritual Use of Hallucinogens in Mesoamerica: New
Evidence for Snuffing from the Preclassic and Early Classic.’’ In:
Religion en Mesoamerica, X11 Mesa Redonda, Sociedad Mexicana
de Antropologia, pp. 61-68.

Gay, Carlo T.E. 1972. NXochipala: The Beginning of Olmec Art. The
Art Museum, Princeton.

Hissink, Karin, and Albert Hahn. 1961. Die Tacana: Ergebnisse der
Frobenius- Expedition nach Bolivien, 1952 bis 1954. Stuttgart: W.
Kohlhammer.

La Barre, Weston. 1972. ‘‘Hallucinogens and the Shamanic Origins
of Religion.’’ In: Flesh of the Gods: The Ritual Use ef Hallucino-
gens, Peter ‘I’. Furst, ed., pp. 261-278. New York: Praeger Pub-
lishers, Inc.

Reis Altschul, Siri von. 1972. The Genus Anadenanthera in Amerin-
dian Cultures. Cambridge: Botanical Museum of Harvard Univer-
sity.

Schultes, Richard Evans. 1972a. “‘On Overview of Hallucinogens in
the Western Hemisphere.’’ In: Flesh of the Gods: The Ritual Use
of Hallucinogens, Peter 1. Furst, ed., pp. 3-54. New York: Prae-
ger Publishers, Inc.

——, 1972b. “Hex Guayusa from 500 A.D. to the Present. Etnolo-
gisker Studier, No. 32, pp. 115-138. Stockholm.

[27 ]

r 2 + ~~ © Ory , wi * , 0
Wassén, S. Henry. 1965. I'he use of some specific kinds of South

American snuff and related paraphernalia.’’ Etnologisker Studier,
No. 28, Stockholm.

—-, 1967. “‘Anthropological Survey of the Use of South American
Snuffs.”’ In: Ethnopharmacologic Search for Psychoactive Drugs,
Daniel H. Efron, ed., pp. 233-289. Washington, D.C.: U.S. Pub-
lic Health Service Publication No. 1645.

1972, “The Anthropological Outlook for Amerindian Medi-
cinal Plants.’? In: Plants in the Development af Modern Medicine,
Tony Swain, ed., pp. 2-65. Cambridge, Harvard University Press.

99

Wasson, R. Gordon. 1972. *“What was the Soma of the Aryans?”’

In: Flesh of the Gods: The Ritual Use of Hallucinogens, Peter T.
Furst, ed., pp. 201-213. New York: Praeger Publishers, Inc.

otk
oe

_-<BOTANICAL MUSEUM LEAFLETS
aa HARVARD UNIVERSITY

Campringr, Massacnusertrs, Ocroper 25, 1974 VoL. 24, No. 2

A STUDY OF SYSTEMATIC WOOD
ANATOMY IN CANNABIS
BY
Loran C. ANDERSON!

Cannabis has been associated with man since very early
times (Ash, 1948), yet, surprisingly, little is known about
its comparative wood anatomy. The reasons are due
probably to the tendencies for (1) anatomists to select
wood from trees and woody shrubs rather than from
herbs for study and for (2) researchers often to disregard
or slight plants associated with man, either as crops or
weeds, in basic scientific enquiries.

Tippo (1988) offered a few general comments on the
wood of C. sativa L. in his extensive study on the anato-
my of the Moraceae and its allies. Stem shape and leaf-
trace number in transections were stressed by Nassonov
(1940) in a report on geographical races of hemp. Hay-
ward (1948) devoted a chapter in his textbook to C.
sattva. The general morphology of that species was
given, but details of seedling anatomy and floral struc-
ture were emphasized; wood anatomy was scarcely men-
tioned. Metcalfe and Chalk (1950) summarized anatomi-
‘al data on Cannabaceae to that date. Shimomura e¢ al.
(1967) emphasized trichomes in their study of leaf and
bract anatomy in Cannabis; they found differences be-
tween C. sativa and C. indica.

' Department of Biological Sciences, Florida State University, Tal-
lahassee, Florida.

FCONOMIC BOTAI?
OF CAKUS &

CWT DA PTT OAT

Rh het nd hae Ue

Some features of Cannabis anatomy are relatively well
known, suchas the economically important phloem (bast)
fibres. These aspects have been reviewed by Hayward,
1948: Metcalfe and Chalk, 1950. Considerable attention
has also been given to cystolithic hairs (Pireyre, 1961)
and laticifers in Cannabis.

With the recent attention devoted to taxonomic prob-
lems in Cannabis (Schultes et al., 1974: Stearn, 1974),
I am pleased to present this introductory account on
comparative wood anatomy. It includes apparently the
first technical description of wood identified with vouch-
ered material as C. ¢ndica Lam.

Merruops AND Matrerias

All materials were collected fresh and preserved in
formalin-propriono-alcohol (FPA). Woods were sec-
tioned on a sliding microtome at 20 h. Some sections of
cach sample were stained in safranin © and counter-
stained with fast green FCF and orange G: others were
stained only with safranin. ‘Tissues were mounted in
Permount.

Xylem features were microscopically measured with a
‘alibrated ocular micrometer: a minimum of 50 measure-
ments were made for each feature reported in ‘Table 1.
Polarizing filters aided study of cell wall structure and
crystals. Statistical analyses were made on a Wang 600
computer with the assistance of Dr. M. P. Johnson.

The material of C. indica came from a wild population
at Pashimool, west of Kandahar, Afghanistan, R. 1.
Schultes 26505 (Keon. Herb. Oakes Ames); that of C.
sativa came trom a naturalized population in Pottawato-
mie County, Kansas, United States, 1. C. Anderson
3668 (Fla. State Univ.).

Resuirs
Details of wood anatomy are illustrated in Figs. 1-6,

30 |

The woods of C. indica and C. sativa differ significantly
in each feature listed in Table 1.

Vessels in C. indica tend to occur in radial chains;
whereas those of C. sativa usually occur singly (as illus-
trated in Hayward, 1948). That difference in distribution
can be seen in Figs. 1-2. Vessel members are angular
to round in transection. ‘They have simple perforation
plates, and the end walls are slightly oblique. Pits are
alternate with elliptic borders. Pit apertures are elon-
gate; they are 6-9 » long in C. indica and 4-8 p» in C.
sativa.

Vessel members and wood fibres differ between the
two samples in average width, length and cell wall thick-
ness (Table 1). In C. indica, both cell types are wider,
have thicker walls, but are shorter in length compared
to those of C. sativa.

Fibres in the secondary xylem must not be confused
with the hemp fibres of commerce, which are phloem or
bast fibres. Wood fibres of C. indica are typical, lignified
libriform fibres. Fibres in C. sativa differ in two respects.
They are dimorphic, with successive tangential bands of

Taste 1. Averaged measurements on wood anatomy in Cannabis.

Feature C. indica C. sativa Significance
level@
vessel number per group 3.05 1.39 ”
vessel member width, / 68.52 62.16 *
vessel member wall thickness, 4 3.50 2.30 =
vessel member length, 209.71 244,54 at
fibre width, /@ 18.41 14,28 eK
fibre wall thickness, - 3.44 0.68 =
fibre length, » 281.10 443.47 sii
ray width (cell number) 2.23 1.63 a
ray height, mm 0.87 0.68 *

* Analysis of variance (F test): * = p < .05, ** = p < .001

[ 31 ]

EXPLANATION OF PLATE X

Figs. 1-6. Cannabis wood sections. Figs. 1,3, 5 are C. indica, and
2,4,6 are C, sativa. Fig. 1, transection showing radial chains of ves-
sels, libriform fibres, and procumbent ray parenchyma. Fig. 2, tran-
section showing tendency for solitary vessels, fibre dimorphism, and
ray parenchyma shorter radially (erect). Fig. 3, transection showing
thick walls of vessel members and libriform fibres ; note cuboidal crys-
tals in ray cells (arrows). Fig. 4, transection showing relatively thin
cell walls; note shrunken secondary walls of gelatinous fibres. Fig. 5,
tangential section showing wood rays with numerous crystals; photo-
graphed with partially polarized light. Fig. 6, tangential section
showing relatively narrower wood rays with erect cells; crystals ab-

sent. Figs. 1-2, 5-6, < 71. Figs. 3-4, < 308.

[ 82 |

PLATE X

C. sativa

indica

C.

*
‘
s
*

e be

2

ob
5

See re ©

t
‘

}

*

cnn nen EE
oem

[vceeteniessnsies ceeessemeeenttaiadidiiteensaiensanediiineeneenee

onal Le aed

thick-walled fibres alternating with bands of thin-walled
fibres. They have irregularly shrunken secondary walls
(more pronounced in the thick-walled fibres) and are
termed gelatinous (Fig. 4). Their staining reaction (note
lighter tones in Figs. 2, 4, 6) and absence of birefringence
under polarized light are similar to that of gelatinous
fibres in other species that I have studied (Anderson,
1963, 1972).

Axial parenchyma is paratracheal. It is very scanty in
C. indica and scanty to vasicentric in C. sativa.

Wood rays are classed as heterogeneous I: ¢.e., both
multiseriates and uniseriates occur, and they are com-
posed of procumbent and erect ray cells. Those of C.
indica are predominantly square to procumbent ; whereas
ray cells in C. sativa are mostly erect with very few square
or procumbent ones. The differences in cell shape are sug-
gested in Figs. 5-6, but they are best viewed in radial
sections. A qualitative difference in wood rays is the
presence of numerous cuboidal or prismatic crystals of
‘calcium oxalate in C. indica. They can be seen in all sec-
tions under normal light but are more obvious with par-
tial polarization of light (Fig. 5). No crystals were found
in C. sativa ray cells (although both species have druses
in their phloem and ground tissues).

Discussion

Many American botanists have thought Cannabis to
be monotypic, possibly because only hemp, C. sativa,
has been cultivated in this country. Most taxonomists
who have studied the genus closely, however, recognize
three species: C. indica Lam., C. ruderalis Janisch., and
C. sativa L. (see Schultes et a/., 1974, for a review of the
taxonomic history of the genus).

Data from wood anatomy have not hitherto been uti-
lized in the taxonomy of Cannabis. Such data might

[ 34 |

help resolve the question of species recognition in the
genus. Nassonov’s study (1940) is of little use, as he
mentioned no binomials. He primarily studied variation
in crop plants (all C. sativa’), where he identified three
basic types of stem structure. He did note that wild and
cultivated forms of hemp could not be distinguished
clearly on the basis of anatomy of stem and bast fibres.

Wood features of C. indica and C. sativa listed in
‘Table 1 are those commonly measured in comparative
studies. They are all significantly different between the
species with four at the 5% level and five at the 0.1%
level! Additional differences in the axial and radial paren-
chyma systems are noted in the text. Woods of the
two species are qualitatively distinct for libriform fibres
versus gelatinous fibres and for presence of crystals in
wood rays. Many examples of the taxonomic signifi-
‘ance of crystals in woods have been noted (Bailey, 1961:
Chattaway, 1955-56).

Although only one sample of each species is discussed
here, the magnitude of differences between the two is
impressive in a system as conservative as wood. In his
exhaustive review on many aspects of wood science, Jane
(1963) stated the following regarding taxonomic wood
anatomy:

Wood structure is probably more conservative than floral struc-
ture, and specific differences, as determined by floral characters,
are often not reflected in the secondary xylem. Indeed, it may
be said that in general the distinguishing features of wood are

at generic, rather than specific, level.

Certainly, the plants used in this study are of the same
genus, but it is my opinion that they represent different
species.

examination of woods from three additional collec-
tions of North American C. sativa shows they are also
distinct from the C. 7ndica wood sample. All vary from

C. indica in the features listed in Table 1, with the ex-
ception of vessel member width. All three samples have
gelatinous fibres. Crystals are absent in wood rays in
two: only a few were found in the rays of the third
sample. Complete data on these samples will be pre-
sented in the future as part of an expanded study on the
wood anatomy of Cannabis.

LITERATURE CITED

Anderson, L.C. 1963. Studies on Petradoria (Compositae): anatomy,
cytology, taxonomy. Trans. Kans. Acad. Sci. 66: 632-684.

—, 1972. Studies on Bigelowia (Asteraceae). I]. Xylary compari-
sons, woodiness, and paedomorphosis. J. Arnold Arb. 53: 499-
Olt,

Ash, A.L. 1948. Hemp —production and utilization. Econ. Bot. 2:
158-169.

Bailey, I.W. 1961. Comparative anatomy of the leaf-bearing Cacta-
ceae, III. Form and distribution of crystals in Pereskia, Pereskiop-
sis and Quiabentia.

Chattaway, M.M. 195
Trop. Woods 102: 5

Hayward, H.E. 1948.
Co., New York.

Jane, F.W. 1968. Botanical aspects of wood science. /n: W.B. Tur-
rill, ed., Vistas in Botany, Vol. I]. Pergamon Press, Oxford.

Metcalfe, C.R., and L. Chalk. 1950. Anatomy of Dicotyledons, Vol.
II. Oxford Univ. Press, London.

Nassonov, V.A. 1940, “Anatomical characteristics of geographical
races of hemp’’ [in Russian]. Vestnik Sotsialisticheskogo Rasten-
ievostva, Institut Rastenievostva, Moscow, 4: 107-120.

Piereyre, N. 1961. Contribution a l°étude morphologique, histologi-

56. Crystals in woody tissues. I and II.
74, 104: 100-124.
he Structure of Economic Plants. Macmillan

5
9)
T

que et physiologique des cystoliths. Rev. Cytol. et Biol. Vég. 23:
93-320.

Schultes, R. K., W.M. Klein, T. Plowman, and T. EK, Lockwood.
1974. Cannabis: an example of taxonomic neglect. Harvard Univ.
Bot. Mus. Leafl. 23: 337-367.

Shimomura, H., M. Shigehiro, KE. Kuriyama, and M. Fujita. 1967.
Studies on Cannabis. 1. Microscopical characters of their internal
morphology and spodogram. J. Jap. Pharm. Soc. 87: 13834-1341.

Stearn, W.T. 1974. Typification of Cannabis sativa L. Harvard Univ.
Bot. Mus. Leafl. 23: 325-336.

Tippo, O. 1938. Comparative anatomy of the Moraceae and their

presumed allies. Bot. Gaz. 100: 1-99.

TWO NEW BRAZILIAN SPECIES
OF BRUNFELSIA
BY
Timotuy C. PLowMaAaNn

In anticipation of my forthcoming revision of the
South American species of Brunfelsia (Solanaceae) based
on a doctoral dissertation at Harvard University (Plow-
man, 1973), [ herewith offer the description of two previ-
ously unrecognized species.

I would like to express my thanks to the curators of the
several herbaria who have graciously loaned specimens
cited in this paper. Abbreviations of herbaria are taken
from Index Herbariorum (I.anjouw and Stafleu, 1964).
Special thanks are due also to Dr. Leslie Garay for
checking the Latin descriptions and to Lynda Bates for
preparing the line drawings which accompany the text.

Research reported in this paper was supported by the
National Institutes of Health Training Grant (T TO]
GM 00036-13) and by the National Science Foundation
Evolutionary Biology Training Grant (GB 7346, Reed
Rollins, Principal Investigator, Harvard University).

Brunfelsia Martiana Plowman sp. nov.

Frutex. Folia subsessilia, oblonga, plerumque elliptico-
oblonga vel oblongo-lanceolata, apice acuminata, basi late
cuneata vel obtusa, utrinque glabra, nervis lateralibus
8-12, petiolo brevissimo. Inflorescentia terminalis vel

[ 37 ]

L

axillaris annotinis ramulis, pauciflora, glabra. Flores 1-7,
albidi vel albo-virescentes. Pedicellus erectus, gracilis,
tlaber. Calyx tubulosus vel tubuloso-campanulatus,

(
Sb
(

|
glaber, dentibus subaequalibus, triangulari-ovatis vel
ovato-lanceolatis, apice acutis vel acuminatis. Corollae
tubus quam calyx duplo longior, rectus, apice parum in-
flatus, glaber, limbo patent, lobis subaequalibus, oblongo-
obovatis, apice rotundatis vel subtruncatis, marginibus
lateralibus abrupte reflexis.. Stamina inclusa in tubi parte
superiori. Ovarium oblongo-ovoideum, Fructus ignotus.

A Brunftelsia guianense foliis oblongis, longioribus,
nervis lateralis pluribus (8-12), calyce tubuloso-campan-
ulato et corollae tubo quam calyce duplo longiori et a
B. amazonica foliis majoribus et pedicellis multo breviori-
bus differt.

Tyre: Brazil: Amazonas: ‘‘Provinciae Rio Negro, habitat in sylvis
ad flum. Japura, prope Sao Joao do Principe,’ Dec. 1819, Martius
[3247] (holotype, M; isotypes, M).

Shrub to 1m. tall. Branches few, naked below, some-
what knobby at nodes, glabrous. Bark cracked longitu-
dinally, dark reddish brown, shiny, furnished with len-
ticels. Leaves scattered along branchlets, subsessile,
mostly oblong, sometimes elliptic-oblong or oblong-
obovate, acuminate at apex, broadly cuneate to blunt
at base, blade 10-25 cm. long, 4-8 cm. wide, glabrous on
both surfaces, firmly membranaceous to subcoriaceous,
upper surface dark green, dull or shiny, lower surface
paler green, sometimes shiny, midrib dark reddish brown,
lateral nerves 8-12 pairs, spreading, straight: petiole
short, 1-4 (12) mm. long, glabrous, becoming cracked,
rugose. Inflorescence terminal or axillary in the upper
leaf axils, glabrous, axis 2-10 (15) mm. long. Flowers
1-7 per inflorescence, sometimes with short peduncle
which articulates with pedicel, greenish white to white.
Bracts 1-8 per flower, linear-lanceolate, concave, 1-10

PLATE XI

| Sic spigec i p

Per 9? she

TYPE

HOLOTL EE festeton of Brvntzfee ‘ Soinne
Brusfelela gent jan

| Me Meirima Sore sey A | 2 a8

Photograph of holotype of Brunfelsta Martiana Plowman

K.F.P. von Martius [3247].

mm. long, ciliolate at margin, caducous. Pedicel short,
slender, 8-6 (10) mm. long, glabrous. Calyx tubular to
tubular-campanulate, 8-12 mm. long, 8-8 mm. in di-
ameter, glabrous, rarely striately veined, teeth subequal,
triangular-ovate to ovate-lanceolate, 2-6 mm. long, acute
to acuminate at apex. Corolla tube twice as long as
calyx, straight, cylindric, inflated at apex, 20-24 mm.
long, 1-8 mm. in diameter, glabrous; limb spreading,
somewhat undulate, 15-22 mm. across, lobes subequal,
oblong-obovate, abruptly reflexed at lateral margins,
rounded-truncate at apex, 5-10 mm. long. Stamens in-
cluded within upper part of tube; filaments subligulate,
longer anterior pair 4 mm. long, shorter posterior pair
3 mm. long; anthers orbicular-renitorm, slightly un-
equal, the upper pair somewhat smaller, about 1 mm. in
diameter. Ovary oblong-ovoid, 2 mm. long: style fila-
mentous, a little broader and curved at apex, about 15
mm. long: stigma briefly bifid, upper lobe slightly
larger, 1 mm. long. Fruit and seed unknown.

Disrrrpurion. Brazil (Amazonas, Para, Bahia); Guyana.

AppITIONAL specIMENS. Brazit: Para: Faro, 22 Jan. 1910, Ducke
s.n. (MG): Bella Vista, Rio Tapajoz, matta das immediacoes do
Campina do Perdido, 12 Feb. 1917, Ducke s.n. (MG): Regiao do
Jutahy de Almeirim. Palhal, 16 Apr. 1923, Ducke s.n. (RB); Gurupa,
25 Feb. 1923, Ducke s.n. (RB); Parad, May 1819, Martius 3300 (M).
AMAZONAS: Manaus and vicinity, road Manaus-Caracarai, km. 22,
21 Mar. 1967, Prance et al., 4704 (WIS), Rio Negro, Cachoeira Baixa
de ‘T'arumao, 11-14 Apr. 1973, Schultes & Rodrigues 26132-A (ECON),
26188 A(ECON). Banta: Distrito Ileus, Ferradas, Dec. 1818, Mar-
tius s.n. (M): Feira de Santana, Apr. 1850, collector unknown (G).

Guyana: Essequibo-Demarara River, Madray-Bubu Trail, Topy
Trysil Forest, 8 Feb. 1944, Forest Department 4422 (K).

Brunfelsia Martiana is named for K.F.P. von Martius
who first collected the plant in the year 1818. This new
specific concept has proven difficult to define, not through
its lack of distinctive characters, but due to the scarcity
of specimens which are scattered in Brazilian and Euro-
pean herbaria.

[ 40 |

Piate XII

BRUNFELSIA = martiana Plowman

Brunfelsia Martiana Plowman. 1, flowering branch, one third natural size. 2,
corolla limb from beneath, approximately natural size. 3, calyx showing attach-
ment to stem and bracts, approximately natural size. 4, excised corolla tube
showing stamens and pistil, one and one-half times natural size.

This interesting shrub grows in the understory of low-
land tropical forests from Guyana throughout the lower
Amazon basin and with disjunct populations in the
coastal forests of the State of Bahia. The species has
been collected only twice in the last fifty years, and no
collections of it are known from Bahia since 1850.

Brunfelsia Martiana is closely related to B. amazonica
Morton and B. guianensis Benth., which also have
rather small, greenish white flowers and occur sym-
patrically with 2B. Martiana in the middle Amazon.
B. Martiana differs from 2B. amazonica in having larger
leaves (10-25 em. vs. 6-12 em.) with more lateral nerves
(8-12 pairs vs. 5-8 pairs) and much shorter pedicels
(8-6 mm. vs. 13-20 mm.). It differs from B. euianen-
sis in having larger (10-25 cm. vs. 6-15 em.) oblong
leaves with more lateral nerves (8-12 pairs vs. 4—5 pairs),
a calyx which is usually narrow-tubular, not ovoid-
‘ampanulate, and a corolla tube which is rarely more
than twice as long as the calyx. In B. euranensis, the
tube is frequently 2) to 8 times as long as the calyx.

Brunfelsia pilosa Plowman sp. nov.

Krutex. Ramuli graciles, pilosi vel villosi. Folia breve
petiolata, anguste elliptica, oblongo-lanceolata vel obo-
rata, apice acuminata, basi cuneata, utrinque ad costam
pilosa, nervis lateralibus 6-9. Inflorescentia terminalis,
sessilis. Flores 1-8, speciosi. Pedicellus brevissimus,
apicem versus incrassatus, pilosus. Calyx tubuloso-
ventricosus, in exsiccatione campanulatus, sparse vel
dense pilosus, raro glaber, membranaceus, dentibus sub-
aequalibus, lanceolatis, acuminatis; calyx in fructu per-
sistens, quam capsula fere aequans. Corollae tubus quam
‘calyx duplo longior, glaber, limbo patenti, lobis sub-
aequalibus, rotundatis, lobo superiori posterior! parum
majori. Stamina et stigma in tubi parte superiori in-

[42]

Piare XIII

Photograph of holotype of Brunfelsia pilosa Plowman, L.B. Smith
& Pe. R. Reitz 12777.

»

=
cs
.
Ne
ee

clusa. Ovarium conico-ovoideum, Capsula ovoidea vel
subglobosa, apice apiculata, laevis, atroviridis. Semina
ellipsoides, reticulato-foveata.

A Bruntelsia cuneifolia calyce tereti pubescentia pilosa
et corollae tubo longiori, a B. uniflora calycis dentibus
longioribus, pubescentia pilosa et corollae tubo longiori,
et a B. australe foliis apice acuminatis, pubescentibus
angustioribus differt.

Tyrer. Brazit: Santa Catarina: Municipio Sao Miguel d’Oeste,
forest above Rio Reperi-guacti, Peperi, ca. 26°32’ S., 53°44’ W.,
300-400 m., 21 Oct. 1964, LL.B. Smith & Pe. R. Reits 12777 (holo-
type, GH; isotypes, MO, R, UC).

Shrub 0.5—-2 m. tall, diffusely branched from near
base. Branches spreading and arching, terete. Bark
rough, longitudinally cracked, shedding in thin, char-
taceous flakes, yellowish brown. Branchlets slender,
pilose to villous, greyish green, dark purple when young.
Leaves appearing two-ranked, scattered on branchlets,
narrowly elliptic, oblong-lanceolate or obovate, acumi-
nate at apex, cuneate at base, blade 8-7.5 cm. long,
1.8-8 em. wide, pilose on both surfaces, primarily at
midrib, firmly membranaceous, dull, dark green above,
pale green beneath, the young leaves dark purple, lateral
nerves 6-9 pairs, straight, often prominulous above;
petiole short, 1-4 mm. long, pilose to villous. Inflores-
cence terminal, sessile, usually with one flower, rarely
2-3. Flowers showy, deep violet fading to pure white
with age, odorless. Bracts 1-8, linear-lanceolate, 1-8
mm. long, pilose to villous, caducous. Pedicel short,
stout, 1-8 mm. long, pilose; in fruit becoming corky,
rugose-verrucose toward apex. Calyx tubular-ventricose,
appearing campanulate in pressed specimens, terete in
cross-section, 12-19 mim. long, sparsely to densely pilose
with long weak hairs, rarely glabrous, membranaceous,
purplish, drying reddish brown, teeth subequal, lanceo-

[ 44]

PLATE XIV

BRUNFELSIA pilosa Plowman

Brunfelsia pilosa Plowman, 1, flowering and fruiting branch, one-
half natural size. 2, corolla limb from above, approximately natural
size. Drawn from living plant in cultivation at Miami, Florida, Plow-
man 2963 (ECON).

(45 |

late, 4-10 mm. long, acuminate; calyx in fruit persist-
ent, partially enclosing capsule, becoming subcoriaceous,
light green, the sinuses becoming more deeply cut. Co-
rolla tube twice as long as calyx, 25-82 mm. long, 1.5-8
mm. in diameter, glabrous; limb spreading, 30-47 mm.
in diameter, thickening at mouth prominent, round and
white, lobes subequal, the uppermost posterior lobe
slightly larger, broadly rounded, 10-15 mm. long. Sta-
mens included in uppermost part of corolla tube; fila-
ments slender, upper pair 38-4 mm. long, lower pair 38-5
mm. long, glabrous; anthers orbicular-reniform, 1—1.5
mm. in diameter. Ovary conical-ovoid, 2—2.5 mm. long,
1 mm. in diameter; style slender, 22-26 mm. long: stig-
ma briefly bifid, in the form of a forceps, 1 mm. long.
Capsule ovoid to subglobose, apiculate at apex, 12 mm.
long, 10 mm. in diameter, smooth, dark green, shiny,
pericarp thin-walled, dry at maturity, tardily dehiscent.
Seeds ca. 10, ellipsoid, 5-6 mm. Jong, 2.5-4 mm. in
diameter, reticulate-pitted.

Distrisution. Brazit: (Sao Paulo, Parani, Santa Catarina, Rio
Grande do Sul); Paraguay ; Argentina (Misiones).

AbbDITIONAL specIMENS. Brazit: Sao Paulo: Itapetininga, 17 Nov.
1887, Léferen 389 (C, R, SP, US); Ypiranga, Nov. 1910, Liiderwaldt
2717 (RB, SP); Campinas, Jundiahy, Mar. 1900, Campos Novaes 216
(US); Villa de Serra Branca, 1 Nov. 1897, Puttemans 4328 (SP);
Carandirti, Dec. 1912, Tamandaré 244 (RB); Mandaquil, 23 Nov.
1906, Usteri s.n. (G), Parana: Curitibi, Bairro Sao Niolau, 18 Nov.
1966, Capriliont 1659 (US); Itaperussti, 17 Nov. 1908, Dusén 7077
(GH, S); Jaguariahyva, 25 Oct. 1910, Dusén 10443 (GH, MICH,
NY, S): Tamandaré, 4 Oct. 1914, Fénssen 1054a (F, GH, NY, S$);
Sao Matheus, 27 Feb. 1929, Gurgel s.n. (RB): Municipio Rio Branco
do Sul, Sao Vicente, 27 Oct. 1967, Hatschbach 17610 (C, UC); Ponta
Grossa, 2 Nov. 1928, Hoehne 23309 (SP, US); Foz do Iguacti, Parque
Nacional de Iguacti, 8 Oct. 1946, J.G. Kuhlmann s.n. (RB) :Ypiranga,
15 Sept. 1934, Reiss 99 (GH, NY). Santa Catarina: Municipio
Cacador, Fazenda dos Carneiros, 1100 m., 7 Dec. 1962, Aletn 3518
(US); Municipio Campos Novos, Palmares, 950 m., 28 Oct. 19683,
Klein 4102 (US); Bituruna, Fazenda Etienne, 11 Feb. 1948, Mello
Filho 793 (R); Nova Teutonia, 25 Oct. 1943, Plaumann 164 (RB);

[ 46 ]

Municipio Ararauja, Rodeio da Areia, 22 Nov. 1943, Reits C-171/
(RB); Santa Cecilia, 1100 m., Reitz & Klein 14136 (NY): Municipio
Sao Miguel d’Oeste, Canela Gaucha, 8 km. northwest of Sao Miguel
d’Oeste, ca. 26°40! S., 53°34’ W., 700-750 m. , Smith & Reitz 12757
(NY, R, US); Manic pio: Abelardo Luz, north bank of Rio Chapeco
at Abelardo Luz, ca. 26°35 S.; 52°20/ W., 900-1000 m., 28 Oct.
1964, Smith & Reitz 12870 (C, F, ae. R); Tubarao, Nov. 1889, Ule
1521 (P). Rio Grande do Sul: Sao Leopoldo, 10 Sept. 1946, Henz
27246 (MO, NY), 20 m., 17 Sept. 1946, Henz 35305 (S): Municipio
Rio Pardo, Fazenda Horticola, 70 m., Oct. 1923, Jiirgens 19 (B);
Colonia Santo Angelo, 4 Feb. 1898, Lindman A-597-b (S), Sept.
1900, Schwarzer 50(L,S); Silveira Martins, 20 Mar. 189 3, Lindmann
A-597-c (S); Rio dos Sinos, 8 Nov. 1949, Rambo 44295 (BR, L);
Bitterberg prope Montenegro, 13 Nov. 1950, Rambo poy (GH
P, W).

Paracuay: In regione fluminis Alto Parana, Yaguarazapa, 1909-
1910, Frebrig 5460 (G, GH, US).

ARGENTINA: Misiones: Departamento de San Javier, Acaragua,
220 m., 30 Sept. 1947, Bertoni 2974 (B, W); Departamento de Can-
delaria; Bompland, 4 Oct. 1909, Jorgenson s.n. (BAB), Nov. 1910,
Jorgenson s.n. (BAB); Santa Ana, Aug. 1901, Llamas 1530 (BAB,
G). Departamento de Iguazt, Cataratas de Iguazu, 13 May 1969,
Plowman 2735 (GH); Delicia, camino a El Dorado, 4 Nov. 1949,
Schwindt 2270 (C, LD, NY); Puerto Aguirre, Rio Iguazti, 19 Sept.
1922, IL.N.T.A. 44365 (B).

Brunfelsia pilosa is a distinct species which has been
consistently confused with several other plants of south-
ern Brazil, especially B. uniflora (Pohl) Benth. and B.
australis Benth. An early collection of B. pilosa (Sellow
1573) with atypically glabrous leaves and calyces was
cited by Bentham as a syntype of his concept of B.
australis (Bentham, 1846).

A small shrub with showy flowers, Brunfelsia pilosa
occurs primarily in the understory of primary and second-
ary woods, as well as in thickets and along water courses.
Flowering takes place from October to December with
fruits appearing from February to May. It has a fairly
wide altitudinal range, occurring from near sea level to
1100 m. ‘The species ranges from the Brazilian State of
Sao Paulo south to Rio Grande do Sul and extreme

[ 47 ]

northeastern Argentina (Misiones). In the central part
of its range, where it occurs most commonly, this species
seems to be associated with the formerly extensive 4 rau-
caria forests of the south Brazilian planalto.

Owing to its attractive flowers, this plant is cultivated
in tropical gardens, although not so extensively as Brun-
felsia australis. Recently, B. pilosa has appeared in the
Florida nursery trade and iS increasing in importance as
an ornamental.

Brunfelsia pilosa may be distinguished from related
species by its very short (1-8 mm. long), stout pedicels
and an inflated tubular calyx with deeply cut teeth
(4-10 mm. long). The twigs, leaves and calyx charac-
teristically contain few to many long, weak trichomes.
The flowers are large and showy with a corolla limb
nearly 5 cm. in diameter.

LITERATURE CITED
Bentham, G. m A. DeCandolle, 1846. Prodromus Systematis Natu-

ralis Regni Vegetabilis. 10: 200.

Lanjouw, J. and F.A. Stafleu, 1964. Index Herbariorum. 5th Fd.
Regnum Vegetabile Vol. 31. International Bureau for Plant Tax-
onomy and Nomenclature, Utrecht.

Plowman, T. 1978. The South American Species of Brunfelsia (So-

lanaceae). Doctoral Dissertation, Harvard University, Cambridge,
Massachusetts.

[ 48 |

BOTANICAL MUSEUM LEAFLETS

HARVARD UNIVERSITY

CaMBRIDGE, MAssacnuseTTs, JANUARY 10, 1975 Vou. 24, No. 3

ARCHAEOLOGICAL MAIZE FROM
NORTHERN CHILE
BY
Paunt C. Mancetsporr AND Gorpon C. PoLLarp*

INTRODUCTION

Now that the modern races of maize of the countries
of this hemisphere have been classified and described in
a series of eleven publications issued by the National
Academy of Sciences- National Research Council, it has
become a common procedure for botanists analyzing
collections of prehistoric remains of corn of a particular
country to relate the ancient specimens, so far as is possi-
ble, to the living races of that country. The modern
maize of Chile has been described by Timothy ef al.
(1961), who recognized nineteen more or less distinct
races.

As part of a study of prehispanic cultural development
in the Atacama Desert of northern Chile, the junior
author of this report obtained several collections of maize
cobs from archaeological sites found near the town of
Chiu Chiu, which lies beside the middle section of the
Loa River. The vicinity of Chiu Chiu, at an elevation
of 2500 meters, is a small oasis along the river which
supports a narrow band of vegetation in the extreme

* Associate Professor of Anthropology, State University of New
York College at Plattsburgh.

[ 49 ]

desert environment. The region around the town has a
long and almost unbroken archaeological sequence begin-
ning at least 10,000 B.C. and lasting until post-Conquest
times. A summary of the cultural sequence since the in-
troduction of ceramic technology (ca. 500 B.C.) is found
in Pollard (1971). The appearance of maize cultivation
in the archaeological record (ca. A.D). 100) is associated
with artificial irrigation technology.

MalIze FROM THE Earnikst SITE

The earliest site found with associated maize is desig-
nated RAnL 100. Although no prehispanic structures
are preserved, the site is believed to be the earliest agri-
cultural village settlement along the middle Loa River.
Sections of the site include a small intact refuse mound,
a looted cemetery, ancient agricultural terraces and irri-
gation canals, and ruins of a historic-period house. -A
1.0 & 2.5 m. cut was excavated in the refuse mound,
revealing five well separated, but culturally homogeneous
layers, the deepest of which (layer 5) was as much as 55
em. below ground surface. Each layer yielded an abun-
dance of sherds, various lithic artifacts, and = sizeable
quantities of Lama sp. bones. Layers 2-5 each con-
tained varying amounts of Lama sp. wool and mesquite
seeds (Prosopis chilensis). Layers 8 and 5 also contained
several gourd fragments (Lagenaria sp.). Twenty maize
cobs and several fragments were recovered from layers
2—5. Radiocarbon analysis of asample of ama sp. wool
and skin from layer 5 resulted ina date of A.D. 105 +
105 (GN-1644). At present, this is the earliest date for
maize agriculture in northern Chile.

Of the nineteen races of maize described by ‘Timothy
et al. (1961), we recognize three in the collection trom
site 100. Descriptions of these follow :

[ 50 |

Capio Chico Chileno

Twenty of the oldest cobs are considered to be related
to this modern race; their characteristics are set forth in
Table 1. All of the cobs represented in this table came
from the test excavation described above. Layer pro-
venience is represented by the last number in the code,
except that layer 2 is represented by two code numbers,
2 and 7. There is no clear relationship between the depth
of the layer and the characteristics of the cobs. Within
each layer the cobs are arranged in order of their lengths,
for identification purposes.

We call particular attention to the last cob listed in
Table 1. This was very small with a length of 2.6 cm.,
a diameter of 1.6 cm., and 12 kernel rows. The senior
author, having shortly before studying the Chilean speci-

TABLE 1

Characteristics of the cobs from the earliest site (RAnL 100),
dated at A.D. 105.

diam/length

code no. length(cem) diameter (cm) ratio kernel-row no.

100-2 Sit 1.8 AQ 22
100-3 3.6 Led 31 14
7 3.1 1.3 42 18
3.4 1.4 Al 16

7 2.1 1.5 72 16
100-7 6.3 1.8 .29 18
6.2 1.6 .26 12

7 6.1 1.7 .28 18

rs 5.3 1.1 21 12
100-4 5.6 2.1 37 18
5.4 1.3 24 16

? 57 1.6 .28 16
5.1 1.5 .29 16

100-5 7.0 2.8 40 30
: 5.2 2.0 38 18
4.2 1.6 38 16

4.1 1.9 46 22

“ 3.6 2.0 56 18

7 3.6 1:5. 42 16

“ 2.6 0.8 31 12

average 4.6 1.6 35 17.2

mens been engaged in studies of the prehistoric wild corn
of the Tehuacin Valley in Mexico, saw in this specimen
some resemblance to the Mexican corn; this raised in
his mind the question of whether there could once have
been a wild corn in Chile. Subsequent correspondence
with the Junior author made it clear that this part of
Chile could scarcely have provided a suitable habitat for
wild corn, and we concluded that this specimen is proba-
bly that of a stunted ear, borne on a depauperate plant.

K.xcept for their smaller size, these earlier Chilean cobs
(see Plate NV, fig. A) are quite similar in their charac-
teristics to cobs of the modern race Capio Chico. The
average data for their lengths, diameters, diameter/
length ratios, and kernel-row numbers are 4.6, 1.6, 0.35,
and 17.2 respectively. For their modern counterparts,
the corresponding averages, published by Timothy et a/.,
are 8.6, 2.38, 0.27, and 17.2 respectively. The similarity
in kernel-row numbers is especially significant.

In their dimensions, the Chilean cobs resemble even
more closely the prehistoric ears excavated by Dr.
Dwight Wallace from the Los Cerillos site in the Ica
Valley on the south coast of Peru, some 1100 km. north
of the Chilean site. This corn, estimated to be 2300-
2500 years old, has been briefly described by Grobman
et al, (1961) as a prehistoric race, Confite [queno; three
ears of this race are illustrated in their figure 19.

The specimens from the Los Cerillos site are unusually
well preserved, the kernels still being attached to their
cobs. To obtain estimates of the diameters of the cobs,
we subtracted from the diameters of the ears 9/10 of the
length of two average kernels, assuming that 1/10 of
their length, on the average, was embedded in the cob;
this figure is based on the data published by ‘Timothy
et al. tor Capio Chico.

Kistimates obtained in this manner from the Los Ceril-

los maize produced the following averages: 5.6, 1.8,
0.32, and 17.8. These are so similar to those set forth
above for the prehistoric Chilean cobs that we may con-
clude, with some degree of confidence, that the two
collections represent essentially the same race, Confite
Iqueno. By the same token, we can assume that if the
Chilean cobs had retained their kernels, the intact ears
would have been similar in their general appearance to
those of the Los Cerillos maize, illustrated by Grobman
et al. in their figure 19.

Polulo

This race is represented among the oldest cobs by a
single specimen which is so different from the remaining
ones that it must be considered as that of a different
race. This specimen is 3.4 cm. long, but it is probably
not intact with respect to length, since it lacks a pedun-
cle. Also, since it comprises only the rachis, the central
stem of the cob, the floral bracts having been lost, its
diameter/length is not comparable with other cobs in
this collection. The kernel-row number is 10; the cu-
pules are distinct, slightly longer than wide and are
hairy; stumps of the rachillae are prominent.

We might not have been able to identify this single
specimen had not we found counterparts of it in a collec-
tion from site RAnL 887-1. Cobs of that lot, illustrated
in Plate XV, fig. B, were derived from the shallow fill
within the foundation of an isolated house dating to the
late prehispanic period.

Anticipating a description of this lot, to be set forth
later, we can say that of the 115 cobs of site 8387-1,
thirty-five are of a very distinctive type. These have
slender rachises and peduncles of about the same diame-
ter as the rachises. The floral bracts that remain attached
to the rachises are relatively long; both lower and upper

Loe |

EXPLANATION OF THE PLATE

Fic. A. Three of the larger cobs from lot 100, dated at ca. A.D.
100. These are quite similar in the average dimensions to the pre-
historic Peruvian race Confite Iqueno, and are related to the modern
Chilean race Capio Chico Chileno. (Actual size. )

Fic. B. Three of the more slender cobs from lot 337-1, dated at
ca. A.D. 1250. These are related to the modern Chilean race Polu-
lo, which may in turn be related to the Peruvian popcorn Confite

Morocho. (Actual size. )

Fic. C. Two cobs from lot 290, dating to the late prehispanic period.
These are prehistoric counterparts of the modern Chilean race Chutu-
cuno Chico, which may be related to the Peruvian race Confite Puntia-
gudo. Note the stiff, curved lower glumes, a characteristic which may
have been derived from one of corn’s relatives, teosinte or 7'ripsacum.

(Actual size. )

[ 54 |

AX “LIV Td

glumes are herbaceous and not at all indurated. They
are typical of certain types of pod-corn involving lower
alleles at the pod-corn locus in combination with 7%, a
major tunicate-inhibiting gene (see Mangelsdorf 1974).

There is only one known modern Chilean race, Polulo,
to which these slender cobs can be assigned. This is a
finger-shaped popcorn with 10-16 kernel rows, grown at
approximately 2700 meters. Its glumes are soft, the
rachillae long and slender.

Timothy et a/. state that the race Polulo seems not to
be duplicated among collections made in other South
American countries, but there seems to us a possibility
that it is related to the Peruvian popcorn race Confite
Morocho. Indeed one of the ears of Polulo illustrated
in figure 5 of Timothy et al. resembles rather closely one
of the ears of Confite Morocho illustrated in figure 48 of
Grobman et a/. (1961). Both have about the same length,
9.1 and 8.5 cm. respectively; both have 10 kernel rows
and their rows are slightly irregular. One of the most
distinctive characteristics of the archaeological specimens
is their peduncles, which have about the same diameters
as the rachises. The data of Grobman ect al. show that
this is true of Confite Morocho.

Galinat (1969) has isolated from the Peruvian race
Confite Morocho types with cobs so slender that he calls
them ‘‘string cobs.” These slender cobs are found also
in the Peruvian race Rabo de Zorro, which Grobman
et al, regard as a hybrid derivative of Confite Morocho.
Moulds thought to represent ears of Rabo de Zorro occur
on a number of ceramic vessels of the Moche culture
(Early Intermediate Period); several of these are illus-
trated in Grobman et al. (figs. 81 and 82), and one in
Mangelsdorf (1974, fig. 17.18). These show that this
slender-cob trait in South American races is an ancient
one. The fact that it is not known in Mexican races may

[ 57

lend some support to the hypothesis (Mangelsdorf 1974)
that there may have been an independent domestication
of maize in South America. [t is in this connection that
the slender-cobbed Chilean specimens are of particular
interest.

Galinat (1972) has recently published a brief descrip-
tion of the archaeological maize turned up by MacNeish
in asitein Ayacucho, Peru. He states that some of the
cobs resemble those of the race Confite Morocho, but
these are not among the earliest cobs from this. site.
They are tentatively dated at 8000 B.P.

Chutucuno Chico

This race, described by Timothy ef af, appears to be
represented by one specimen obtained from near the top
of looters’ backdirt adjacent to the cemetery on site 100.

~

This single cob is 7.6 em. long: its most prominent
feature is the stiff indurated lower glumes.

This specimen could be a precursor of cobs with stiff
indurated lower glumes occurring in lots from. sites
RAnL 186, 268B, and 290, described briefly below.
These in turn appear to be related to the living Chilean
race Chutucuno Chico which is grown at altitudes of
2260 to 2500 meters. Timothy et al. describe this race
as a small yellow popcorn with large cobs having 16-22
or more kernel rows. Both red and white cobs occur in
the modern as well as the late prehistoric collections.

Chutucuno is in some respects similar to the fasciated
form of the Peruvian primitive race Confite Puntiagudo
described by Grobman ef a/. and illustrated in their fig.
56. These authors attribute the indurated tissues of the
rachis and lower glumes to hybridization with corn’s
relative T'ripsacum.

Maize Frou LATER Sires
RAnL 2
[ 58 |

This site is a cemetery of more than 100 niche and pit
graves at the locality of Chiu Chiu, and appears to have
been utilized ca. A.D. 800. All the burials were found
to have been looted, with broken artifacts and organic
remains, including maize cobs, strewn on the ground
surface.

Fifty intact or almost intact cobs and six fragments
were collected from the surface scatter. Of the intact
cobs, twenty-nine are red and twenty-one white. Many
are quite tripsacoid in having indurated tissues of the
rachis and glumes. These cobs are quite similar to those
of the lot from site RAnL 290 except that they are
smaller. In shape they are tapered and rounded at the
butts. The shanks are intermediate in thickness; the
lower glumes are stiff, and the rachillae long. Measure-
ments of five red and five white cobs show no significant
differences in the two types except perhaps in kernel row
numbers. The averages for lengths, diameters, and kernel
row numbers are 7.6 cm., 1.7 ¢m., and 18.0 for the red
cobs, and 7.5 cm., 1.6 em., and 15.6 for the white.

RAnL 1

This is a village site located at the present dispersed
settlement of Lasana beside the Loa River, and may
have been occupied from ca. A.D. 800 until Spanish
arrival in the early 16th century. Maize cobs were col-
lected from scattered surface refuse, and comprise eleven
intact or almost intact specimens and six fragments.

Three of the specimens resemble the slender cobs in
the lot from site 837-1, and three resemble the thicker
cobs from that same site. Five cobs are quite tripsacoid,
with their tissues highly indurated either naturally or
hardened through some kind of impregnation. One of
the cobs has several single spikelets, a characteristic of
corn’s relatives teosinte and T7ipsacum; both of the lat-

[ 59 ]

ter differ from corn in having solitary instead of paired
pistillate spikelets. Three of the cobs have stumps of
what may have been staminate spikes,

RAnL 290

This is a small burial site of niche and shaft graves
near Chiu Chiu, dating to the late prehispanic period.
All graves were found looted, but a collection of maize
cobs was made from the surface of the burial chambers.

Thirty-seven cobs were collected, of which eighteen
are red or pink, seven variegated, and twelve white. The
predominating shape is tapering with rounded butts: the
shanks are intermediate in thickness. The lower glumes
are stiff, indurated and glabrous. On several of the cobs
some of the lower glumes have spots of brown pigmenta-
tion characteristic of maize-teosinte or maize- T’ripsacum
hybrids. The rachillae are prominent and in some speci-
mens protrude beyond the lower glumes.

There are no marked differences between the red cobs
and the white. Averages for lengths, diameters, and
kernel row numbers for eight red cobs are 9.0, 2.4, and
18.3; for six white cobs these measurements are 8.9,
2.6, and 18.3. ‘Two of the red cobs from this site are
illustrated in Plate NV, fig. C. Most of the cobs appear
to be related to the modern race Chutucuno Chico, which
still occurs in Antofagasta Province.

RAnL 186

Dating to the late prehispanic period, this site consists
of a single isolated shelter/dwelling situated on barren
terrain 2 km. from the Loa River. A cache of between
400 and 500 maize cobs was found in the earth floor of
the structure.

The cobs are quite variable in shape and other charac-
teristics. Some are tapering at both ends; others are

' 60 |

more slender and more nearly cylindrical. With respect
to diameter/length ratios, the cobs in a selected sample
of twenty specimens vary from 0.13 to 0.86. In kernel-
row numbers these same specimens vary from ten to
twenty-six. Many of the cobs are twisted and contorted,
and some are fused into conglomerate masses. ‘They must
have been somewhat soft and pliable at one time, perhaps
as the result of some chemical action in the soil.

With respect to race, these cobs appear to represent
a kind of ‘‘melting pot™’ in which the principal partici-
pants were the races Capio Chico and Polulo; a minor
component may have been Chutucuno Chico.

RA nN iL 268B

This is another small, looted cemetery of niche graves
near Chiu Chiu, dating to the late prehispanic period.
A cache of fifty-three cobs was found within one of the
burial chambers.

The specimens are similar to the cobs from site 290
with respect to color; twenty-nine are red or pink, six
variegated, and eighteen white, but are brighter due to
better preservation. Another similarity is the stiff lower
glumes, which in some specimens are curved like the
teeth of a wood rasp. Most of the cobs appear to be re-
lated to the modern Chilean race Chutucuno Chico.

RAnL 337-1

This isolated house site is located beside the Loa River
among scrub vegetation, and is at least 12 km. from the
nearest prehispanic settlement. The remains consist of a
surtace scatter of artifacts and a 3.0 by 8.5 meter house
foundation which contained a maximum of 28 cm. of
unstratified refuse. A sample of wood from the excava-
tion yielded a radiocarbon date of A.D. 1250 + 90 years
(1-5899), thus placing the site within the late prehispanic
period.

[ 61 ]

In one respect, the maize from the refuse is the most
interesting of all of the lots recovered because it contains
cobs of a slender type not previously described in the
archaeological record of South America.

The lot comprises 115 cobs, of which fifty are rather
thick, tapering at both ends. The diameter/length ratios
of these vary from 0.25 to 0.82 in five typical cobs that
were measured; the kernel-row number varied from
fourteen to eighteen. These cobs appear to be related to
the modern Chilean race Capio Chico Chileno.

The slender cobs (see Plate XV, fig. B) are thirty-five
in number. In five cobs measured, diameter/length ratios
varied from 0.09 to 0.14, and kernel row numbers from
twelve to sixteen. Both upper and lower glumes are her-
baceous, quite different from the stiff, indurated glumes
of Chutucuno Chico, and they probably represent a form
of pod-corn. As mentioned earlier in this report, they
appear to be related to the Chilean race Polulo, which
in turn may be a descendant of the primitive popcorn
race Confite Morocho.

Four kernels were found in this collection. All are
popcorn: three are yellow in color and one is brown.

Discussion

The cobs of these collections, the earliest as well as the
more recent, appear to be related to three still living
Chilean races: Capio Chico Chileno, Polulo, and Chutu-
cuno Chico. These in turn are related respectively to
three prehistoric Peruvian races: Confite Iqueno, the
popcorn race Confite Morocho, and the popcorn race
Confite Puntiagudo. The last-named race resembles in
a number of characteristics the Mexican popcorn Polo-
mero ‘Toluqueno, but the remaining two races have no
close counterparts in Mexican maize, either prehistoric
or modern. This fact is consistent with the suggestion

[ 62 |

(see Mangelsdort 1974) that there may have been an in-
dependent domestication of corn in South America.

ACKNOWLEDGMENTS

The field research during which the maize specimens
were obtained was supported by a grant from. the
National Science Foundation. The research was part of
a Columbia University archaeological project in conjunc-
tion with the Universidad del Norte in Antofagasta.

| 63 |

LITERATURE CITED

Galinat, W.C. 1969. The evolution under domestication of the maize
ear: string cob maize. Univ. Mass. Expt. Sta. Bull. 577.

——, 1972. Common ancestry of the primitive races of maize indi-
genous to the Ayacucho area in Peru. Maize Gen. Coop. News
letter 46: 108-109.

Grobman, A.; W. Salhuana, and R. Sevilla, in collaboration with
P.C. Mangelsdorf. 1961. Races of maize in Peru. Nat. Acad. Sci.-
Nat. Res. Council, Publ. 915.

Mangelsdorf, P.C. 1974. Corn, its origin, evolution, and improve-
ment. Harvard Univ. Press, Cambridge.

Pollard, G.C. 1971. Cultural change and adaptation in the central
Atacama Desert of northern Chile. Nawpa Pacha 9: 41-64.

Timothy, D.H.: B. Pefa V., and R. Ramirez E., in collaboration
with W.L. Brown and E. Anderson. 1961. Races of maize in Chile.
Nat. Acad. Sci.-Nat. Res. Council, Publ. 847.

[ G4 |

ger 10 19/5

NICAL MUSEUM LEAFLETS
HARVARD UNIVERSITY

CamspripGr, Massacnusetts, JUNE 16, 1975 Vor. 24, No. 4

THE CONTRIBUTION OF LINNAEUS
TO ORCHIDOLOGY'!
BY
Wintuiam T. STearNn?

The presentation to the Botanical Museum, Harvard
University of a collection of works by Carl Linnaeus
(1707-1778), which formed part of the library of Oakes
Ames (1874-1950), provides a fitting occasion on which
to review concisely Linnaeus’s pioneer contribution to
orchidology. Orchids were Ames’s life-study and, while
himself contributing much new knowledge, he was too
wise and well-informed to ignore the historic background
of modern work; the scholarly essays gathered together
in his Orchids in Retrospect (1948) manifest the range of
his orchidological learning and contain many references
to earlier literature.

Orchidology is the study of orchids as a special group
within the Plant Kingdom. Its existence necessarily de-
pends upon recognition of them as a definable group
worthy of particular study, and that implies a high degree
of botanical sophistication. A group embracing plants so

'This paper embodies a lecture given at the Botanical Museum of
Harvard University on September 18, 1969 when the heirs of Mrs.
Blanche Ames presented Professor Oakes Ames’s collection of Lin-
naeana to the Botanical Museum. It includes part of an earlier lec-
ture, “I'wo thousand years of orchidology’, printed in Proc. Third
World Orchid Conf. 1960, 26-42 (1960).

* Department of Botany, British Museum (Natural History), Lon-

don, England.
5

[ 65 ]

ECONOMIC BOTANY LIBRARY
OF OAKES AMES
HARVARD BOTANICAL MUSEUM

distinct in habit as Hria microphylla, with leaves only 1
cm. long, creeping over the bark of tropical trees, and the
tall erect terrestrial savanna herbs of the genus Hudophia,
as well as the saprophytes exemplified by phyllorchis,
Corallorhiza, Didymoplexis, E:pipogium, Gastrodia and
Neottia, which spend most of their lives hidden in sub-
terranean darkness, can be held together in a single bo-
tanical family, the Orchidaceae, only by emphasis on a
common basic floral pattern by no means obvious to the
uninitiated. The existence of such a ground plan of floral
organization tends to be masked by fantastic variations
upon it in size, shape and colour associated with pollina-
tion mechanisms so complex as to be unbelievable were
they not so well attested. Consequently, the concept of
the Orchidaceae as a group of petaloid monocotyledons
technically distinguished by such associated characters
as a perigon of six segments zygomorphically developed,
stamens reduced to one, two or three, ovary inferior and
usually twisted (the whole flower then upside down) a
stigma modified into a peculiar organ, the rostellum,
seeds very numerous, minute and lacking endosperm—
this concept has been long in the making.

The orchids lack a single marked feature comparable
to the umbel of Umbelliferae (Apiaceae), which led to the
‘arly recognition of that aromatic economically impor-
tant family. Nevertheless, they now present themselves
to us as a natural group without conscious reference to
their individual technical features. This recognition is
the result of familiarity. The firm waxy texture of their
flowers, in shape generally unlike those of other plants,
their frequent use in floristry and their former association
with wealth have established orchids as a generally known
and popular group. They continually provide botanists
with new problems and gardeners with new joys un-
paralleled in any other botanical family.

| 66 |

When discussing the evolution of the orchid flower in
Amer. Orchid Soc. Bull. 15: 18-19 (1946), Ames re-
marked, with light-hearted scholarship, that ‘certain
floral structures and the story of their probable evolution
intensify the belief that orchids possess qualities which
set them apart from other plants. Therefore (in the bosom
of the family, with no botanists present) it is delightful to
regard orchids as beings which have shaped their own des-
tiny and to speak of their floral structures, of the rostel-
lum, for example, as organs purposefully developed to
serve special and fantastic functions’. Inthe 18th century,
Linnaeus recognized most of their distinctive features asa
group, although he knew only a few species. He thus pro-
vided a base upon which, in the 19th century, other bota-
nists, John Lindley their chief, built modern orchidology.

The words orchis (6pxes) ‘testicle’ and logos (Aoyos)
‘discourse’ are Greek, but the distinctive technical details
of the family Orchidaceae if made known to a learned and
intelligent Ancient Greek herbalist, even Dioscorides
himself, in the Ist century A.D. would have had little
or no meaning or interest. ‘To people whose interest in
plants was primarily utilitarian, the division of the Plant
Kingdom into major groups defined by inconspicuous and
apparently trivial features such as these floral subtleties
of organization—groups which Magnol called familae
and Linnaeus ordines naturales—would have seemed
pointless. The Ancient Greeks and Romans, like their
successors in Kurope down to the 17th century, knew
nothing about the functions of the different parts of the
flower; they lacked lenses suitable for supplementing and
training the naked eye in observation; consequently,
they gave little attention to the floral structure on which
the concept of the Orchidaceae rests. For them the word
orchiologos would have suggested either a medical treatise
or an obscene comedy.

[ 67 |

To survive the long, dry, fiercely hot Mediterranean
summer, plants must either have features enabling them
to withstand dessication, such as small, hard or woolly
leaves, or else disappear underground, having stored
during the relatively mild moist winter and spring
enough food in bulbs or tubers to give them a quick
start into growth when the rains come. The storage
organs of a number of Mediterranean orchids such as
Orchis and Ophrys are paired rounded tubers which sug-
gested testicles to the far from prudish Greeks, and they
named these plants orchis (dpyes) and cynosorchis
(kuvvocopxts) ‘dog’s testicle’ accordingly, a matter dis-
cussed by Ames in his little essay on the ‘Origin of the
term Orchis’ in Amer. Orchid Soc. Bull. 11: 146-147
(1942). Dioscorides described the root as bulbous, with
paired swellings, one full and solid, the other wrinkled
and soft, and mentioned asa folk-tale the belief that men
should eat the full one to beget sons but women the soft
one to conceive daughters, presumably avoiding syn-
chronisation! The extension of the word orchis and cog-
nate words to tropical plants with long slender dangling
roots would have seemed an absurdity for them, a con-
tradiction in terms. The extension of the concept of
‘orchid’ to cover such plants owes much to Carl Linn-
aeus (after 1762 Carl von Linné) but did not originate
with him. It developed out of the studies of his predeces-
sors who had a feeling for affinities between plants as ex-
pressed in overall resemblance.

Thus Gaspard (or Caspar) Bauhin (1560-1624), who
attempted in his Prnav (1628) to regulate the names of
all the plants then known, devoted section 6 of his book
Il mainly to orchids, with headings now unfamiliar such
as Cynosorchis seu Testiculus canis, Orchis Serapias, ete.,
although somewhat spoiling the picture by associating
Dens canis (ce. Hrythronium) and Orobanche with them.

| 68 |

J

PuatTE XVI

TRIORCHIS SERA Knabentrautmennle das tleines,

PIAS MAS,
Orchis morio L., illustrating the paired tubers which suggested the
generic name. Reproduced from Leonhard Fuchs, De historia stirpium,
page 559, 1542, Basel.

[ 69 ]

Towards the end of the 17th century came the huge
works of John Ray (1628-1705) and Robert Morison
(1620-1683). Ray’s account occupies pages 1212-12338
of volume 2 of his Historia Plantarum (1688), the spe-
cies being listed under the names Orchis, Tragorchis,
Cynosorchis, Chamaeorchis, Monorchis, Triorchis, Palm-
ata, Pseudo-Orchis, Nidus-avis, Pscudo-limodoron, Helle-
borine and Bifolium. He possessed a good knowledge of
these plants derived from his extensive travels in Britain
and on the Continent of Kurope, and his descriptions
give significant details as regards the species but no for-
mal generic characters. His comments on the great difh-
culty of studying orchids on account of the numerous
allied species and the inadequate descriptions by bota-
nists, of which one description can apply to several spe-
cies and several different descriptions to one species, has
unfortunately not ceased to be true: ‘Caeterum cum
Orchidum historia magna confusione et obscuritate lab-
oret, partim ob multitudinem specierum, et nonnullarum
etiam similitudinem, partim ob earum apud_ botanicos
descriptiones, adeo breves interdum et generales ut et
una pluribus, et plures uni alicul speciel accommodari
possint: unde perspicacissimum quemque et in his studiis
versatissimum torqueant et perplexum reddant necesse
est’ (ist. 2: 1212: 1685). This was a state of affairs
which Linnaeus set out to remedy.

In the third volume (1699) of Morison’s Plantarum
FAistoria universalis Oxvoniensis, which was completed by
Jacob Bobart after Morison’s death, 108 kinds of Orchid-
aceae are listed together, though associated with other
plants having minute seeds such as Orobanche and Pyr-
ola. These are put in six genera: Helleborine, Bifolium
sive Ophris, Triorchis et Monorchis, Orchis palmata,
Pseud-orchis.

Joseph Pitton de Tournefort (1656-1708) in his In-

[ 71 ]

stitutiones Rei Herbariae 1: 481-488 (1701) dealt with
them as a special group under the heading ‘De Herbis,
flore polypetalo, anomalo’, this including Orchis, Helle-
borine, Calceolus (i.e. Cypripedium), Ophris (i.e. Listera),
Limodorum and Nidus avis (i.e. Neottia). Unlike his
predecessors, he provided concise generic descriptions.
Within Orehis he included plants with flowers resembling
‘nunc Hominem nudum, modo Papilionem, Fucum,
Columbam, Simiam, Lacertam, Psittacum, Muscam,
caeterave repraesentante’ now put in Orchis, Dactylo-
rhiza, Ophrys, Himantoglossum, Aceras, ete. This ex-
traordinary mimicry of animal forms by the flowers of
orchids had already excited the admiration of Jacob
Breyne (1637-1697) in his Hwvoticarum aliorumque minus
cognitarum Plantarum, 94 (1678) as Ames noted in his
essay, ‘Orchids in retrospect’ (in Amer. Orchid Soc. Bull.
11: 102-106; 1942). The orchids known to these early
authors were north-temperate species. Of the astonish-
ing richness of the tropics in orchids they knew nothing.

In 1708, H.A. van Rheede tot Draakestein (1685
1691) posthumously recorded six orchids from Malabar,
southern India, in his Hortus Indicus Malabaricus, these
being now known as Rhynchostylis retusa, Vanda spath-
ulata, Acampe Wightiana, Sarcanthus peninsularis, Den-
drobium ovatum and Cymbidium alotfolium.

On his visit to Jamaica in 1687 to 1689, Sir Hans
Sloane (1660-1753) found a number of epiphytic orchids.
Those recorded in his Voyage to the Islands Madera xxx
and Jamaica (1707-25) as species of Viscum, Cardamon,
ete. are now known as Brassavola cordata, Oncidium
guttatum, Broughtonia sanguinea, Vanilla claviculata,
Stenorrhynchus speciosus and Hrythrodes plantaginea.

About the same time, another doctor, Engelbert
Kaempfer (1651-1716), was becoming acquainted with
a few of the many orchids of Java and Japan. He illus-

>

[ 72]

trated three of these in his Amoenitatum Exvoticarum
Fasciculi V (1712), the species now known as Dendrobium
moniliforme, Vanilla domestica and Arachnis flos-aeris.
Such was the state of knowledge of orchids at and
after the birth of Carl Linnaeus in 1707. His career has
been sketched in so many publications, notably B.D.
Jackson, Linnaeus (afterwards Carl von Linné), the
Story of his Life (1923), Kk. Hagberg, Carl Linnaeus
(1952), N. Gourlie, The Prince of Botanists (19538), and
W. Blunt and W.'T. Stearn, The Compleat Naturalist
(1971), that a bare outline will suffice here. He was born
in 1707 at Rashult, Skane, southern Sweden, entered
the University of Lund in 1727, changed next year to
the University of Uppsala, made an important journey
to Lapland in 1782, left Sweden for Holland in 1785 and
returned in 1788, having acquired a doctor’s degree at
Harderwijk and published his Systema Naturae (1735),
Fundamenta botanica (1736), Genera Plantarum (1787),
Flora Lapponica (1737), Critica botanica (1737), Hortus
Chffortianus (1788). He was appointed a professor at the
University of Uppsala in 1741 and that year visited the
Baltic islands of Oland and Gotland. His Species Plan-
tarum, the starting point of modern botanical nomen-
clature, appeared in 1758, and his Systema Naturae, 10th
ed., vol. i, the starting point of modern zoological no-
menclature, in 1758. He died at the age of 70 in 1778.
On his journey to Lapland in 17382, Linnaeus became
acquainted with seven species now known as Dactylo-
rhiza maculata, Coeloglossum viride, Goodyera repens,
Corallorhiza trifida, Listera cordata, Herminium mon-
orchis and Cypripedium calceolus, admittedly but not sur-
prisingly few, considering their remoteness from the
tropics, and also atypical of this essentially tropical
family. He knew Calypso bulbosa only from a coloured
drawing made in Lapland by Rudbeck the Younger.

[ 73

These species he enumerated in his Flora Lapponica
(1737).

In 1738, Linnaeus published his Classes Plantarum
which summarizes the earlier classifications of the Vege-
table Kingdom by his predecessors, Cesalpino, Morison,
Ray, Knaut, Hermann, Boerhaave, Rivinus, Rupp, Lud-
wig, Tournefort, Pontedera and Magnol, as well as his
own artificial “sexual system’, wherein the genera now
included in Orchidaceae form his class Gynandria order
Diandria. He appended to this a ‘Fragmenta methodi
naturalis’ which is an attempt at a more natural system
of classification. Here again, as in his ‘systema sexuale’,
the orchid genera, Orehis, Satyrium, Serapias, Hermi-
nium, Neottia, Ophrys, Cypripedium and Epidendrum,
form a group distinct from other plants, his Ordo IV.
Much later, in his Genera Plantarum, 6th ed. (1764),
he published a modified version under the heading
‘Ordines naturales’; this is reprinted, with a discussion
by Stearn, in the Ray Society facsimile of Iinnaeus,
Species Plantarum, vol. 2, Appendix: 98-102 (1959).
Here appears an Ordo VII Orchidaceae comprising the
above genera. After his death, two former students,
P. 1). Giseke and J.C. Fabricius, published a detailed
account of lectures on the natural system delivered by
Linnaeus at Uppsala in 1764 and 1771 entitled Caroli a
Linne Praelectiones in) Ordines naturales Plantarum
(1792). Here again the Orchideae from a distinct “Ordo™
(in modern terminology, family) of which Linnaeus
stated the characters, the ecology and the cultivation,
then so little understood that most tropical orchids died
soon after their introduction. “Omnium difficillime in
hortis plantantur, et introductae vix per aliquot annos
perennant’, he said, ‘I seminibus satis in hortis non pro-
veniunt: certe si possent seri, nullus hortulus amoeni-
oribus floribus superbiret’. Their successful cultivation

[ 74 J

and raising from seed was not achieved until the 19th
century; John Smith (1798-1888) provided in_ his
Records of the Royal Botanic Gardens, Kew, 227-241
(1880) a first-hand account of progress in orchid-growing
during his lifetime.

On his way back to Sweden from Holland in 1738,
Linnaeus visited Paris and was conducted to Fontain-
bleau, where the wild orchids in particular delighted him.
Before taking up his professorship at Uppsala in October
1741, Linnaeus with six young companions made an ex-
pedition to Oland and Gotland at the request of the
Swedish Estates of the Realm. The purpose of this Jour-
ney, which extended from May to August 1741, was
primarily economic, to search for dye-yielding and medi-
cinal plants and to make observations on soils, minerals,
farming, manufacturing methods and general natural
history. These islands, then scientifically almost unex-
plored, differ from mainland Sweden in being of lime-
stone and having a relatively mild climate, so that they
shelter many organisms rare or unknown on the main-
land, among them various orchids. The two islands natu-
rally excited Linnaeus’s eager curiosity, and he described
them in detail in his entertaining account of the journey
entitled Oliindsha och Gothlindska Resa (1745), of which
an annotated English translation by Marie Asberg and
W.'T. Stearn is now available in Biol. J. Linnean Soc.
5: 1-220 (1973). Their orchid flora includes 22 species.
Linnaeus visited the islands when most of the orchids
were in full flower. As soon as he and his companions
touched the shore of Oland on June 12 (June 1, Old
Style, of Linnaeus’s journal) they realized that this was
a land altogether different from the other Swedish prov-
inces and they decided to make meticulous notes about
everything. Their second day provided them with the
joy of finding Ophrys insectifera, Orchis militaris, Orchis

[ 75 |

ustulata, Orchis morio, Gymnadenia conopsea, Dactylo-
rhiza sambucina, Dactylorhiza maculata and Dactylorhiza
incarnata and caused Linnaeus to remark that, such
plants having previously been unrecorded for Sweden,
he had never expected when seeing them at Fontainbleau
that he would ever see them again. These grew between
Forjestaden and Bornholm. Listera cordata was found
on June 15 (June 4, Old Style). Later (June 25, Old
Style), he came across Orehis mascula in Lummeland
parish north of Visby; he also found Herminium mon-
orchis (July 5, Old Style), Mpipactis helleborine (July 25,
Old Style) and Mpipactis atrorubens (July 8, Old Style).

These Baltic orchids inspired Linnaeus to give special
attention to all of the species then known. Many had
been described by earlier authors under a diversity of
names. Linnaeus reduced them to 41 species in a sur-
vey entitled ‘Species orchidum et affinitum plantarum’ in
Acta Soc. Regiae Sei. Upsal. 1744: 1-87 (1744) cited
by Linnaeus as ‘Act. ups. 1740°. For each one he gave
very detailed synonymy and a concise diagnosis. There
are no generic descriptions here, as he had already pro-
vided them in his Genera Plantarum, 2nd edition (1742).
His genera were few: Orchis (including Dactylorhiza),
Satyrium, Serapias, Herminium, Neottia, Ophrys, Cy-
pripedium and I/pidendrum.

In 1745 Linnaeus published the account of his Oland
and Gotland journey, Oltindska och Gothlindshka Resa,
already mentioned. An important feature of this inter-
esting book is its index because it foreshadows Linnaeus’s
later introduction of consistent binomial nomenclature
for all species of plants and animals. In the text he used
diagnostic polynomials such as Orchis bulbis indivisis,
nectaru labio quinquefido punctis scabro, cornu obtuso,
petalis conniventibus and Orehis bulbis palmatis, nectari
cornu setaceo germinibus longiore, labio crenato, names

[ 76 |

which indicate the distinguishing features of the species
but are inconvenient to write and almost impossible to
remember. In the index these are listed as Orchis 725
militaris and Orchis 727 longicalcar, the number referring
to the numbered entry with diagnostic details in Linn-
aeus’s Hlora Suecica (1745). The generic name and the
added single epithet were easy to remember; the num-
ber took the place of the long descriptive phrase by serv-
ing as a concise reference to one in a standard work.
These names thus provide an intermediate stage between
the polynomial nomenclature of Linnaeus’s Flora Sue-
cica (1745) and the binomial nomenclature of his Species
Plantarum (17538), where, for example, the two species
mentioned above are simply designated Orchis militaris
and Orchis conopsea; the last species, however, has long
been removed from Orc/is and is now known as Gym-
nadenia conopsea. Apart, therefore, from its interest as
an 18th-century travel book replete with first hand ob-
servations on local customs and superstitions—plants,
animals, minerals, quarrying, household remedies, ver-
nacular names, runic inscriptions and much else— Linn-
acus’s Oland and Gotland Journey 1741 is an important
work in the history of biology.

Orchids for Linnaeus were only a small part of the
Vegetable Kingdom which he had set himself the daunt-
ing task of classifying and naming. He accomplished this
in his Species Plantarum (1758) and the associated fitth
edition (1754) of his Genera Plantarum. Contrary to
common belief, the introduction of consistent binomial
nomenclature for species, i.e. the designation of each one
by a two-word name, e.g. Orchis ustulata, consisting of
the generic name, e.g. Orchis, followed a single specific
epithet, e.g. wstu/ata, was not the primary intent of the
Species Plantarum, although this nomenclatural innova-
tion has given it outstanding importance. The addition

[77 |

of specific epithets came almost as an after-thought: as
late as 1748, Linnaeus had not added them to his manu-
script draft now in the library of the Linnean Society of
London. The main purpose of the Species Plantarum was
to provide diagnostic phrase-names (polynomials), such as
Cypripedium radicibus fibrosis, foltis ovato-lanceolatis cau-
linis and Cypripedium bulbo subrotundo, folio subrotundo
radical, which, taken together, functioned like the con-
trasts in a modern key and enabled the species to be dis-
tinguished from their congeners, with relevant synonyms
cited as a guide to the existing literature about each.
Hence, the material, either specimens or illustrations, sup-
plying the information epitomised in the phrase-name, is
of first importance for the typification of the associated
specific epithet. Hach phrase-name set out to give as con-
cisely as possible a statement of the main distinguishing
features. Thus, in his synopsis of the genus Orc/is in the
Species Plantarum (1: 989-944), the phrase-names of spe-
cies 1-10 begin with ‘bulbis indivisis’ (‘with undivided
root-tubers’) and so distinguish them from species 11-14
with ‘bulbis palmatis’ (‘with palmately lobed root-tubers’ )
as well as from species 15-17 with ‘bulbis fasciculatis’
(‘with bunched fibrous roots’). A number of species are
further distinguished by the length of the spur in relation
to the ovary, e.g. species 12. (maculata) with ‘nectarii
cornu germinibus breviore’ (‘with spur shorter than the
ovary ) being thereby distinguished from species 18
(conopsea) with ‘nectaril cornu setaceo germinibus lon-
giore’ (‘with spur setaceous longer than the ovary’) and
species 14 (flava) with ‘nectarii cornu filiformi longitu-
dine germinis’ (‘with spur filiform as long as the ovary’).
A name such as that for species 5 (pyramidalis), Orchis
bulbis indivisis, nectaru labio trifido antice bidentato, cornu
longo, petals acuminatis served to distinguish this spe-
cies from all other orchids then known.

[ 78 |

Descriptions of the genera were provided in the fifth
edition of the Genera Plantarum. Here, Linnaeus recog-
nized as distinet Orchis (which included Tournefort’s
Limodorum), Ophrys, Serapias (which included 'Tourne-
fort’s Helleborine), Limodorum, Arethusa, Cypripedium
(which included Tournefort’s Calceolus) and I/pidendrum
(which included Plumier’s V’ani//a). The content of these
genera can be best illustrated by listing the species in-
cluded in each genus by Linnaeus in the Species Plant-
arum and then giving the currently accepted name for
each species and also the name which Oakes Ames
adopted or would probably have adopted when dealing
with the same species. This list, compiled with the help
of Mr. Peter Fk. Hunt, is as follows:

Linnaeus, Species Plantarum 2: 989-954 (1752)

oe WwW bd ee

-p OOOND

tp

12
13

14
15
16
17

Linnaean name

Currently accepted name
(if different from Linnaean name)

Probable name used by Ames
(if different from

currently accepted name)

Orchis susannae Pecteilis susannae (L.) Raf. Habenaria susannae (L.)
R.Br.
O. ciliaris Blephariglottis ciliaris (L.) Rydb. H. ciliaris (L.) R.Br.
O. bifolia Platanthera bifolia (L.) L.C.Rich. H. bifolia (L.) R.Br.
O. cucullata Neottianthe cucullata (L.) Schltr. H. cucullata (L.) Hoefft.
O. pyramidalis Anacamptis pyramidalis (L.)
L.C.Rich. 7
O. coriophora — —
O. cubitalis Peristylus cubitalis (L.) Kraenzlin H. cubitalis (L.) R.Br.
O. morio _ —
O. ustulata _— —
O. militaris — a
O. latifolia Dactylorhiza latifolia (L.) Soé O. latifolia L.
(D. majalis (Rchb.) P.F.Hunt &
Summerh. )
O. maculata D. maculata (L.) So6 O. maculata L.
O. conopsea Gymnadenia conopsea (L.) R.Br. Habenaria conopsea (L.)
Benth. (non Rchb.f.)
O. flava Tulotis flava (L.) Sengh. H. flava (L.) R.Br.
O. fuscescens Tulotis fuscescens (L.) Raf. H. fuscescens (L.) Torr.
O. strateumatica Zeuxine strateumatica (L.) Schltr. —
O. abortiva Limodorum abortivum (L.) Swartz —

[79 ]

Linnaeus, Species Plantarum (cont. )

Currently accepted name
(if different from Linnaean name)

Linnaean name Probable name used by Ames
(if different from
currently accepted name)
Blephariglottis psycodes (L.) H. psycodes (L.) Sprengel

Rydb.

Galearis spectabilis (L.) Raf.

18 O. psycodes

19 O. spectabilis Orchis spectabilis L.

Koch _
2S. viride Coeloglossum viride (L.) Hartman H. viridis (L.) R.Br.
3. S. nigrum Nigritella nigra (L.) Rchb. —
4 §S. albidum Pseudorchis albida (L.) Love & H. albida (L.) R.Br.

Satyrium hircinum

Himantoglossum hircinum (L.)

Love

5 S. epipogium Epipogium aphyllum Swartz —
6 S. repens Goodyera repens (L.) R.Br. =
1 Ophrys nidus-avis Neottia nidus-avis (L.) L.C.Rich. —
2 O. corallorhiza Corallorhiza trifida Chatelain —
3 O.spiralis Spiranthes spiralis (L.) Chevall. —
4 ©. cernua S. cernua (L.) L.C.Rich. —
5 O. ovata Listera ovata (L.) R.Br. —_
6 QO. cordata L. cordata (L.) R.Br. —_—
7 O. lilifolia Liparis lilifolia (L.) L.C.Rich.

8 O. loeselil L. loeselii (L.) L.C.Rich. —
9 QO. paludosa Malaxis paludosa (L.) Swartz —
10 O. monophyllos M. monophyllos (L.) Swartz —
11. O. monorchis Herminium monorchis (L.) R.Br. a
(‘Monochris’ )
12 O. alpina Chamorchis alpina (L.) L.C.Rich. —
13. O. camtschatea Neottia camtschatea (L.) Rchb.f. —
14 O. anthropophora Aceras anthropophorum (L.) R.Br. —
15 O. insectifera — (O. muscifera Hudson) —
1 Serapias hellecorine Epipactis helleborine (L.) Crantz —
latifolia “ (E. latifolia (L.) All. —_
longifolia Cephalanthera longifolia (L.)
Fritsch _
palustris Epipactis palustris (L.) Crantz —
2 S. lingua = —
1 Limodorum tuberosum Bletia tuberosa (L.) Ames —
1 Arethusa bulbosa — —
2 A. ophioglossoides Pogonia ophioglossoides (L.)
Ker-Gawl. —_—
3. A. divaricata Cleistes divaricata (L.) Ames —_
1 Cypripedium calceolus — —
2 C. bulbosum Calypso bulbosa (L.) Oakes —
1 Epidendrum vanilla Vanilla planifolia Andrews Vanilla fragrans Ames
2 E. domesticum Vanilla domestica (L.) Druce —

ex Lindl.

' 80 |

Linnaeus, Species Plantarum (cont. )

Linnaean name Currently accepted name Probable name used by Ames
(if different from Linnaean name) (if different from
currently accepted name)

3. E. flos-aéris Arachnis flos-aéris (L.) Rchb.f. —
4 E. tenuifolium Cleisostoma tenuifolium (L.) Garay —
5 E. spathulatum Vanda spathulata (L.) Sprengel —_
6 E. ovatum Dendrobium ovatum (L.) Kraenzlin —
7 E. nodosum Brassavola nodosa (L.) Lindl. —
8 E. carinatum Dendrobium carinatum (L.) Willd. —
9 E. aloifolium Cymbidium aloifolium (L.) Swartz —
10 ~E. guttatum Oncidium guttatum (L.) Rchb.f.
11 E. retusum Rhynchostylis retusa (L.) Blume —
12. E. amabile Phalaenopsis amabilis (L.) Blume —
13 E. ensifolium Cymbidium ensifolium (L.) Swartz —
14. E. moniliforme Dendrobium moniliforme (L.) Swartz —

From the above concordance, it will be evident that
Linnaeus’s 19 species of Orchis can be referred to 15
genera or at least 5 genera; his 6 species of Satyrium to
6 genera or at least 5 genera; his 15 species of Ophrys
to 10 genera; his two species of Serapias to 8 genera:
his 8 species of Arethusa to 8 genera; his two species of
Cypripedium to 2 genera; his 14 species of H/pidendrum
to 10 genera. Thus, in a group of species where Linn-
aeus saw only 7 genera, modern orchidologists see 50.
This change of viewpoint has come largely from the in-
troduction into gardens of allied species during the 19th
century, a consequence both of Kuropean botanical ex-
ploration outside Europe and of improved methods of
cultivation, the two together providing living specimens
for study and greatly stimulating this study. Of the
species accepted by Linnaeus in 1758, Europe provided
30, eastern North America 11, Indonesia and the Philip-
pines 4, Siberia 4, India 6, tropical America 8, Ceylon
2, China 1; only 4 were previously undescribed species.
He had a first-hand knowledge of many European spe-
cies in a living state, but his acquaintance with tropical
orchids was meagre. He knew his Orchis cubitalis (Peri-

[ 81 ]

stylus cubitalis) and O.stateumatica (Zeuvine stateumatica)
from Ceylonese specimens in Hermann’s herbarium, now
in the British Museum (Natural History), his Z’piden-
drum amabile (Phalaenopsis amabilis) and I. ensifolium
(Cymbidium ensifolium) trom Chinese specimens collected
by Osbeck and his 14. Vanilla (Vanilla planifolia) trom
cultivated material and 24. nodosum (Brassavola nodosa)
from cultivated material and illustrations, but the other
9 he knew only from illustrations published by Rheede
and Kaempfer. Wisely he kept these little-known species
in a single genus, Mpidendrum; he could record their
existence, give them names but add nothing to an under-
standing of them. This had to wait for Olof Swartz
(1760-1858), who studied many species while in the
West Indies, Robert Brown (1778-1858), who studied
many in Australia, Carl Ludwig Blume (1796-1862),
who studied many others in Java, and John Lindley
(1799-1865), who never visited the tropics but studied
their products more conveniently in the glasshouses of
Kngland and published some 120 currently accepted
genera.

Assessment of Linnaeus’s main publications relating
to orchids depends upon one’s viewpoint and this is con-
ditioned by both the knowledge and the scientific fash-
ions of the time. If, forexample, we imagine Carl Linn-
aeus as a Harvard student submitting his account of
Orchidaceae to Professor Ames as a thesis, which ad-
mittedly requires some imagination, we can be sure poor
student Linnaeus would have been taken to task for the
superficiality of his work, for ignoring so many minute
floral details in his definition of genera, for failing to
separate groups which, in the Professor’s eyes, were ob-
viously distinct; Linnaeus would have been told to ex-
amine their floral structure with greater care. On the
other hand, if we imagine Oakes Ames as an Uppsala

[ 82 |

student making his work on orchidaceous genera known
to Professor Linnaeus, we can also imagine the latter
severely questioning the propriety of his taxonomic pro-
cedure. Linnaeus could well have criticized him for
ignoring the overall resemblance of his plants which
justified keeping them together, for giving excessive
importance to insignificant details and for revealing by
his absurd breaking up of natural genera (although Ames
in fact was more conservative in his approach than some
of his contemporaries) a lack of understanding of the
essentials in classification as then understood.

Both professors would have been right within the con-
text of their times. When the known species of a family
are few, it is probably more convenient to keep them
together within a few genera. When the same species
have had added to them a host of others, it may be more
convenient to divide them all among a much greater
number of genera. Provided the facts about them are
understood, it is often a matter of convenience or pre-
vailing fashion or consistency of treatment in comparable
situations, or even plain egotism, as to the taxonomic
rank these groups should be assigned.

The value of Linnaeus’s contributions to orchidology
lies primarily not in his treating orchids as a special group
but, on the contrary, in giving them the same nomen-
clatural treatment as other plants in accordance with his
methodology. The essentials of this were that:

1) the genera should be named euphoniously and clear-
ly, avoiding names which could be confused ;

2) the generashould be defined according to a consistent
formal plan, using a fixed terminology ;

3) the species should be defined by key characters en-
abling them to be distinguished and recognized ;

4) the species should be given convenient binomial
names.

[ 83 |

Such procedure has been fundamental to the develop-
ment and expansion of systematic botany in general and
to orchidology in particular, to the work of Oakes Ames
no less than to his predecessors, from Linnaeus onwards;
but it now seems so commonplace and self-evident that
only by comparison with the work of Linnaeus’s prede-
cessors does its originality become manifest. Linnaeus
distinguished in the Species Plantarum (1758) only 7
genera and 62 species out of the 700 or so genera and
20,000 or so species of Orchidaceae now known and, by
the end of his life, had become acquainted with only 107
or so genera. Nevertheless, his treatment of these few
exemplified the methods which, with due modification,
have made possible the recording of the many. Thus, in
the works of Linnaeus lay the potentialities of the studies
in orchidology so honourably associated with Ames and
the Botanical Museum of Harvard University.

[ 84 ]

ECONOMIC BOTANY LIBRARY
OF OAKES AMES
HARVARD BOTANICAL MUSEUM

BOTANICAL MUSEUM LEAFLETS

HARVARD UNIVERSITY

Campripngr, Massacuuserts, Jung 27, 1975 Voi. 24, No.

K'THNOBOTANY OF THE WESTERN
TARAHUMARA OF CHIHUAHUA,
MEXICO

I. Nores ON THE GENUS Agave
BY
Roserr A. Byrg, Jr.', Don Burcess’,
AND ALBINO Mares Trias’

INTRODUCTION

The genus 4 gave, native to Mexico and southwestern
United States, has played an important role in the ma-
terial culture of the peoples who lived in its range. In
the Greater Southwest, various species played a part in
almost every aspect of aboriginal life (Castetter, Bell and
Grove, 1988; Gentry, 1972: 9-11). Kven in industri-
alized Mexico, the maguey plants are important sources
of commercial fibre (sisal and henequén) and fermented
beverages (pulque).

In southwestern Chihuahua, several species of fave
are still important to the Tarahumara Indians, although
the intensity of some uses has decreased. From the first
European observations to the present, numerous writers

' Botanical Museum of Harvard University.

* Missionary-linguist, Summer Institute of Linguistics, Rocoroibo,
Municipio de Guazapares, Chihuahua.

* A Western Tarahumara from Municipio of Guazapares who is writ-
ing a series of booklets in Tarahumara and Spanish under the direc-

tion of Mr. Burgess (cf. Mares, 1972 and 1974).

(85 |

Or

have commented on the many uses and functions of
maguey or the mescal plant. Ina future paper, Bye will
review these reports with respect to the T'arahumara as
2n whole. We now propose to review the utilization and
significance of the magueys to one group, the Western
Tarahumara‘’. This paper is the first part in a series on
Western Tarahumara ethnobotany in which we plan to
present information on and discussions of plants used
by the natives living on the eastern slopes of the Rio
Chinipas drainage area in the Municipios of Guazapares,
Chinipas and Uruachi in southwestern Chihuahua. The
Rio Chinipas, a tributary of the Rio Fuerte, originates
west of the Continental Divide near Creel.

Kight species of leave which occur in western ‘Tara-
humara region will be discussed’. Most of the species are
restricted to the northern Sierra Madre Occidental and
range from the arid sub-tropical vegetation of the barran-
cas to the cool pine forests of the sierras. Table I sum-
marizes the range, elevation and ecological association
of the Western Tarahumara agaves.

UTILIZATION

Agave is first and foremost a source of food. The com-
mon method in preparing the heart (.e., the secondary
thickened stem with compacted nodes) and leaf bases
consists of pit-baking, usually by the men. Plants are
selected by size and degree of development of the inflo-
rescence bud, The leaves near the centre around the
swollen flower stalk bud are usually disposed in a series
of smaller leaves contrasting with the previous larger
ones. Harvesting the plant Just prior to the elongation of
the large inflorescence when the concentration of energy-

'Tarahumara Baja, according to Mexico’s Instituto Nacional Indi-
genista.

” Consideration of other species found to the east and not recognized
by the Western Tarahumara will be deferred to a later paper.

if a]
86 |

rich sugars is highest is done by all people who eat ma-
guey hearts. Pit-baked magueys are an important food
source at the time of year when food stores of the pre-
vious growing season are dwindling and before the arrival
of summer rains which are needed for the growth of wild
greens and cultivated plants. Occasionally, magueys are
pit-baked at other times of the year.

The plants are dug up with a pointed stick, an iron
bar or an old ax. ‘Then the outer leaves are trimmed
with a machete, leaving the hearts and leaf-bases. A
wooden hook is inserted into the bottom of each heart,
two of which are tied together and placed over a burro
to carry to the pit. If people are carrying the hearts and
the distance to the pit is great, the leaf-bases are cut off
and carried separately in a basket.

A fire, usually of green oak wood (Quercus spp.) is
built in the pit with rocks placed on top of the wood.
Hard, heavy rocks are chosen, since they hold heat longer.
The round pit is usually about three feet deep and five
feet wide. The same pit is used year after year. About
noon, when the fire has burned down to coals and the
smoke, which would give the hearts a bad taste, no longer
comes out, the pit is ready for use. If the coals are too
hot, a layer of earth is placed over them. Then the mes-
‘al hearts with the leaf-bases still intact are placed in the
pit. Sometimes a stick of ‘‘rolobusi’’® (Bouvardia sp.)
is inserted into each heart to assure sweetness in cooking.
If ‘‘rolobusi”* is not available, leaves of ‘‘usabi’’ (Prunus
Gentry) are placed on the hot rocks at the bottom for
the same purpose. After the hearts are in the pit, the
pit is covered with leaves of a palm (Sabal wresana).
Then a layer of soil is placed over the top and left for

° Western Tarahumara belongs to the Uto-Aztecan family of lan-
guages. It is written here using the Spanish orthography with the

addition of the glottal (*). The r’s and 1’s differ from Spanish in that
they are retroflexed.

=

[ 87 |

two nights. On the morning after the second night, if
the tops of the mescals are not done, the hearts are re-
moved and the fire rebuilt under the rocks. Then the
hearts are turned over, replaced in the pit, covered again
and left for another day. If still not sufficiently cooked,
they may remain in the pit-oven an extra day.

Upon removal, the hearts and leat-bases are loaded
into gunny sacks and baskets and taken to the house.
They are either eaten by themselves or oftentimes with
acorn drink, *‘gue huali’’ (Spanish: esquiate), or a corn
gruel, ‘‘atonali’’ (Spanish: atole). The sweet flesh of the
heart is eaten and the fibres spit out. The heart 1s some-
times ground and mixed with tortilla dough. The prepa-
ration and consumption of agaves have been described
for other areas of the Tarahumara region (Pennington,
1963: 129-131; Bennett and Zingg, 1985: 148-149).
The hearts of Agave pacifica, A. Shreve, A. multifilifera
and f. polianthiflora and the leaf-bases of 4. americana
var. evpansa, A. multifilifera and A. polianthiflora are all
pit-baked and eaten.

‘Sugui’’ (Spanish: tesgiiino), a fermented beverage,
‘an also be made from the well cooked mescal. In areas
where agaves are plentiful, only the leaf-bases are used.
In other areas, the hearts are also used. ‘The mescal is
thoroughly mashed in a trough or hollow rock with a
large mallet cut from the limb of an oak (Quercus spp. ).
The juice is collected and strained, first through a coarse
basket, then a cloth. After straining, the sweet Juice is
cooked ina large pot by women, until all of the impuri-
ties have come to the top and are skimmed off. Follow-
ing the cooking, the juice is placed in a large-mouthed
pot to cool and then put into a fermenting pot, ‘‘sicoli
ronela’’ (Spanish: botija), where it is left for two or three
days. The process can be speeded up by adding the juice
of sprouted corn (Zea Mays) that has been ground, cooked

.

| 88 |

and strained. Water is added if the ‘“‘imé”’ juice is too
strong. When the bubbling of the fermenting process
begins to slow down, a handful of ground wheat (7'riti-
cum aestivum) is added. Then, within five to six hours,
the ‘‘sugui™’ is ready to drink. If the drink is not to be
used for several days, the pot is sealed with a board and
mud. This will keep the drink from going bad for six
to eight days. Agave bovicornuta is said to make a
stronger ‘‘sugui’” than the other species. 4. americana
and 4. vilmoriana are the only species not used for
making the fermented drink.

Tarahumaras also dry the roasted mescal so that it will
last five to six months (during the dry season). The outer
fibres are removed from the cooked leaf-bases and the
meat is mashed, as when making “‘sugui’’, and then
ground on the ‘‘majta’’ (Spanish: metate). It is then
shaped into a cylindrical cheese form and placed in the sun
to dry for a week. This *‘mesagoli”” or “‘imé cha’ poli’,
mescal bread, is especially good food for long trips. It is
‘carried in a bag of ‘‘cobisi’’ (Spanish: pinole), toasted
corn grain that is ground dry into a powder and later
mixed with water to drink, and used as a trail food.

.

al. pacifica is not used to make ‘‘mesagoli’’, since it is not
so sweet, has less meat and the fibres are harder to chew.

The emergent flower stalks, ‘‘balila’* (Spanish : quiote),
are eaten as well. A young, tender stalk can be cut from
the plant and roasted on coals. Then it is peeled and
eaten. The taste is similar to that of the squash. It can
also be chopped up, boiled, and mixed with the whey
left over from making ranch-style cheese. When a stalk
has elongated about a yard, the top foot is cut off and
the remainder is peeled and roasted in the pit-oven along
with mescal hearts and leaf-bases. The decapitated plant
is allowed to grow and is cut the following year for pit-
baking when it is sweeter. In other parts of the Tara-

"89 |

humara region, the prepared stalks are eaten with ‘‘gue’-
huali’’ (Spanish: esquiate), toasted corn ground with
water, and ‘‘cobisi’’, but this practice is not common in
the western region. ‘The stalks of species of the subgenus
Agave are eaten, while those of the members of sub-
genus Littaeca are not because of their bitterness or
small size.

If the stalks be allowed to develop, the unopened flow-
ers with the inferior ovaries removed are cooked and
washed to remove the bitter constituent. The cooked
flowers can be eaten boiled, fried or made into tortillas.
The sweet nectar is likewise gathered with a small tube
for consumption. The flowers of sfeave bovicornuta, A.
pacifica, A. Shrevet and A. Wocomahi are eaten. The
flowers of 4. bovicornuta are preferred as food because
of their better taste and as they are available when there
is little other food. On the other hand, flowers of 4.
Wocomah, which appear during the rainy season, are not
often collected, since there are many other preferred
greens (Spanish: quelites) available at that time.

Tortillas made from the perianth of the maguey flow-
ers are especially valued by Western Tarahumara. The
unopened flowers are cut from the large inflorescence.
Usually a hooked pole is used to gather the flowers, so
that the stalk is not broken, and the underdeveloped
flower buds are allowed to develop tor future gathering.
The flowers are put on the house roof and collected early
the next morning, when they are much sweeter. They
are cooked well and washed overnight in the arroyo to
remove the bitterness. The next day they are ground
with lime-cooked maize, or “*najpili” (Spanish: nixta-
mal), and made into tortillas. Some cooked flowers are
retained for later frying to be eaten with tortillas.

Fibre is extracted from the leaves of Aeave multifih-
fera and A. pacifica. The larger leaves are cut, beaten

— 90

and placed in water. The tissue surrounding the fibre
bundles rots away. The resulting fibres are washed in
the arroyo and commonly used in making lassos and
domestic rope.

Stupefying fish with agaves is well known among the
Tarahumara (cf. Gentry, 1972: 11, 89). The Western
Tarahumara use Agave Vilmoriniana, which is rich in
sapogenins, and 4. bovicornuta and A. Wocomahi, the
toxic constituents of which are unknown, to stupefy fish
trapped in rock enclosures in the rivers. The plants are
mashed on a rock, the juices are allowed to drain into
the river and, finally, the mashed plants are thrown into
the water. Several plants are needed rapidly to stupefy
the trapped fish which, upon reaching the surface, are
collected and prepared.

The crushed leaves of Agave Vilmoriniana, which is
rich in the sapogenin smilagenin, are prized as soap for
bathing and washing clothes, wool, blankets and dishes.
Not all ‘‘ahué”” leaves are good for washing. The whiter
leaves are chosen over the greener ones which produce
too strong a soap, especially for washing hair. Another
method of selection is to twist the end of the leaf. If it
breaks, that leaf is not used.

Ritual curing with agaves is important to the Tara-
humara ceremonies (Lumholtz, 1902: 821-822, 368).
Small plants of fave Shreve: are placed in pots of
water in front of the cross during fertility rites for sheep
and goats and in ceremonies for the dead.

Discussion

About 300 years ago, the Tarahumara retreated west-
ward and assimilated various tribes of close cultural and
linguistic affiliations (Pennington, 1986: 8, 9, 229, 2380,
Map I). Today they number about 50,000 (Pennington,
1963: preface). The Western Tarahumara are geographi-

91)

‘ally isolated from the mainstream 'Tarahumara and live
east of the Guarijio (Warihio). They number tewer than
10,000 (Burgess, 1970: 64).

The Western Tarahumara recognize two groups of
Agave’. They recognize the group ‘‘imé™” (subgenus
Agave) through its leaf with terminal and marginal
spines and through the paniculate inflorescence. The
‘imé”” group includes: *‘galime’’ (fl. americana var. ev-
pansa), “sa pult”” (A. bovicornuta), ““gusime™” (4. paci-
fica), “‘o'tosi™” (CL. Shrever), and *‘ojcome™” (P. Woco-
mahi). The introduced succulent, -floe vera (Liliaceae),

superficially resembling the maguey plant, is not widely
considered an *‘imé™ although it is called *‘imé me’ ta-
gochi’. The second group (subgenus Littaea) is recog-
nized through its leaf with a terminal spine and smooth
margin and its racemose inflorescence. This group in-
cludes ‘‘chahut’’ (Ll. multifilifera), rv yéchili” (A. polian-
thiflora) and “tahué™” (A. vidmoriana). The group carries
no generic term. Some Western ‘Tarahumara include
“chahul” and ‘ra’yéchili”” in the ‘‘imé*” group because
they are pit-baked and eaten.

There isa striking similarity and contrast in the recog-
nition of the species of .feave among the Western ‘Tara-
humara, the Guariyio to the west and the Tarahumara
to the east. It may be explained by the geographical
distribution of the magueys and, in part, by the histori-
cal past. Table Ll summarizes the names of the eight
species of -fgave occurring in the Western ‘Tarahumara
region.

In subgenus Littaca, Agave multifilifera and A. Vil-
morimana are similarly recognized and domestically used
by all three groups. One possible explanation is that the

“The Spanish word, ‘“mescal’’, is locally used for the plants of

Agave, but in other parts of Mexico this term is applied to a distilled
beverage made from the Agave plant.

leaves and fibres of ‘‘chahui’’ and the sapogenin-rich
leaves of ‘‘ahué’” can be easily transported and traded
throughout the area. On the other hand, food and fer-
mented beverage made from the members of subgenus
A gave are prepared and consumed near dwellings and
‘an not easily be transported (except mescal bread). A.
polianthiflora is presently known only from Western
Tarahumara region and adjacent Sonora. fl. americana
var. evpansa 1s not reported from the Guarijio or Tara-
humara regions.

Another significant point is that the species of eave
here discussed are commoner on the western slopes and
barrancas of the Sierra Madre Occidental. The diversity
of species of subgenus f.2ave is greater toward the west
(cf. Gentry, 1972). The increased diversity of ‘‘imé”’
encountered by the westward moving ancestors of West-
ern Tarahumara could be handled by an expanded no-
menclature borrowed from an assimilated tribe, such as
the Guazapar. The language of the Guazapar was said
to be the same as Guarijio, although it resembled Tara-
humara (Pennington, 1968: 9).

Boranical List or A gave
Subgenus Agave

Agave americana LL. var. expansa (Jacobi) Gentry,
U.S.D.A., Agric. Handbook no. 899, p. 80. 1972.
Western Tarahumara: GALIME (house maguey)
Mexican: mescal maguey
This large maguey, cultivated near houses, produces
easily transplanted sucker shoots or ‘‘ranala’’ (Spanish:
hijos). Its antiquity among the Western T'arahumara is
unknown, and it is not reported from any other nearby
regions of southwestern Chihuahua and adjacent Sonora.
Gentry (1972: 84) suggests that it was introduced north-

| [ 93 ]

ward into Sonora and southwestern United States, proba-
bly as an ornamental, after 1850.

The bases of large leaves, pit-baked and eaten or pre-
pared as mescal bread, are sweet, bland and lack fibres.
One large plant usually provides sufficient material to
fill a large baking-pit. Since other parts of the plant—
the stalk, flowers and heart—are bitter, they are not
consumed.

Specimen: Curmmuanua: Municipio de Guazapares, Rocoroibo, ca.
6000 ft. Cultivated plants about 6-8 feet high near governor’s house,
associated with spineless Opuntia, apple and peach trees. Nov. 10,
1973. Bye and Burgess 5827.

Onservation (by R. Bye):

Cuimuanvua: Municipios de Guazapares y Chinipas, Wasachi, ca.
6000 ft. A recently transplanted linear plantation near walls of a
ranchito in the open flat above arroyo, July 23, 1974.

Agave bovicornuta Gentry, Carnegie Inst. Wash.
Pub. 527, p. 92. 1942.

Western Tarahumara: sA°’puLi (meat barranca: mean-

Ing uncertain)

Mexican: mescal lechuguilla

This attractive maguey is found in the middle zone of
the barrancas associated with oaks (Quercus spp.). The
flowers, well washed to remove the bitterness, are eaten
cooked or in the form of tortillas and are preferred to
those of other species in making tortillas. The hearts are

.

baked and used in preparing a strong “‘sugui’’ which is
said to be similar to ‘*pisto’’, the local Mexican distilled
alcohol from cfeave and Yucca. The caustic juice of the
leaves causes a burning irritation, followed by white blis-
ters, on sensitive skin. The plant is employed to stupefy
fish.

Specimen: Cuiuuanua: Municipios de Guazapares y Chinipas, be-
tween Cusdrare and Nopalero, ca. 3500 ft. Seattered plants on rocky
slopes of the barranca associated with oaks (Quercus spp.) and below
the mixed oak and pine forests. Nov. 8, 1973. Bye, Burgess and
Mares 5806.

94 |

Opservation (by R. Bye):

Curnuanua: Municipios de Guazapares y Chinipas, between Wasa-
chi and La Paz, ca. 5000. Scattered plants in the lower section of
the mixed oak and pine forest. The leaves are a light yellow-green.
July 23, 1974.

A gave pacifica 'Trel., U.S. Natl. Herb. Contrib. 28:
118. 1920.
Western Tarahumara: GUSIME (maguey of the woods
or stick maguey)
Mexican: mescal del monte (wild plant)
mescal casero (cultivated plant)

Found in the lower parts of the arid sub-tropical bar-
rancas in short thorn forest, fave pacifica produces
sucker shoots which are transplanted near dwellings in
the lower oak (Quercus albocincta) zone of the middle
barranca zone. Considered the most delicious and diffi-
cult to collect in the field, this species is often grown
near the houses. It also became popular with the
Spaniards and Mexicans who settled in the deep can-
yons and later transplanted and cultivated it.

The hearts of the cultivated plants are larger than
those of the wild plants. The baked hearts are sweet and
consumed baked or in the form of fermented drink. The
flowers are eaten cooked or in the form of tortillas. The
fibres from the leaves provide cordage and thread.

Specimen: Curmuanvua: Municipios de Guazapares y Chinipas, Cu-
sirare, ca. 3000 ft. Cultivated plants, whose suckers originated from
the wild in the lower parts of the barranca, in a hollow at the base
of a field ‘‘trinchera’’ or stone retaining wall near a ranchito, Plants
about 5 feet high with the ‘‘quiotes’’ cut off. Associated with such
woody plants as /pomoea arborescens and Randia sp. Nov. 8, 19738.
Bye, Burgess and Mares 5796.

Opservation (by R. Bye):

Curmuanua: Municipios de Guazapares y Chinipas, below La Paz,
ea, 3500 ft. Cultivated plants along astone fence and margin of field

near house surrounded by a peach orchard. In the lower oak (Quer-
cus albocineta) zone. July 25, 1974.

Agave Shreve: Gentry, Carnegie Inst. Wash. Pub.
527, p. 95. 1942.

Western Tarahumara: o'ros. (white, plural form)

Mexican: mescal blanco

This whitish appearing plant is scattered in open areas
in the middle and upper zones of the barrancas in the
oak and mixed oak and pine vegetation. The hearts are
pit-baked for eating or are employed in making mescal
bread or a fermented drink. The stalks are also cooked
for food. Small plants are used in curing and in death
ceremonies.

Specimen: Curavuanvua: Municipios de Guazapares y Chinipas, No-
palero, ca. 4000 ft. Small scattered plants on open, rocky slope with
mixture of oaks (Quercus spp.), pines (Pinus spp.) and Nolina mata-
pensis. Some plants with “‘quiotes’’ removed. Nov. 9, 19873. Bye

and Burgess 5813.

Agave Wocomaht Gentry, Carnegie Inst. Wash. Pub.
527, p. 96. 1942.

Western ‘Tarahumara: OJCOME (pine maguey)

Mexican: mescal verde

A small maguey, -feave Wocomahi is found scattered
in rock crevices in the sierras among pines. Baked hearts

‘an be used to make *‘sugui’’ and “‘mesagoli’’, although

some hearts are said to have a bad taste and are not in-
gested. The flowers are an acceptable food but are not
eaten, as they appear at a time when there is an abun-
dance of other greens.

Specimens: Cuimuanua: Municipios de Guazapares y Chinipas, No-
palero, ca. 4000 ft. Few scattered plants on open, rocky slope with
a mixture of oaks (Quercus spp.) and pines (Pinus spp.). The plants
are commoner above. Nov. 9, 1973. Bye and Burgess 5812.— Muni-
cipio de Guazapares, Rocoroibo, ca. 6000 ft. On rocky, open slope
above arroyo with mixed oaks (Quercus spp.), pines (Pinus Engelmannit
and P. ponderosa) and madronos (Arbutus spp.). This plant was more
robust than usual because of the moist site. Nov. 10, 1973. Bye and
Burgess 58385.

96 |

Subgenus Littaea

A gave multifilifera Gentry, U.S.D.A., Agric. Hand-
book no. 899, p. 46. 1972.
Western Tarahumara: CHAHUI

A medium sized plant, ‘‘chahui’’ grows in rock out-
crops in the upper zone of the barrancas in the mixed
oak and pine vegetation. The hearts and leaf bases are
pit-baked, eaten or made into ‘‘mesagoli’” and ‘‘sugui’’.
The leaves are an important source of fibre.

Specimens: Curnuanua: Municipios de Guazapares y Chinipas, near
Nopalero, ca. 4000 ft. In crevices of rocks in the mixed oaks (Quer-
cus spp.) and pines (Pinus spp.). Nov. 8, 1973. Bye, Burgess and
Mares 5783. —Municipio de Guazapares, along ridge above Nopalero,
cea, 5000 ft. Few scattered plants in rock crevices on outcrops in
mixed oaks (Quercus spp.), pines (Pinus spp.) and Arctostaphylos pun-
gens. Nov. 9, 1973. Bye and Burgess 5819.

Agave polanthifiora Gentry, U.S.D.A., Agric.

§ i
Handbook no. 399, p. 51. 1972.
Western ‘Tarahumara: RI YECHILI

This small, attractive and infrequent maguey grows
on partially open rocky slopes and outcrops in the mixed
oak and pine forest in the upper zone of the barrancas.
The hearts and leaf bases were formerly eaten as food
and employed to make a fermented drink.

Specimens: Cutuuanua: Municipios de Guazapares y Chinipas, above
Nopalero, ca. 4500 ft. Few scattered plants in whiterock crevice in
open spots with oaks (Quercens spp.), pines (Pinus spp.) and Arcto-
staphylos pungens. *“‘Quiotes’’? about one foot high. Nov. 7, 1973.
Bye, Burgess and Mares 5779.— Below Wasachi, ca. 5500 ft. Few
scattered plants on open, moist rocky surface with mixed oaks (Quer-
cus spp.) and pines (Pinus spp.). “‘Quiote’’ about one foot high with
flowers rose. July 23, 1974. Bye, Burgess and Mundy 6414.

Agave Vilmoriniana Berger, in Fedde Repert. Spec.

Nov. Reg. Veg. 12: 508. 19138.
Western Tarahumara: AHUE
Mexican: amole

[ 97 |

These spider-like plants cling to the vertical rock walls
of the barrancas from the lower arid sub-tropical region
to the higher cooler region. The plant is not eaten. The
leaves provide a soap for bathing and washing clothes,
wool blankets and utensils, while the macerated plants
are used as a fish poison.

Specimen: Cuimuanua: Municipios de Guazapares y Chinipas, be-
tween Nopalero and Cusidrare, ca. 3500 ft. Scattered plants found on
protected vertical walls. Nov. 8, 1973. Bye, Burgess and Mares 5805.

CONCLUSIONS

Kight species of fave are recognized and used by the
Western Tarahumara. ‘The members of subgenus -f cave,
generically called “‘imé’’, are distinguished from the
members of subgenus Littaca. The two groups are differ-
entiated on gross morphology of leaves and inflores-
cences. Some Western ‘Tarahumara include all of the
magueys that are pit-baked under “‘imé”’.

ach species of “‘imé™” has a name which reflects the
habitat preference or a characteristic of the species. Al-
though the Tarahumara to the east also recognize sub-
genus -feave, ““mé’, as different from that of subgenus
Littaea, the Western ‘Tarahumara recognize a greater
diversity of species, and their names are similar to
Guarijio names. In subgenus Littaca, two of the three
species on the eastern slopes of the Rio Chinipas area
have names in common with Guarijio names to the west
and ‘Tarahumara names to the east. The third species,
A. polianthiflora, has not been reported east of the
Western Tarahumara region.

There is a greater specific diversity of .feave on the
western slopes of the Sierra Madre Occidental than on
the eastern slopes. The similarity of the Western ‘Tara-
humara names to those of the Guarijio for the members
of subgenus .f@ave may be explained, in part, by the
western movement of ancestors of the Tarahumara from

[ 98 ]

the eastern foothills and uplands to the western uplands
and barrancas. Subsequent assimilation of former tribes
of that region may have included additions to Tarahu-
mara 4 gave nomenclature. The similar names for Af.
multifilifera and A. Vilmoriniana may reflect the wide
familiarity of the leaf, fibre and soap products of the
plants. These plants and products would probably be
encountered in travel, transport and trading. The other
species provide food and beverage prepared and gener-
ally eaten near dwellings and usually not traded or
offered to strangers.

‘Two species of f4gave produce sucker shoots, and the
Western Tarahumara have used this characteristic to ad-
vantage in their propagation. The antiquity of propaga-
tion of 4. americana var. expansa is unknown, but its
presence in northwestern Mexico is thought to be only
a century old. ‘The much valued “f. pacifica has been
propagated closer to the locality of its preparation and
consumption and at higher elevations than its normal
range.

SPECIMENS AND PHOTOGRAPHS

Voucher specimens were collected, prepared and iden-
tified by R. Bye. Dr. H. S. Gentry was very helptul
in instructing him in the preparation and identification
of sterile and fertile plants and specimens. ‘The speci-
mens are deposited in the Economic Botany Herbarium
of Oakes Ames in the Botanical Museum, Harvard Uni-
versity (KCON). Duplicates will be distributed to the
Jniversidad Nacional Aut6noma de México (MENU),
Gray Herbarium of Harvard University (GH), and the
Herbarium of H. S. Gentry in Phoenix, Arizona.

All photographs were taken by Don Burgess.

TABLE I

Western Tarahumara Agave

Ecological
Range Elevation Association}!
subgenus Agave
americana w Mexico to sw U.S. (100)-6000 ft. cult. (PO)
var. expansa Jalisco to Ariz. & Calif.
bovicornuta nw Sierra Madre Occidental 3000-6000 ft. Q-PO
Son., Chih. & Sin.
pacifica nw Mexico (0)-3500 ft. sT
Son., Chih. & Sin. (cult.sT-O)
Shrevet nw Mexico 3000-6000 ft. O-PO
Son. & Chih.
Wocomaht nw Sierra Madre Occidental 5000-7500 ft. © PO-P
Son., Chih., Sin. & Dgo.
subgenus Litlaea
multifilifera nw Sierra Madre Occidental 5000-6000 ft. O-PO
Chih.
polianthiflora nw Sierra Madre Occidental 4000-6000 ft. ©Q-PO
Son. & Chih.
Vilmoriniana w Mexico 2000-4000 ft. sT

Son. & Chih. to Jalisco &
Aguascalientes

1—_cult.=cultivated; O=oak forest (Quercus spp., e.g. Y. albocincta); P=pine
forest (Pinus spp., e.g. P. Engelmannii, P. arizonica, P. ponderosa); PO=
pine-oak forest; sT=arid sub-tropical forest with Ipomoea arborescens, Bur-

sera spp. and various woody Leguminosae.

' 100 |

TABLE II
Nomenclature of Western Tarahumara Agave

Species Guarijio! W. Tar. Tarahumara?
subgenus Agave _ imé mé
americana 0 galime 0°

var. expansa

borvicornuta —sapuri sa’puli _
pacifica — gusime ku’uri
Shrevei totosa o’tosa mé*
Wocomaht wocomahi ojcome mésagori
mé
subgenus Littaea
multifilifera chahui chahui chawi (or chahui)

polianthiflora taiehcholi ri’yéchili 0°

Vilmoriniana hauwé ahueé awé (or ahué)

1—Guarijio (or Warihio) names, from Gentry (1942 and 1963).

2—Tarahumara names collected by R. Bye in the region of Creel-Barranca
del Cobre-Barranca de Batopilas.

3—Western Tarahumara names, plant photographs and specimens not recog-
nized by native informants.

4—One informant, who grew up west of Cerocahui near Western Tarahumara
Zo

region but now lives in Barranca de Batopilas, recalled “retosa” as the
name used for this maguey in his youth.
“0” =species not recorded from the area.

“__” — data uncertain.

( 101 |

ACKNOWLEDGMENTS

We would like gratefully to acknowledge the help and
encouragement of our families and friends in the United
States and Mexico who generously supported our studies.
A special thanks is extended to those Tarahumara who
through friendship shared their knowledge. Without the
continual encouragement and critical evaluation of Pro-
fessor C. W. Pennington, Department of Geography,
Texas A & M University, the research upon which these
notes are based might have faltered. Professor Richard
Evans Schultes, Botanical Museum, and Dr. Bernice
Schubert, Arnold Arboretum, kindly read the manu-
script and offered suggestions. Robert Bye would like
personally to thank Dr. Margaret A. Towle, Curator of
the Ethnobotanical Laboratory of the Botanical Mu-
seum, for her support and valuable suggestions. Financial
support between 1971 and 1974 was generously extended
to Robert Bye by the Botanical Museum and the De-
partment of Biology, Harvard University, the National
Geographic Society and the National Science Founda-
tion Grant GB-35047 for Improvement of Doctoral
Dissertation Research.

[ 102 ]

BIBLIOGRAPHY

Bennett, W.C., and R.M. Zingg. 1935. The Tarahumara. Univ. of
Chicago Press, Chicago.

Burgess, D.H. 1970. ‘‘Tarahumara Phonology (Rocoroibo Dialect). ”’
In R.W. Ewton, Jr., and J. Ornstein (eds.). 1970. Studies in
Language and Linguistics (1969-1970), pp. 45-66, Texas Western
Press, El Paso.

Castetter, E.F., W.H. Bell and A.R. Grove. 1938. ‘‘The early uti-
lization and the distribution of Agave in the American Southwest. ”’
Univ. New Mex. Bull., Bio. Ser. 5(4).

Gentry, H.S. 1942. Rio Mayo Plants. Carnegie Institution of Wash-
ington, Publ. no. 527.

——, 1963. ‘‘The Warihio Indians of Sonora-Chihuahua: an ethno-
graphic survey.”’’ Bur. Amer. Ethnol., Bull. 186: 61-144, pl. 28-38,

——, 1972. The Agave Family in Sonora. U.S.D.A., Agric. Hand-
book, no. 399.

Lumholtz, Carl. 1902. Unknown Mexico. Charles Scribner’s Sons,
New York. Volume I.

Mares Trias, A. 1972. Hue’cd E’cartigame Nehualiame Ju Imé. (Haec-
mos muchas cosas con el mezcal). Instituto Lingiiistico de Verano
en coordinacion con la Secretaria de Educacién Ptiblica de México,
México, D.F.

——, 1974, Rega Me’liame Ju Ye Ajagame. (Como cazamos y pes-
camos). Instituto Lingiistico de Verano en coordinacioén con la
Secretaraia de Educaci6n Publica de México, México, D.F.

Pennington, C.W. 1963. The Tarahumar of Mexico. Univ. of Utah
Press, Salt Lake City.

[ 103 ]

[ILLUSTRATIONS

{ 105 |

PLATE XVII

7 =
=

PO, ON a:
fae. cata

Agave americana var, expansa (— galime’”)

Mature leaves about 4) feet long.

| 106 |

PLaTE XVIII

; 66 739
Agave bovicornuta ( ‘sa’ puli’’)

Mature leaves about 2 feet long.

PLATE XIX

ee 29)

Agave pacifica (~ gusime

Mature leaves about 3 feet long.

[ 108 ]

PLaTrE XX

Agave Shrevei (“‘o Ai

Mature leaves about 13 . feet long.

| 109 |

PLATE XX]

Agave Wocomahi (‘ojcome’’)

Mature leaves about | foot long.

[ 110 |

PLATE XXII

hili’’)

éc
aves about 4 inches long.

ri’yé

e polianthiflora (**

)

4 gar

N

Agave multifilifera (“‘chahui’’)

{ature le

+ feet long.

2

Mature leaves about

PLATE XNIII

Agave Vilmoriniana (- ahué’’)

Mature leaves about 5 feet long.

[112 ]

HARVARD UNIVERSITY

CampripGre, Massacnuserts, Ocroser 31, 1975 VoL. 24, No. 6

NUTRITIONAL VALUE OF COCA
BY
JAMES A. Duke’, Davip AULIK* AND
Timoruy PLlowMan®

Leaves of wond’rous nourishment

Whose Juice Suce’d in, and to the Stomach tak’n

Long Hunger and long Labour can sustain;

From which our faint and weary Bodies find

More Succor, more they cheer the drooping Mind,

Than can your Bacchus and your Ceres join’d.
—AsBrAHAM Cow ry in Mortimer’s

History of Coca

Abstract. Coca leaves (Erythroxvylum Coca am.)
from Chapare, Bolivia, compared to an average of 50
other Latin American vegetable products, are higher in
calories (805 per 100 g compared to 279), protein (18.9
g: 11.4 g¢), carbohydrate (46.2 g: 87.1 g), fiber (14.4 g:
3.2 g), ash (9.0 g: 2.0 g), calcium (1540 mg: 99 mg),
phosphorus (911 mg: 270 mg), iron (45.8 mg: 3.6 mg),
vitamin A (11,000 IU: 185 LU), and riboflavin (1.91
mg: 0.18 mg). Coca was lower than the average for the
50 plant foods in oil content (5.0 g per 100 g compared
to 9.9 g), moisture (6.5 g: 40.0 g), thiamin (0.35 mg:
0.388 mg), niacin (1.8 mg: 2.2 mg), and ascorbic acid

‘Chief, Plant Taxonomy Laboratory, Plant Genetics and Germ-
plasm, Institute, Agricultural Research Service, Beltsville, Maryland.

> WARE Institute, Inc., Box 2599, Madison, Wisconsin.

3 Fy ° 7 F .
Research Associate in Economic Botany, Botanical Museum,
Harvard University, Cambridge, Massachusetts.

[ 113 ]

(1.4 mg: 18.0 mg). Ingestion of 100 g of the Bolivian
coca leaves tested would more than satisfy the Recom-
mended Dietary Allowance for reference man and
woman of calcium, iron, phosphorus, vitamin A, vita-
min B, and vitamin EK. However, the leaves also contain
alkaloids and may harbor pesticide residues.

Unable to establish the nutritional value of coca leaves
(Mrythrovylum Coca Lam. )afterconsulting many sources,
we obtained a one kilogram sun-dried sample from San
Francisco, Province of Chapare, Bolivia, in June, 1974.
Using methods listed in References and Notes, we ob-
tained the following nutritional analysis (1): calories, 805
per 100 @: moisture, 6.5 g; protein, 18.9 @: carbohy-
drate, 46.2 @; fat, 5.0 g: vitamin A, 11,000 LU (as
beta-carotene); vitamin C, 1.4 mg; vitamin B, (thia-
mine), 0.85 mg; vitamin B, (riboflavin), 1.9 m@: niacin,
1.29 meg: calcium, 1,540 mg: iron, 45.8 me: vitamin
Ki, 43.5 [LU (as d-alpha tocopherol): vitamin B,, 0.508
mg: folicacid, 0.1830 mg: vitamin B,,, 1.05 meg: iodine,
5.0 meg: phosphorus, 911 mg: magnesium, 213 mg;
zinc, 2.70 mg; copper, 1.21 mg: biotin, 0.0868 mg;
pantothenic acid, 0.684 mg: and sodium, 40.6 mg. The
analysis of other elements by emission spectroscopy
yielded the following amounts: potassium, 2.02 @ per
r: stronti-

S

_

100 @; aluminum, 89.5 mg; barium, 4.67 m
um, 9.71 mg; boron, 5.85 mg: zine, 2.70 mg; manga-
nese, 6.65 mg: and chromium, 0.359 mg.

Surprised by the high values, especially in calcium and
iron, we tabulated nutritional averages for other plant
products ingested by Latin Americans (‘Table 1). Com-
pared with an average from ten nuts and oilseeds (2)
(Sesamum indicum, Terminalia Catappa, Prunus A myg-
dalus, Corylus spp., Arachis hypogaca, Castanea spp.,
Bertholletia evcelsa, Helianthus annuus, Anacardium
occidentale and Inga spp.), the San Francisco coca leaves

( 114 ]

were higher in protein, carbohydrate, ash, calcium, phos-
phorus, iron, vitamin A and riboflavin. Coca was lower
in calories, moisture, fat, thiamin, niacin, and vitamin
C. Compared with an average of ten pulses (Vigna un-
guiculata, Cicer arietinum, Cajanus Cajyan, Pisum sativum,
Vicia Faba, Phaseolus vulgaris, Dolchos Lablab, Lens
spp., Glycine Max and Lupinus mutabilis), coca was
equal in fat; higher in fiber, ash, calcium, phosphorus,
iron, vitamin A, and riboflavin; and lower in calories,
moisture, protein, carbohydrate, thiamin, niacin, and
vitamin C. Compared with an average of ten cereals
(Amaranthus caudatus, Oryza sativa, Avena sativa, Chen-
opodium pallidicaule, Chenopodium Quinoa, Hordeum
vulgare, Secale cereale, Coix Lachryma-jobi, Zea Mays
and Triticum aestivum), coca was higher in protein, fat,
fiber, ash, calcium, phosphorus, iron, vitamin A, ribo-
flavin, and vitamin C; lower in calories, moisture, carbo-
hydrate, thiamin, and niacin. Compared with an average
of ten vegetables (Canna edulis, Capsicum spp., Allium
sativum, Arracacha wanthorrhiza, Ipomoea Batatas, Cy-
clanthera pedata, Cucurbita mavima, Allium Cepa, Bras-
sica oleracea, and Tropaeolum tuberosum) and an average
of ten fruits (Persea americana, Ananas comosus, Musa
sapientum, Cocos nucifera, Passiflora mollissima, Annona
Cherimolia, Prunus persica, Fragaria spp., Annona mu-
ricata, and Meus Carica), coca was high on all counts
except moisture and vitamin C.

The present coca analyses are comparable to an average
of three earlier coca analyses from Bolivia (8) and three
recently reported from Peru (4). Frequent reports that
coca has no nutritional value should be re-evaluated in
view of these findings. The comparatively high nutri-
tional values for coca are due partly to the fact that the
leaves are dry (less than 10% moisture) when purchased,
whereas most other foods are higher in moisture.

[ 115 |

In most areas where it is used, coca should be con-
sidered a masticatory since it is not wholly consumed by
the chewer. Typically, the leaves are first moistened in
the mouth with saliva, then formed into a quid with the
tongue and pushed into the upper cheek cavity. They
are then sucked to extract the rich, green Juice which is
subsequently swallowed. Usually some form of alkali
is added to facilitate this extraction. When the chew is
exhausted, it is usually spat out. Thus, the full comple-
ment of nutrients present in the coca leaf is not consumed
entirely, and the nutritional amounts reported here may
be somewhat higher than the amounts actually ingested
by the coca chewer. To our knowledge, no studies have
been made on the nutritional value of the swallowed
extract.

In the Colombian Amazon, a variation of coca use is
practiced by several tribes. Coca leaves are pulverized
to a fine powder along with the ashes of Pourouma or
Cecropia leaves. The mixture is placed in the mouth on
the gums and inner cheeks and is eventually swallowed
(8). In this case, and in instances where a coca chewer
swallows his quid, the full complement of the leaf nutri-
ents would be ingested. There is essentially no difference
here between the use of coca and the direct consumption
of food, in terms of nutrition.

The amounts of coca consumed may contribute signifi-
‘cantly to the diets of Andean coca chewers (5). If the
average chewer ingests 60 g of Peruvian coca per day
(5,6), he more than satisfies his requirements for calcium,
even without the supplemental ash or lime usually added
to the coca quid (6). No other food in the INC AP Food
Composition Tables (2) approaches coca for calcium con-
tent (1,789 mg). Other food items are high in calcium:
sesame seeds, 1212 mg per 100 @: spinach flour, 488 mg:
leaves of Laurus nobilis, 803 mg: leaves of Justicia pecto-

[ 116 ]

ralis, 668 mg; West Indian Almond ( Terminalia Catap-
pa), 497 mg; powdered skim milk, 1,801 mg; whole
milk, 921 mg; and alligator meat, 1,281 mg; but none
equals coca. Few food plants can satisfy the calcium and
iron in the Recommended Dietary Allowance (RDA)
of reference man ingesting 100 g. The Bolivian coca
leaves reported here do satisfy the RDA.

Coca leaves may, however, contain 0.25 to 2.25% toxic
alkaloids, including benzoylecgonine, benzoyltropine,
cinnamylcocaine, cocaine, cuscohygrine, dihydroxytro-
pane, hygrine, hygroline, methyleocaine, methylecgoni-
dine, nicotine, tropacocaine, and a - and £ -truxilline
(4,7). These alone could make the nutritious coca leaf
undesirable as a source of nutrients. he average coca
chewer could also ingest 442 mg of copper in a year if
the San Francisco leaves are typical; but this amount
is not excessive.

Many coca growers in both high (Chapare, Bolivia)
and low (Yungas, Bolivia) rainfall areas may use insecti-
cides. The leaves are not intentionally washed, and har-
vest and curing are timed to avoid rainfall. In considering
coca for human consumption, the leaves should be ana-
lyzed for insecticide residues.

Although coca leaves contain relatively high levels of
certain nutrients, the presence of alkaloids and the pos-
sible presence of insecticide residues suggest caution in
coca chewing.

[ 117 ]

REFERENCES AND NOTES

Methods: Protein (Yo N 6.25), Association of Official Analyti-
cal Chemists (A.O.A.C.), 11th Ed., 16. 1970; moisture, vacuum
oven, A.O.A.C., 122. 1970; ash, A.O.A.C., 123. 1970; fat,
A.O.A.C., 129. 1970; carbohydrates, by difference; calories, by
calculation; vitamin A, Moore & Ely, Ind. Eng. Chem. Anal.
Ed., 13: 600. 1941; vitamin C, J. Biol. Chem., 147: 399. 1948;
vitamin B,, A.O.A.C., 771. 1970; vitamin Bz, A.O.A.C., 789.
1970; niacin, A.O.A.C., 787. 1970; elemental, J.A.O.A.C., 51:
10038. 1968; vitamin E, Acta Chemica Scandinavica 11: 34-43.
1957; vitamin Be (Streptococcus carlsbergensis), Atkins, Schultz,
Williams & Frey, Ind. & Eng. Chem., Anal. Ed., 15: 141. 1948;
folic acid, A.O.A.C., 786. 1970; vitamin B,2, U.S.P. 17: 864.
1965; iodine, ashing, A.O.A.C., 674. 1970; colorimetry, W.T.
Binnerts, Anal. Chemica Acta 10: 78. 1954: biotin (Lactobacillus
arabinosus), Wright & Skeggs, Proc. Soc. Exp. Biol. & Med., 56:
95, 1944; pantothenic acid, Nielands & Strong, Arch. Biol., 19:
2. 1948,

Wu Leung, W.& M. Flores. 1961. Tabla de Composicién de Ali-
mentos para Uso en América Latina. Instituto de Nutrici6n de
Centro América y Panama (INCAP) and Interdepartmental Com-
mittee on Nutrition for National Defense (ICNND). U.S. Govern-
ment Printing Office. Washington, D.C.

del Granado, J.T. 1931. Plantas Bolivianas. Arno Hermanos.
La Paz.
Machado, E. 1972. El género Erythroxylon en el Peru. Ray-

mondiana 5: 5-101.

Hanna, J.M. 1974. Coca Leaf Use in Southern Peru: Some Bio-
social Aspects. Am. Anthropologist 76 (2): 281-296.

Baker, P.T. & R.B. Mazess. 19638. Calcium: Unusual Sources in
the Highland Peruvian Diet. Science 142: 1466-7.

Willaman, J.J. & B.G. Schubert. 1961. Alkaloid-bearing plants
and their Contained Alkaloids. U.S. Department of Agriculture.
Washington, D.C.

Schultes, R.E. 1957. A new method of coca preparation in the
Colombian Amazon. Bot. Mus. Leafl. Harvard Univ. 17 (9): 241-

246,

[ 118 ]

e1ep OU »

JUIJOIBI-BjJ9q SB I

062 800 900 SOO CE 80 €& 02 LO FT Ihl Sh eT 96L £6 (OT) Lt |
OTE OT S00 =660°0 = S6S ant cS 96 60 ST 691 vO 8T E28 bl (OT) S9]qe}a39 A
80 Le ScO 170 €I SP 9PE FL Tc OF OTL L2E€ LIT SIl ce (OT) s]ee1o)
61 S@e #20 8S0 02 TZ 86€ col ee gs Toss 0S VSt EIT HSE (OT) ses[nd
Te ¢S 8¢0 820 LI eV G6S ELS Te Ge 68¢ O98 8S9T 66 T2o (OT) SP99S FY SInN
. es : : : , ; . ; . ADVAAAV
O'ET 2 810 880 SEI gE OLE 66 0G GE Tle 66 VIL OOF 622 (0S) dOOd LNW 1d
VI Le €ZT 890 O00;0T s9c L269 6841 £9 El EC Cf BRI Ss — (L) AOVAAAV VOOO
— Lv9 SST 180 0006 62 €9€ 8806 OF GLI = — £51 £0 = () 2909 Nilo
= — = = = = a 7 eS O08 ver OT — 88 a (¢) Bote BANE
VI 6¢T I6T SEO OOOIT Sch TII6 OFT O06 FFT ZF OS 6ST G9 SOE (T) 2905 oosIouRIYy URS
su sur sul BU OI su Su su 3 3 3 3 3 3 a]duies
OVA BIN: GY PUL WHA d PD USVI “qiey 63eY 301d = OFH [8D uy ae

‘Spooy JuR[q UeeWYy Ue] Jayi0 YIM SaAvaT] B09 jo 3 QOT Jed uostedwosd jeuONaNN

T ATaAVL

[119 J

a4

DE PLANTIS TOXICARIIS EF MUNDO
NOVO TROPICALE COMMENTATIONES XIII

NOTES ON POISONOUS OR MEDICINAL MALPIGHIACEOUS
SPECIES OF THE AMAZON
BY
Ricuarp Evans ScCHULTES

This paper is offered in continuation of a series pre-
senting ethnopharmacological observations made in the
northwest Amazon. The malpighiaceous species enumer-
ated in the following pages—all apparently unstudied
from the phytochemical and pharmacological points of
view—have been found to have interesting applications
amongst the Indians of the region.

The identifications of the voucher specimens were
made by Dr. José Cuatrecasas of the Smithsonian In-
stitution and are cited and described in his monograph
of Colombian Malpighiaceae which appeared in Webbia
13 (1958) 843-664. The drawings herein produced were
prepared by Mr. Joshua B. Clark.

The ethnopharmacological observations were made by
the writer in the Amazonian basin of Colombia between
1941 and 1954. They are offered in the hope of stimu-
lating research to discover new biodynamic principles or
to find new potentially valuable medicinal plants. Re-
search along these lines is especially needed for this family
of 60 genera and more than 880 species which has re-
cently been described as ‘‘scarcely touched, chemotaxo-

[121 J

nomically speaking’. (Gibbs, R.D. ‘‘Chemotaxonomy
of Flowering Plants” 8 (1974) 1676-1677.)

The best known members, from the point of view of
their phytochemistry, are undoubtedly the several closely
allied species of Banisteriopsis—B. Caapi, B. inebrians,
B. Rusbyana—employed in northern South America in
elaborating the narcotic drink known variously as aya-
huasca, caapi, natema, pinde and yajé. ‘These species
contain psychoactive alkaloids—the first two, B-carboline
alkaloids; thethird, N, N-dimethyltry ptamine( Deulofeu,
V. ‘Chemical Compounds Isolated from Banisteriopsis
and Relative Species’ in [| Ed.D. Efron] ‘‘Ethnopharm-
acologic Search for Psychoactive Drugs’” Public Health
Service Publ. No. 1645, Washington, D.C. (1967) 893-
402).

One hundred and twenty-five years ago, when Spruce
collected the type material of Banisteriopsis Caapi and
described its narcotic use in the northwest Amazon of
Brazil, he despatched material for chemical analysis—
material which was not analyzed until 1968 (Schultes,
R.E., B. Holmstedt and J.-E. Lindgren *‘De Plantis
‘Toxicariis e Mundo Novo Tropicale I11. Phytochemical
Kxamination of Spruce’s Original Collection of Banis-
teriopsis Caapr” in Bot. Mus. Leafl. Harvard Univ. 22
(1969) 121-182). Spruce was struck by the presence in
this family of such a potent narcotic. In 1852, he wrote:
‘2... Lsaw, not without surprise, that it belonged to
the order Malpighiaceae and the genus Banisteria, of
which I made it out to be an undescribed species. .. .
My surprise arose from the fact that there was no narcotic
Malpighiad on record, nor indeed any species of that
order with strong medicinal properties of any kind... .
The seed [of Bunchosia] is described in books as poison-
ous, and if it be really so, then it is the only instance,
so faras I know, of the existence of any hurtful principle

[ 122 ]

in the entire family of Malpighiads, always excepting
that of the Caapi. Yet strong poisons may lurk undis-
covered in many others of the order, which is very large
.... (Spruce, R. [Ed. A.R. Wallace] ‘‘Notes of a
Botanist on the Amazon and Andes”’ 2 (1908) 421-422).

Banisteriopsis Martiniana (Juss.) Cuatrecasas
var. laevis Cuatrecasas in Webbia 18 (1958) 502.

Cotompia: Comisaria del Amazonas, Vaupés, Rio Apaporis, Raudal
de Jirijirimo. H. Garcia-Barriga 13706.—Same locality. November
25,1951. R.E. Schultes et I, Cabrera 14569, —Comisaria del Amazonas,
Rio Apaporis, Soratama. August 20, 1951. Schultes et Cabrera 13615.

Garcia- Barriga (Flora Medicinal de Colombia 2 (1975)
69) has recently reported that Banisteriopsis Martiniana
var. /aevis may be used by the Makuna Indians of the
middle Apaporis as one source of the narcotic prepa-
ration yajé.

Heteropteris riparia Cuatrecasas in Webbia 13
(1958) 483.

Cotomsia : Comisaria del Putumayo, Rio Putumayo, Puerto Ospina.
Alt. 300 m. ‘‘Extensive liana along river’s edge.’’ July 1942. R.E.
Schultes 4029.—Comisaraia del Vaupés, Rio Apaporis, Raudal de
Jirijirimo. Alt. 250 m. June 12, 1951. R.E. Schultes et I. Cabrera
12428,—Comisaria de] Amazonas, Rio Apaporis, Soratama. June 18,
1951. Schultes et Cabrera 12654.—Comisaria del Vaupés, Rio Apaporis,
Jinogojé (near mouth of Rio Piraparana). Alt. 260 m. ‘Vine. Flowers
yellow.’’ June 5, 1952. Schultes et Cabrera 16591.—Same locality.
June 8, 1952. Schultes et Cabrera 16665.

The bark of this extensive liana is reported in these
widely separated localities to be extremely poisonous,
but the Indians in the middle Apaporis rasp the bark
and prepare a tea which is employed internally in the
treatment of gonorrhoea. The Makuna name on the Rio
Apaporis 1s mee-see-gaw.

Saponins, tannins and phenolic acids have been re-
ported from the genus Heteropteris (Gibbs, loc. cit. ).

| 223.

Heteropteris macrostachya 4. Jussieu Malpigh.
Synop. (1840) 275.

CoLtombBia: Comisaria del Amazonas, Rio Apaporis, Soratama. Alt.
250 m. ‘‘Bejuco; hojas doradas envés; flores amarillas. Nombre en
lengua ‘geral’ capituriiva.’’ August 16, 1951. R. FE. Schultes et I.
Cabrera 13533,

The seeds of Heteropteris macrostachya are reputedly
taken inatea by the Taiwano Indians of the Rio Kana-
nari to treat diarrhoea.

It may be significant that the seeds of Heteropteris
suberosa Griseb., commonly called sarabatuciv, are simi-
larly employed on the Rio Mauhés in the central Amazon
of Brazil (Le Cointe, P.: **A Amazonia Brasileira’ 3
(1934) 406).

A collection of Hleteropteris macrostachya from Pana-
ma (1..M. Johnston 1577) bears an annotation that the
plant is ‘‘toxic™.

Hiraea apaporiensis (watrecasas Webbia 18 (1958)
404.

Cotomsia: Comisaria del Vaupés, Rio Apaporis, Raudal Yayacopi
(La Playa). August 18, 1952. R.E. Schultes et I. Cabrera 16969.—
Jinogojé (near mouth of Rio Piraparana). August 25, 1952. Schulltes
et Cabrera 17022,.—Comisaria del Vaupés, Rio Piraparana. August
1952. H. Garcia Barriga 14215.

The Maku Indians, who know Hiraea apaporiensis as
ye-aing, employ a tea prepared from the leaves which is
valued in the treatment of conjunctivitis.

Hiraea Schultesii Cuatrecasas Webbia 18 (1958)
408.

CoLomspia: Comisaria del Vaupés, Rio Apaporis, Raudal Yayacopi
(La Playa). March 16, 1952. R.E. Schultes et I. Cabrera 15996.

A wash of the leaves of Hiraea Schultesii is valued by
the Makuna Indians of the middle Apaporis in treating
severe conjunctivitis.

[ 124 |

PiatTe AXTV

paria
Cuatr.

pry

€ 5
(th.

=——

So pe ==

Mascagnia glandulifera Cuatrecasas Webbia 13
(1958) 865.

Cotompia: Comisaria del Amazonas, Rio Apaporis, Soratama. July
31, 1951. RE. Schultes et I. Cabrera 15208,—Same locality. August
16, 1951. Schultes et Cabrera 13594.

A poultice of crushed and boiled leaves of Mascagnia
glandulifera is commonly applied to boils and similar
infections by Indians of the middle Apaporis.

Saponins have been reported from a species of this
genus (Gibbs loc. cit. ).

Mezia includens (Benth.) Cuatrecasas in Webbia
13 (1958) 450.

Cotompia: Comisaria del Vaupés, Rio Apaporis, Jinogojé (at mouth
of Rio Piraparand) and vicinity. “‘Vine. Flowers bright yellow.”
RE. Schultes et I. Cabrera 15691.

This vine represents one of the medicinal plants with
most uses amongst the Makuna Indians of the middle
Apaporis. It is unusual that this should be so, in view
of the relative scarcity of the plant.

The bark, which contains apparently high concentra-
tions of tannin, is prepared in an infusion which is taken
warm in large quantities to treat urinary troubles and to
provoke urination. Perhaps because of its diuretic proper-
ties, it isemployed in the form of a tea in cases of swol-
len legs—obviously edemas due to age and circulatory
ailments.

The root is considered a strong laxative: it is crushed
and soaked together with water in which /arvva (the flour
prepared from Manihot esculenta has been setting for
several hours.

The leaves, boiled for long periods into a tea, provide
a strong emetic drink. They are also considered, when
applied over the abdomen in the form of a cataplasm, a
help in the treatment of what appears to be jaundice.

The Makuna Indians call this vine ce’-taw-gai.

[ 126 }

PLATE XXV

HIRAEA apaporiensis cuatr.

<a be . hy \
AIG \ yy

WN

ns
Pé

4

\\ . \\ \
Sy \ \
NY

Tetrapteris mucronata Cavanilles Diss. 9 (1790)
4354, t. 262.

CotomBia: Comisaria del Amazonas, Rio Apaporis, Soratama. June
15, 1951. RE. Schultes et I. Cabrera 12554,—Comisaria del Vaupés,
Rio Apaporis, mouth of Rio Pacoa. June 15, 1951. Schultes et Cabrera
13559.—Rio Apaporis, Jinogojé (near mouth of Rio Piraparana). June
20, 1952. Schultes et Cabrera 16771.

The natives of the lower Rio Piraparana prepare a
weak type of curare by boiling together for four or five
hours the bark of Tetrapteris mucronata and of Strychnos
Erichsonti R. Schomb. (Schultes et Cabrera 16770).

Tetrapteris mucronata is said to be ‘employed by the
Karaparana tribe in preparing yajé’ (Schultes et Cabrera
12107). In this connection, it is of interest to note the
report of the narcotic use of a species of Tetrapteris, 7.
methystica (Schultes et Lopez 10184), in the Rio Negro
basin of Brazil (Schultes, R. E. in Bot. Mus. Leafl.,
Harvard Univ. 16 (1954) 202-205.

Tetrapteris silvatica Cuatrecasas Webbia 13
(1958) 425.

Cotompia: Comisaria del Vaupés, Rio Apaporis, Raudal Yayacopi
(La Playa). August 18, 1952. R.E. Schultes et I. Cabrera 16960,

The leaves of Tetrapteris silvatica are burned by the
Makuna Indians, and the ashes, mixed with any oil or
grease, are applied to what appear to be fungal patches
on the skin. The Makunas refer to this vine as ¢fee-mee-
a-mee-see-ma and recognize its very close relationship
with the following species, Tetrapteris styloptera.

Tetrapteris styloptera Jussieu in Ann. Sci. Nat.,
ser. 2 Bot. 13 (1840) 202.

Cotompia: Comisaria del Amazonas, Rio Igaraparana, La Chorrera
and vicinity. June 1942. R.E. Schultes 3914.—Same locality and
date. Schultes 3927.—Rio Apaporis, Soratama and vicinity. August

3, 1951. RIE. Schultes et I. Cabrera 12880,.—Same locality. August

(128 ]

Piatt XX VI

MASCAGNIA) glandulifera cuatr.

[ 129 ]

16, 1951. Schultes et Cabrera 13573, —Comisaria del Vaupés, Rio Apa-
poris, mouth of Rio Pacoa. June 17, 1951. Schultes et Cabrera 12597 .—
Jinogojé (near mouth of Rio Piraparana). June 5, 1952. Schultes et
Cabrera 16591.—Same locality. June 8, 1952. Schultes et Cabrera
16667.—Same locality. June 20, 1952. Schultes et Cabrera 16778.—
Same locality. September 20, 1952. Schultes et Cabrera 17600. — Raudal
Yayacopi (La Playa). August 18, 1952. Schultes et Cabrera 169523 ,.—
Same locality and date. Schultes et Cabrera 16961.—Cachivera Jirijiri-
ma. November 1951. H. Garcia-Barriga 13700,

The ‘Tanimuka Indians on the Rio Miritiparana know
this yellow-flowered vine as wee-po-awh. The bark is
rasped and boiled to prepare an extremely bitter drink
taken as a febrifuge.

Amongst the Makunas, the leaves, reduced to ashes,
are applied to itching infections of the skin that may be
due to fungal growth. The ashes are mixed with oil or
fat for application. This vine is one of the most impor-
tant medicines amongst the Makunas, who have three
distinct names for it: bce-ra-ree-a-ma, ho-ree-a-mee-see
and né-na’-mee-hoo-ma.

(130 |

Puate XXVII

TETRAPTERIS
styloptera Juss.

vA TETRAPTERIS
| mucronata Cav.

[131 |

ECONOMIC BOTANY LIBRARY
OF OAKES AME)

RVARD BOTANICAL MUSEUM

BOTANICAL MUSEUM LEAFLETS
HARVARD UNIVERSITY

CampripGr, Massacuusetrts, December 12, 1975 VoL. 24, No. 7

STUDIES IN THE GENUS ONCIDIUM. IL.
BY
Joun E. Stacy*

The genus Oncidium, one of the largest orchid genera
in the New World, has been of continuous intrigue to
both botanists and horticulturists. Although during the
past 175 years several attempts have been made to clarify
the various species that comprise this genus, no satis-
factory compendium is presently at hand. It was Lind-
ley who first produced a kind of monograph, or rather a
revision of the genus in 1841, in his Sertum Orchidaceum
under t. 48 with 101 species, and an expanded version
of it in 1855 in his Folia Orchidacea with 198 species.
These treatments were in use almost uninterruptedly
until 1922 when anew attempt was made by Kraenzlin.
It is unfortunate that the Kraenzlinian version was any-
thing but an improvement on the former treatments.
Since that time two synopses have appeared, one in 1970
by Garay (Taxon 19: 443-467, 1970), and the other
jointly by Garay and me in 1974 (Bradea 1(40): 893-
424, 1974). IL believe that in this latter synopsis we have
clearly outlined the infrageneric divisions which I intend
to follow section by section in this series. These papers
will not follow the sequence outlined in the key to the
sections, but rather will treat them randomly depending
on the demand and need of immediate attention. Conse-

* Research Fellow, Orchid Herbarium of Oakes Ames.

[ 183 ]

quently this first study deals with the Section Cucullata.

The actual identity of the plants with colorful and
attractive flowers of the Section Cucullata have been
plagued with confusion and misunderstanding since the
early days of their introduction to European horticul-
ture. During the nineteenth century hundreds of these
plants were flowered in England and Europe and many
of their floral variations were noted. For the most part,
the early botanists considered the morphological differ-
ences as being minor and not worthy of specific rank.
Lindley, for example, chose to combine several different
plants as varieties of O. cucudlatum including his earher
described O. nubigenum. This action, since no univer-
sally accepted Code of Nomenclature existed at that
time, prompted Reichenbach to remark (Gard. Chron.
367, 1867) “°... he (Lindley) called this plant a good
species (nubigenum) but when he was older it had come
into fashion to combine very heterogeneous types into
one species because there were too many species’. De-
spite Reichenbach’s attempt to separate these plants into
distinct entities, the practice started by Lindley con-
tinued and O. Phalaenopsis, O. oltvaceum as well as O.
andigenum were declared to be varieties of O. cucullatum
by Nichols in 1886, Veitch in 1892, and Gower in 1899,
respectively.

Reichenbach maintained a vigorous separation of the
species in this affinity, attested by his drawings and de-
scriptions. In 1922, Kraenzlin, in his monograph of On-
cidium, attempted to expand the work of Reichenbach
by establishing the Section Cucullata as a distinct unit.
This effort, although noteworthy, included several spe-
cies which belong to other affinities. He also omitted
O. andigenum which he placed ina section together with
QO. cornigerum tor reasons known only to him. In general
Kraenzlin well understood his Section Cucullata, but he

[ 184 |

was rather careless and inconsistent in the allocation of
the species.

The circumscription of a section composed of seem-
ingly similar flowering plants in itself is not a difficult or
tedious task, but the delimitation of the species assigned
to a given section is altogether a different matter. This
latter approach demands a careful examination of all type
specimens along with the associated descriptive data.
Living material must be studied in the field and in cul-
tivation, and compared with the types. Camera lucida
drawings of the floral parts of type specimens as well as
of additional material must be prepared for comparative
purposes. Data presented in this paper, and those to be
published in the future, are assembled through these
methods.

The plants of the Section Cucullata are endemic to the
Andes of South America and grow in the cold and humid
climate found at about 8,000 meters. Their range is from
Venezuela southward to northern Peru, with an obvious
concentration in Keuador. In the mid-nineteenth cen-
tury, several collectors shipped plants by the thousands
to England and Europe, which formed the basis for many
of the herbarium specimens. These plants are relatively
common in living collections today, although as, ex-
pected, in rather small numbers, for repeated field trips
to previously lucrative areas vividly tell of the encroach-
ment by man. There is no doubt that these plants will
continue to be found, but certainly not in the presently
accessible areas known to current collectors.

Because of the rather peculiar relationship between the
‘allus of the lip and base of the column, I entertained
the thought that the Section Cucullata should be ex-
cluded from the genus Oncidium. In all of the species,
the base of the column has a tumid member on either
side projecting forward and enclosing the base of the

[ 135 ]

‘allus. Most of the species of the other sections of On-
cidium have a short, variously tumid column-base, but
only the Cucullata types enjoy the horseshoe-like shape
which shelters part of the callus. Yet, analogous situa-
tions exist in the Section Stellata, which is a different
type of modification of the basic pattern. Until we have
a satisfactory key to the various genera of the Oncidium
alliance, I decided to maintain this section under
Oncidium.

In the past, several Odontoglossum species of the O.
rigidum relationship have been referred to the Section
Cucullata because of the connate lateral sepals and the
sessile sepals and petals. A study of all the characters has
necessitated the exclusion of these species from the genus
Oneidium tor the above listed characters document an
obvious case of parallel evolution rather than similarity
based on close relationships. Likewise Oneidium Storhki
and O. Warscewiczi must also be excluded from the
Section Cucullata due to the different columnar struc-
tures of the flowers.

Several noteworthy characters within the section serve
to distinguish the species. Of the 19 species recorded in
this paper, ten have a tuft of hairs at the base of the
elongate callus, while the remainder have a short tumid
‘allus without hairs. With respect to the calli, three
species—O. andigenum, O. mimeticum and O. sanguino-
lentum—have five subglobose tubercles in three rows.
All others are with three-membered, variously merged,
tumid or keel-like structures. The familiar long nose-like
tubercle is found in O. eucullatum, O. olivaceum, O.
Kennedy and O. rhodostictum; long, sulcate callus which
is tridenticulate at the apex, on the other hand, is limited
to O. spathulatum, O. tripterygium and O. Dayanum.

Five species—O. cucullatum, O. mimeticum, O. oliva-
ceum, O. rhodostictum and O. sanguinolentum—have a

prominently cucullate clinandrium, while the remainder
have only a low collar which partially encircles the anther.

The lip is the most conspicuous part of the flower in
each species and contributes greatly to the overall size.
Most species have a variously constricted lip with a well-
defined isthmus or with a cuneate base to the midlobe
suggesting a trend toward an isthmus. Otherwise, the
midlobe is sessile and cordate, precluding an isthmus as
in O. nubigenum.

The most important non-floral character within the
section is found in the inflorescence. Only O. cucullatum,
O. Kennedyi and O. oltvaceum have a long, erect, rigid
peduncle with a fractiflex rachis, while all others have a
flaccid, arcuate peduncle with a sinuously flexuous rachis.
This character alone is sufficient to separate the three
species mentioned from the rest in the section.

Section Cucullata A7vz/. in Pflanzenr. Heft 80:
128: 1922.

Type: Oneidium cucullatum Vind).

The flowers range between 2 to 4 cm. vertically and
have attractive color patterns. The sepals and petals are
subsimilar, ovate-oblong, sessile, acute, and often con-
cave. The color is variable, frequently mottled, often
solidly colored with a well-defined margin. The lateral
sepals are variously connate, always shorter than, and
hidden by, the lip. The lip is conspicuous, usually con-
stricted in the middle, with or without a well-defined
isthmus; the side lobes subquadrate to rounded and often
deflexed. The midlobe is spreading, emarginate, vari-
ously maculated, occasionally on both sides. The crest
consists of an odd number of keels or tumors with or
without a hirsute base. The column is short, glabrous,
occasionally with a cucullate clinandrium, and a thick-
ened base; the rostellum is short.

[ 137 ]

The plants are rather small with aggregate or approxi-
mate ovoid to pyriform and slightly compressed pseudo-
bulbs which are enveloped by several imbricating sheaths
of which the uppermost may be leaf-bearing. The one or
more terminal leaves are linear to elliptical or narrowly
lanceolate and subcoriaceous. The inflorescence, fre-
quently precocious, emerges laterally from the basal
bracts and is usually slender, arching, few-flowered, and
racemose; occasionally it is rigid and erect, or many-
Howered and paniculate, often conspicuously fractiflex.

Key to Species

Part of callus enclosed by the auriculate base
of column adorned with a tuft of hairs . . . . . . . . . 2

_

la. Part of callus enclosed by the auriculate base

of column without a tuft of hairs . . . . . . . . 2. +. . 158
8 Inflorescence rigid, erect, with more or less
flexuous rachis . . . 0.0.0. ee ee ee ee SB

Ya. Inflorescence flaccid, arcuate with sinuous

PCN a ke RRS es ee RO eS ee
3. Column with erect clinandrium, callus of lip

keel-like, one-fourth to one-sixth of entire

length of lip, triangular in lateral view, with

a small, divaricately spreading lobe resembling

nostrils on each side « « 6% 6 @ 6 se eee ee we &
3a, Column with dorsally reclining clinandrium ;

callus of lip keel-like, one-third of entire

length of lip, triangular in lateral view consist-

ing of three, parallel, distinct ridges of which

the median is longer and nose-like without

small spreading lobes on each side. . . . . 2. 2. 2 we OD
4, Inflorescence racemose, rarely with a few short

branches; sepals at least 16 mm. long; lip 30

mm. long... 0... ee ee ee ee ee OF olivaceum

var. olivaceum

4a, Inflorescence paniculate with numerous, long,

diffused branches; sepals up to 10 mm. long;

lip up to20mm.long. . . . . . . 2 2. . . OW olivaceum

var. giganteum

[ 138 ]

6a,

8a.

io}

10a,

Ll,

lla.

Column with distinct, linear processes on each

side of the stigma; disc of lip with scattered

wart-like papillae in center. . . Wee ew . OJ Kennedyi

Column without any processes on each side of stigma;

dise of lip without any wart-like papillae in center .... 6

Flowers medium size; sepals not more than

15 mm. long; lateral lobes of lip oblong-

rectangular with obliquely truncate apex . . . O. cucullatum
var. cucullatum

Flowers large; sepals at least 20 mm. long;

lateral lobes of lip never oblong-rectangular .

Lateral lobes of lip rectangularly spreading
obliquely triangular or dolabriform . . . . . O. cucullatum
var. dolabratum
Lateral lobes of lip recurved toward midlobe,
obliquely ovate. 2. 2... 2 ww ee ee OW cucullatum
var. macrochilum
Callus of lip elongate, consisting of three
variously confluent, parallel ridges .

&%

Callus of lip short, consisting of five confluent

tubercles in three rows, two on each side and

the fifth in the middle. . ..........2.2.2. 2.18
Petals unicolored except margins; clinandrium

cucullate; callus of lip compressed, keel-like,

the median ridge much longer than the very

obscure lateral ridges, nose-like laterally . . . O. rhodostictum
Petals marmorate to barred; clinandrium

collar-like, marginate; callus of lip not com-

pressed, the median ridge often considerably

shorter than the lamella-like lateral ridges,

never nose-like . . . .. . ae . ee eae ae
Lateral sepals connate into a bifid sy aoa

petals froma cuneate base elliptic-lanceolate,

abruptly acute to acuminate .............421
Lateral sepals free at least to middle, spread-

ing; petals broadly elliptic, sessile, obtuse . . O. spathulatum

Callus of lip obovate in outline, concave with

entire, raised margins, bidentate at apex and

with a shorter keel in the center originating

from the middle of the callus . . . . . . . OJ Phalaenopsis
Callus of lip quadrate to oblong in outline,

conver, consisting of three parallel ridges of

which the two outer ones are obscurely lobu-

[ 139 }

12

a

12a.

12.

13a.

14,

14a,

7a.

18a.

19a.

20.

late, and the third in the center originating
from the base of the lip between the column-

WINGS: 2 6 oo Se

Median keel much shorter than the beens ones

Median keel as long as the lateral ones .
Column slender; clinandrium low,

clinandrium high, cucullate

Column stout;
Flowers rather small: lip when expanded
distinctly pandurate, i.e.,
middle; midlobe of lip truncate to cuneate
OCDASE 4 ce Se ee Ee

Flowers medium to large, lip when expanded
never pandurate, but constricted above the

middle; midlobe of lip transversely reniform,

constricted in

deeply cordate at base .

Lip when flattened truncate to subcordate at
BOR % ee & 6 4 Re eS ee ee eS
Lip when flattened cuneate at base

Midlobe of lip sessile, transversely elliptic
to cordate-reniform ew we ee
Midlobe of lip cuneate to cuneate-flabellate
Sepals and petals conspicuously muricate-
papillose dorsally, especially towards base;
petals distinctly mucronate at the more or
less emarginate apex... ae ae
Sepals and petals glabrous ee oe petals
neither mucronate nor emarginate at apex
Lateral lobes of lip ovate-oblong, forming an
acute to subacuminate, open sinus with
median lobe; callus large, basally produced
into a retrorse, conical point . . . . . . .
Lateral lobes of lip subrotund to obliquely
obovate with rounded lobes, often imbricating

the base of median lobe ; callus small without

a retrorse, conical point . ........

Flowers medium size; lip 15-17 mm. long;
callus quadrate in outline, porrect, composed

of three ridges . . . . .. ee

Flowers large: lip 22-24 mm. eee eallus
transversely rectangular-oblong, distinctly
curved upwards in front without any ridges
but slightly depressed. . .....25.-.

Petals with transverse bars; lateral lobes of

(140 |

marginate .

12

0. Dayanum
O. tripterygium
O. andigenum
14

O. sanguinolentum

O. mimelicum

O. azuayense

O. nubigenum

. O. aequinoctiale

O. alticola

O. tarquiense

lip quadrate, truncate; midlobe of lip with a

rectangularly transverse, erose apex . . . . . QO. erosilabium
20a. Petals without transverse bars; lateral lobes

of lip rounded; midlobe of lip 2-lobulate

without an erose apex oo a oe ey -21
21. Pseudobulbs one-leaved ; lip wider than long,

transversely elliptic in outline . . . . . . O. chimborazoénse
21a. Pseudobulbs two-leaved ; lip as long as wide,

rhombic in outline . 2... . . . . . . . OF. tunguraguense

SYSTEMATICS OF SPECIES

Oncidium olivaceum H.B.K., Nov. Gen. et Sp. PI.

1: 847, 1816.

Syn. : Oneidium cucullatum subvar. olivaceum (H.B.K.)
Veitch, Man. Orch. PI. pt. 8: 31, 1892.

Oncidium cucullatum var. olivaceum (H.B.1K.) Gower

in Garden 55: 88, 1899.

Type: Colombia, Depto. de Cauca, Puracé near Pop-
ayan. Coll. Humboldt no. 2020! (P).

Observation: Dorsal sepal 16 mm. long, 6 mm. wide; lateral sepals
cymbiform, bifid, 16 mm. long, 7 mm. wide; petals 15 mm. long, 8
mm. wide; lip 3-lobed, lateral lobes obliquely ovate, 12 mm. long,
7 mm. wide, midlobe from a narrow isthmus reniform, deeply cordate
at base, bilobed in front, whole lip 30 mm. long, 32 mm. wide across
midlobe. The callus is triangular laterally, keel-like, with a small,
divaricately-spreading lobe, resembling nostrils on each side.

Oncidium olivaceum is readily separable from O. cucullatum by its
callus which is one-fourth to one-sixth of the length of the entire lip
and by the presence of nostril-like lobes on each side of the nose-like
keel, It is noteworthy that I have not seen any material of this spe-
cies outside the Cauca area in Colombia.

Oncidium olivaceum var. giganteum ( Hort.) Stacy,

comb, nov.

Basionym: Oncidium cucullatum var. giganteum Hort.
in Garden 22: 166, 1882.

Type: Reported only from cultivation; no specimen
is known to exist. The original color plate is regarded
here to represent the holotype.

[ 141 ]

Syn. : Oneidium olivaceum var. Lawrenceanum Gower
in Garden 52: 26, 1897.

Neotype: Ecuador, without exact locality. Coll.
Lehmann no. 10081! (x, AMES). Since no type speci-
men has ever been designated, the above-cited collection
by Lehmann is selected here because duplicates of it al-
ready exist in various herbaria under this name.

Observation: The long, almost straight branches produced in pro-
fusion lend a very distinct appearance to the plants of this variety.
I have seen specimens from the Imbabura province of Kcuador and
from the Cauca, near Popayan in Colombia. When more information
will be available of this remarkable plant, it may prove to be a dis-
tinct species.

Illustrations : Cogniaux, Dict. leon. Orch. Oncidium pl. 28, 1899, as.
O. cucullatum; Amer. Orch. Soc. Bull. 37: 13, 1968, as O. cucullatum.

Oncidium Kennedyi Stacy, sp. nov.

Type: Ecuador, along road from Quevedo to Lata-
cunga, ca. 6000 ft. alt. Coll. G. Kennedy s.7./( AMES).

KE. piphytica, caespitosa, usque ad 65 cm. alta; radici-
bus crassiusculis, glabris: pseudobulbis ovoideis, bifoli-
atis, usque ad 7 cm. altis; foliis lineari-oblongis, usque
ad 22 em. longis, 2.5 cm. latis; inflorescentiis arcuatis,
paniculatis, usque ad 65 cm. longis; pedunculo rigidius-
culo, tereti; rhachide fractiflexa; sepalo postico valde
concavo, anguste ovato-oblongo, acuto, 15 mm. longo,
5 mm. lato; sepalis lateralibus inter se connatis, bifidis,
ovato-cymbiformibus, 16 mm. longis, 7 mm. latis: pe-
talis reflexis, ovato-ellipticis, undulatis, obtusis, 15 mm.
longis, 6 mm. latis; labello 8-lobo, lobis lateralibus ob-
lique ovatis, obtusis, lobo intermedio cuneato-flabellato,
antice bilobulato, margine undulato, disco sparse papil-
loso, callo lateraliter compresso, obscure 38-carinato;
columna erecta, auriculis oblongo-linearibus Juxta stigma
ornata; clinandrio suborbiculari, crenulato; ovario pedi-
cellato 8.5 cm. longo.

[ 142 |

Prare NNVIII

O. olivaceum

O. olivaceum

var. giganteum

Observation: In many respects both the plants and flowers belong-
ing to this new species resemble those of O. cucul/atum Lindl. It is
unique in the whole alliance on account of the presence of distinct
ears on each side of the stigma as well as the papillose lip. No doubt
that many more undescribed species and forms are yet to be reported

in this alliance.

Oncidium cucullatum Lindl., Sert. Orch. sub. t. 21,
1858.
Type: Keuador: On the western declivity of Pichin-
cha. Coll. : Jameson s.n./ (IX).

Observation; The plants belonging to this species are rather small-
flowered. Dorsal sepal 12 mm. long, 6 mm. wide; lateral sepals deeply
concave, connate, bifid, 13 mm. long, 8 mm. wide; petals elliptic,
obtuse, 12 mm. long, 8 mm. wide: lip 3-lobed, lateral lobes quad-
rate, midlobe froma narrow isthmus reniform, deeply cordate, bilobed
in front, 20 mm. long, 15 mm. across lateral lobes, 25 mm. across
midlobe. The callus consists of three, keel-like, parallel ridges, the
median ridge is triangular in lateral view, nose-like and longer than
the lateral ridges which divaricate at front.

The callus of the lip in the typical form as well as in the recognized
varieties is always one-third of the length of the entire lip. Most
probably because the type belonged to the Hooker Herbarium and
because Lindley did not make a copy of it for himself, he confused
it with O. sanguinolentum and with others now kept separate under
O. olivaceum and O. mimeticum.

Oncidium cucullatum is essentially Ecuadorian in distribution, but it
may extend into Southern Colombia across the political boundary, up

to Pasto in Depto. de Narino.

Oncidium cucullatum var. dolabratum Stacy,
var, NOV.

Type: Ecuador, without proper locality. Coll. Jame-
son s.n.! (IX).

A var. cucullatum, floribus multo majoribus et lobis
lateralibus labelli oblique triangulari-dolabriformibus
differt.

Dorsal sepal oblong-lanceolate, acute, 20 mm. long,
6 mm. wide: lateral sepals connate, bifid, acute, 22 mm.
long,7 mm. wide; petals ovate-lanceolate, acute, 19 mm.

[ 144 ]

PLATE X XIX

. cucullatum

var. dolabratum

LAG

[ 145 ]

long, 10 mm. wide; lip 8-lobed, lateral lobes obliquely
triangular-dolabriform, acute, 10 mm. long, midlobe
from a narrow isthmus reniform, cordate at base, bilobed
in front; callus an elevated, laterally compressed, nose-
like keel, with a tuft of hairs at base, 10 mm. long; whole
lip 30 mm. long, 22 mm. across lateral lobes, 80 mm.

>

across terminal lobe.

Oncidium cucullatum var. macrochilum Lindl.,

Folia Orch. Oncidium 22, 1855.

Syn.: Oneidium olivaceum var. macrochilum (Lind].)
Sander, Orch. Guide 187, 1901.

Type: Keuador, Quitinian Andes, on trunks of tree,
at 4000 m., Coll. Jameson s.n./ (IXK-L, AMES-GH).

Observation: Dorsal sepal narrowly elliptic, abruptly acute, 18 mm.
long, 8 mm. wide; lateral sepals connate, bifid, deeply navicular, 20
mm. long, 10 mm. wide; petals broadly ovate-elliptic, obtuse, 17
mm. long, 12 mm, wide; lip 3-lobed; lateral lobes obliquely ovate,
rounded, recurved toward midlobe; midlobe from a short isthmus
reniform, cordate at base, bilobed in front; whole lip 28 mm. long,
20 mm. across lateral lobes, 38 mm. across terminal lobe.

Oncidium rhodostictum Krz]. in Pflanzenr. Heft.
80: 2385, 1922.
Type: Specimen without proper locality, cultivated
by Sander s.n./ (W).
Syn.: Oncidium cucullatum var. maculosum Laindl.,
Foha Orch. Oncidium 22, 1855.
Oncidium olivaccum var. maculosum (Laindl.) Sander,
Orch. Guide 188, 1901.
Lectotype: Colombia, near Mariquita, in the forest
of Quindiu. Coll. Linden 1887! (K—L), 7 hoe loco.
Oncidium cucullatum var. purpurascens Hort. in Gar-
den 21: 94, 1882.
Type: Reported only from cultivation; no specimen

is known to exist.

[146 ]

Prater XXX

O. rhodostictum

O. andigenum

LAG

[ 147 ]

Oncidium cucullatum var. Chestertont Hort. in Garden

23: 574, 1888.

Type: Reported only from cultivation; no specimen
is known to exist.

Observation: Sepals 18 mm. long, 5 mm. wide; petals 18 mm.

long, 7 mm. wide; lip 20 mm. long, 10 mm. across lateral lobes, 20
mm. across terminal lobe.

In appearance, this plant is very similar to O. mimeticum in growth
habit as well as in the floral structure, especially the proportions of
the lip. Yet, the two species can easily be separated because of the
configuration of the callus. In O. rhodostictum, the callus of the lip is
always nose-like, triangular in lateral view. In O. mimeticum, the callus
is always a five-parted tubercle which is rectangular in lateral view,

but never nose-like.

Oncidium spathulatum (Lindl.) Stacy, comb. nov.

Basionym: Oneidium cucullatum var. spathulatum
Lindl., Folia Orch. Oncidium 22, 1855.

Syn.: Oneidium nubigenum var. spathulatum (Lindl. )
Rehb.f. in Gard. Chron. 539, 1872.

Oncidium cucullatum subvar. spathulatum (Lindl. )

Veitch, Man. Orch. Pl. pt. 8: 31, 1892.

Type. Ecuador, without proper locality. Coll. Lobb
s.n./ Veitch sale 1847 (IK—L).

Oncidium Phalaenopsis var. excellens Linden in Lind-

enia 12:58, t. 558, 1897.

Lectotype: Ecuador, without precise locality. Im-
ported and flowered by Linden s.n. No specimen is
known to exist. The illustration in Lindenia is chosen
here for the Lectotype.

Observation: Dorsal sepal 16 mm. long, 7 mm. wide; lateral sepals
connate to middle, 17 mm. long, 6.5 mm. wide; petals broadly el-
liptic, obtuse, more or less mucronate, 14 mm. long, 8.5 mm. wide;
lip constricted above middle with open, rounded sinuses, up to 25
mm. long, 18 mm. across lateral lobes, 22-32 mm. across terminal
lobe.

Illustrations: Warner & Williams, Orch. Album 2; t. 96, 1883, as
O. Phalaenopsis; Veitch, Man. Orch. PI. pt. 8: 30, 1892, as O. cucul-
latum var. Phalaenopsis; Amer. Orch. Soc. Bull. 27: 758, 1958, as

[ 148 ]

PLateE XXXI

O. spathulatum

O. Dayanum

[ 149 ]

O. Phalaenopsis, Amer. Orch, Soc. Bull. 42: 1009, 1973, as O.
Phalaenopsis.

Oncidium Phalaenopsis Linden & Rchb.f. in Gard.
Chron. 416, 1869.
Syn. : Oncidium cucullatum var. Phalaenopsis (Linden
& Rehb.f.) Nichols., Hl. Dict. Gard. 2: 485, 1886.
Oncidium olivaceum var. Phalaenopsis (Linden &
Rehb.f.) Sander, Orch. Guide 188, 1901, in syn.
Type: Ecuador, without precise locality. Coll. Wallis
sn! (W).
Observation: Dorsal sepal 18 mm. long, 9 mm. wide; lateral sepals

connate, bifid, 18 mm. long, 7 mm. wide; petals acute, 17 mm. long,
10 mm. wide; lip 33 mm. long, 22 mm. across lateral lobes, 32 mm.
across terminal lobe.

The callus of the lip is very characteristic, obovate in outline, con-
cave with entire raised margins, bidentate at apex, and with a third,
shorter keel in the center originating from the middle of the callus.
There is a good colored drawing from the type plant among Day’s
sketches at Kew, Book 13, page 33, prepared in 1868, a year earlier
than Reichenbach’s description.

Oncidium Dayanum (Rchb.f.) Stacy, comb. nov.

Basionym: Oncidium cucullatum var. Dayanum
Rehb.f. in Gard. Chron. 834, 1871.

Type: Ecuador, without proper locality. Imported
and cultivated by Day, s.n./ (W).

Syn.: Oncidium Phalaenopsis var. Brandtiae Hort. in
Sem. Hortic. 2: 175, 1898.

Type: Reported from cultivation, no specimens known
to exist. The photograph published subsequently under
O. Phalaenopsis in Orchid Rev. 9: 169, 1901, most proba-
bly represents a picture of the type.

Observation: Oncidium Dayanum is much smaller than O. triptery-
gium. Dorsal sepal narrowly elliptic, acute, 13 mm. long, 6 mm. wide;
lateral sepals connate, bifid, 16 mm. long, 7 mm. wide; petals sub-
spathulate, elliptic, acute, 13 mm. long, 7 mm. wide; lip 3-lobed,
wider than long, constricted above the middle, without an isthmus;
whole lip 20 mm. long, 16 mm. across lateral lobes, 24 mm. across

terminal lobe.

| 150 |

Oncidium Dayanum has been imported together with O. Phalaenopsis
by Linden. Excellent color drawings exist at Kew among Day’s draw-
ings, Book 15, page 56, drawn on Febr. 16 and 28, 1871.

Illustrations: Amer. Orch. Soc. Bull. 27: 729, 1958; ibid. 42:
1009, 1973; ibid. 44: 580, 1975, as O. Phalaenopsis; Dodson & Gilles-
pie, The Biology of the Orchids, Frontispiece, 1967 as O. Phalaenopsis.

Oncidium tripterygium Rchb.f. in Bot. Zeit. 10:
694, 1852.
Type: Ecuador, Loja. Coll. Warscewicz s.2./ (W).
Observation: Dorsal sepal 18 mm. long; lateral sepals connate,
bifid, 18 mm. long, 6 mm. wide; petals from a cuneate base elliptic-
lanceolate, acute, 17 mm. long, 8 mm. wide; lip longer than wide,
33 mm. long, 22 mm. across lateral lobes, 30 mm. across terminal lobe.
The type material consists of a single flower in a rather poor state
of preservation. In size it is very close to O. Phalaenopsis, but the
callus of the lip is very different. It is quadrate to oblong in outline,
convex, consisting of three parallel ridges of which the two outer ones
are obscurely lobulate, the third in the center originates from the base
of the lip between the column-wings and as long as the lateral ones.
Illustration: Il]. Hortic. 17: t. 8, 1870, as O. Phalaenopsis.

Oncidium andigenum Linden & Rchb.f. in Gard.

Chron. 416, 1869; ibid. 539, 1872.

Syn.: Oneidium cucullatum var. andigenum (Linden
& Rehb.f.) Hort. in Garden 22: 166, 1882.

Oneidium cucullatum subvar. andigenum (Linden &

Rchb.f.) Veitch, Man. Orch. PI. pt. 8:31, 1892.

Oneidium olivaceum var. andigenum (Linden &

Rehb.f.) Sander, Orch. Guide 184, 1901.

Type: Ecuador, without proper locality. Coll. Wallis
s.n./ (W).

Observation: Dorsal sepal 6 mm. long, 2.5 mm. wide; lateral se-
pals connate, bifid, cymbiform, 7 mm. long, 3 mm. wide; petals elliptic-
oblong, obtuse, 6mm. long, 2.5 mm. wide; lip pandurate, 7 mm. long,
4.5 mm. across lateral lobes, 5 mm. across terminal lobe.

Although references have been made to this species in literature,
I have seen only one collection in addition to the type to which I can
satisfactorily refer here. There is a color drawing with details of flow-

ers of this species in the Kew Herbarium among Day’s drawings,
Book 18: page 54, drawn on October 19, 1874. The corresponding

[151 ]

specimen was sent to Reichenbach by Day, and it is mounted on the
type sheet. Oncidium andigenum is the smallest flowered in the group
and has a five-parted callus.

Oncidium sanguinolentum (Lindl. ) Schltr. in Fedde

Rep. Beih. 6: 98, 1919.

Basionym: Leochilus sanguinolentus Wind. in Bot.
Reg. 80: Misc. p. 91, 1844.

Syn.: Oneidium cucullatum var. sanguinolentum
(Lindl.) Lindl., Folia Orch. Oncidium 22, 1855.

Oncidium olivaceum var. sanguinolentum (Lindl. ) San-

der, Orch. Guide 188, 1901.

Type: Colombia, La Guayra. Coll. Barkers.n./ (KX-L).

Observation: The type specimen consists of two separately mounted
flowers and a colored drawing apparently prepared froma living flower
by Lindley. The flowers are rather small: dorsal sepal 11 mm. long,
5 mm. wide; lateral sepals 13 mm. long, 5.5 mm. wide; petals 11
mm. long, 6 mm. wide; lip 15 mm. long, 10 mm. across base, 14
mm. across the bilobed terminal part.

The reduction of this species by Lindley to O. cucul/atum is difficult
to understand because of the distinctions in growth habit and the
differently constructed callus. Oncidium sanguinolentum flowers from an
undeveloped new growth and has a 5-parted callus: while O. cucul-
latum flowers from mature pseudobulbs and has a nose-like, keeled
callus.

Illustrations: Dunsterville and Garay, Venezuelan Orchids Illustr.
5: 220, 1972, only Fig. B, as O. andigenum. Foldats in Lasser, Flora
Venez. Orch. 15 (5): 360, 1970, as O. nubigenum: Bertha Ospina,
Orquideas de la Clarita, 75, 1972, as O. cucullatum.

Oncidium mimeticum Stacy, sp. nov.

Type: Columbia, Depto. Norte de Santander, Ocana.
Coll. Schlim no. 411! (IKK-L).

Epiphytica, caespitosa, variabili; radicibus flexuosis,
glabris; pseudobulbis approximatis, ovoideis vel pyri-
formibus, vulgo unifoliatis, cataphyllis obtectis; foliis
anguste oblongo-ellipticis vel lanceolato-oblongis, obtu-
sis, basi conduplicatis, quam inflorescentiis vulgo brevior-
ibus; inflorescentiis elongatis, prominenter pedunculatis,

Led

PLATE XA XATI

O, sanguinolentum

LAG

supra laxe paucifloris: floribus vulgo secundis; bracteis
late triangularibus, acutis, pedicellis multo brevioribus ;
sepalo postico elliptico, obtusiusculo, valde concavo,
usque ad 15 mm. longo, 6 mm. lato; sepalis lateralibus
inter se fere usque ad apicem in synsepalo connatis,
cochleato-cymbiformibus, apice divaricatim bidentatis,
usque ad 16 mm. longis, 7 mm. latis; petalis late ellip-
ticis vel obovatis, obtusiusculis, reflexis, usque ad 13 mm.
longis, 8 mm. latis; labello distincte lobulato, lobis later-
alibus auriculatis vel subquadratis, lobo intermedio trans-
verse reniformi, bilobo, antice vulgo sinuato vel promi-
nenter exciso, basi valde cordato, margine undulato, disco
tuberculis 5-nis, cariniformibus elevatis ornatis: toto la-
bello usque ad 25 mm. longo, 30 mm. lato, basin inter
lobos laterales 12 mm. lato; columna humili, 6 mm.
alta; clinandrio cucullato; ovario pedicellato ca. 15-20
mm. longo.

Observation: The rather abundant material of this species present
in the various herbaria has been confused repeatedly with O. cuculla-
tum Lindl. The specimens belonging to this new species can readily
be differentiated by the shape of the callus which consists of 5 tuber-
cles or humps arranged in three rows; 2 humps on each side and a
fifth one in the middle which isshorter than the lateral ones. Although
the structure of the calli is similar to O. sanguinolentum ( Lindl.) Schltr.,
the flowers of the latter species differ, in addition to being much
smaller, in the construction of the lip which is pandurate and the
midlobe of lip is truncate to cuneate at the base.

The original error dates back to Lindley who confused this appar-
ently distinct species with O. cucu//atum in Paxton’s Flower Garden.
It should be noted that the plate in Paxton’s Flower Garden is based
on an abnormal specimen, showing a terminal inflorescence.

Illustrations: Paxton Fl. Gard. 3: t. 87, 1852, as O. cucullatum;
Fl. des Serres 8: t. 853, 1853, as O. cucullatum; ibid. 23: t. 2457,
1880, as O. cucullatum; Veiteh, Man. Orch. Pl., pt. 8: 30, 1892, as
O. cucullatum; Amer. Orch. Soc. Bull. 27: 750, 1958; ibid. 44: 581,
1975, as O. cucullatum; Dunsterville & Garay, Venez. Orch. Ill. 5:
220, 1972, Fig. A only, as O. andigenum; Foldats in Lasser, Flora
Venez. Orch. 15(5): 360, 1970, dotted outline of lip only, as O.

nubigenum.,

[ 154 ]

XXXITI

PLATE

g
=
s
ra
£
E
S

Oncidium mimeticum var. flavidum (B.S. Wms.)

Stacy, comb. nov.

Basionym: Oncidium cucullatum var. flavidum B.S.
Whis., Orch. Grow. Man. ed. 4, 229, 1871.

Syn.: Oncidium cucullatum var. flavidum Linden, Cat-
alogue under nos. 88 to 86, 1870.

Oneidium olivaceum var. flavidum (B.S.Wmms. ) Sander.

Orch. Guide 187, 1901.

Neotype: Since no material is known to exist of the
original introduction by Linden in 1870, Catalogue nos.
33 and 36, the watercolors prepared by Day of the flow-
ers sold under no. 38 and 86, and which are now kept at
Ikew are regarded here as representing the holotype.
Although this unusual color variant was noticed at first
by Linden, the plant became officially described only by
Williams in 1871. The color variant is maintained here
distinct on account of its flowering time which is in De-
cember. The flowering period of O. mimeticum is April—
May.

Illustration: Ill. Hortic. 25: t. 305, 1878, as Oncidium cucullatum.

Oncidium azuayense Krzl. in Pflanzenr. Heft 80:
3821, 1922.
Type: Ecuador, southern slope of the Azuay moun-
tains. Coll. Lehmann no. 587! (G).

Observation: Dorsal sepal broadly elliptic, obtuse to truncate, oc-
casionally somewhat retuse at mucronate apex, dorsally sparsely but
distinetly papillose-muriculate, 9 mm. long, 5 mm. wide; lateral sepals
connate, bilobulate in front, the inside margins of lobes often imbri-
cate, dorsally prominently keeled and sparsely papillose-muriculate,
especially near base, 10 mm. long, 8 mm. wide; petals broadly ellip-
tic, rounded to deeply emarginate at the mucronate apex, dorsally
with a few papillae near the base, 10 mm. long, 7 mm. wide; lip
3-lobed, lateral lobes oblong to subquadrate, midlobe sessile, cordate-
reniform, bilobed in front, 16 mm. long, 14 mm. across lateral lobes,
24 mm. across terminal lobe; callus with a retrorse median tooth.

The plants of this rather commonly occurring species in the Azuay
region can easily be recognized by the dorsally papillose-muriculate

[ 156 |

PLATE XXXIV

O. nubigenum
LAG

sepals and petals and the frequently deeply emarginate petals with
a mucronate tip. The sepals and petals are always fleshy-coriaceous.

Illustrations: Amer. Orch. Soc. Bull. 27: 748, 1958, as O. nubi-
genum; ibid. 27: 757, 1958; Florida Orchidist 10: 14, 1967, as O.

nubigenum.

Oncidium nubigenum Lindl., Gen. and Sp. Orch.
Pl. 197, 1833.
Syn.: Oneidium cucullatum var. nubigenum (Lindl. )
Lindl. Folia Orch. Oncidium 22, 1855.
Oncidium olivaccum var. nubigenum (Lindl.) Sander,
Orch. Guide 188, 1901.
Type: Ecuador, on the ridge of Azuay mountains.
Coll. Jameson .s.2/ (K,K-L).

Observation: Dorsal sepal elliptic, dorsally carinate, 11 mm. long,
6 mm. wide; lateral sepals, connate basally to } of their length, cari-
nate toward apex, 13 mm. long, 5 mm. wide; petals elliptic, obtuse
to abruptly subapiculate, never mucronate, 1] mm. long, 7 mm. wide;
lip 3-lobed, lateral lobes semiovate, without a sinus and imbricating
the reniform midlobe, 17 mm. long, 13 mm. across lateral lobes, 23
mm. across terminal lobe; callus very large in comparison with others
of the immediate relationship, with a retrorse conical point.

The flowers of O. nubigenum can be compared only with those of
O. azuayense, but the two differ from one another in many important
points. The flowers in O. nubigenum are of thin texture, sepals and
petals never papillose dorsally, petals never mucronate, lateral sepals
free almost to base, and the callus of the lip is robust. The flowers in
O. azuayense are rather coriaceous in texture, sepals and petals papil-
lose dorsally, petals mucronate and commonly emarginate, lateral
sepals connate nearly to apex, and the callus of lip is insignificant.

What prompted Lindley to reduce this species to a variety of O.
cucullatum in 1855 will never be understood. But, it surely has confused
the issue ever since. Already in 1867 Reichenbach complained (Gard.
Chron. p. 376) * . when the lamented Dr. Lindley was a young
lynx-eyed observer, he called this plant a good species. When he was
older, and it had come into fashion to combine very heterogeneous
types in ‘one species” because ‘there were too many species’, he be-
lieved it was his duty to cancel some of his own, and this Oncidium
was degraded to the rank of a variety. We believe this was a mistake,
since the many flowers... . never show any cucullate anther-bed,
never have a nasiform keel on the lip, nor is there ever an isthmus
to the lip’.

[ 158 |

Notwithstanding Reichenbach’s clear analysis of the distinctions,
Lindley’s error has been carried on to the present. Veitch in his
Manual of Orchidaceous Plants pt. 8: 30, 1887, suggests that since
O. nubigenum was described first, Lindley should have reversed the
reduction by calling O. cucullatum a variety of O. nubigenum. This
opinion is echoed by Dodson and Frymire in Amer. Orch. Soc. Bull.
27: 750-751, 1958, by calling O. cucullatum and O. Phalaenopsis a
subspecies of O. nubigenum. Fortunately, no legitimate transfers were
made at that time.

Oncidium alticola Stacy, sp. nov.

Type: Ecuador, Prov. Azuay, Nudo de Potate, pass
between headwaters Rio Tarqui and Gir6n, ca. 9,000 ft.
alt. ‘‘Kpiphyte. 5 segments pale to ochraceous-yellow
with purplish tinge; 6th white with purple splotch at
base. Base of column purple, apex white. ‘Bulbs’ (pseu-
dobulbs) called ‘guagra saccha’ and when ground said to
be diuretic; but natives often eat them when thirsty.”
Coll. Camp no. EK-2163! (AMES).

E;piphytica, caespitosa, usque ad 16 em. alta; radici-
bus flexuosis, glabris; rhizomate abbreviato, crasso ; pseu-
dobulbis approximatis, lateraliter compressis, anguste
cylindraceis vel cylindraceo-ovoideis, 2-foliatis, usque ad
4 cm. altis: foliis lineari-oblongis vel oblongo-oblanceo-
latis, acutis, usque ad 11 cm. longis, 1.2 cm. latis; in-
florescentiis, satis gracilibus, remote paucifloris, usque
ad 16cm. longis; bracteis ovato-cucullatis, acutis, 4mm.
longis; sepalo postico obovato-oblanceolato, acuto, 12
mm. longo, 4 mm. lato; sepalis lateralibus inter se usque
ad medium connatis, anguste obovatis, acutis, 12 mm.
longis, 4 mm. latis; petalis ellipticis, obtusis, breviter
apiculatis, 10 mm. longis, 6 mm. latis: labello 4-lobo,
lobis lateralibus subrotundis, inter se 9 mm. latis, lobo
intermedio e cuneata basi flabellato, antice valde bilobo,
lobis subrotundis, disco ima basi callo 8-partito, tumido
ornato, toto labello 19 mm. longo, antice 17 mm. lato;
columna humili; clinandrio tenuiter marginato; ovario
pedicellato usque ad 2 cm. long.

[ 159 ]

Observation: The large lateral lobes of the lip and the cuneate-
flabellate midlobe readily identify the plants belonging to this new
species. It appears to be rather common in the Azuay region.

The following additional specimens belong here; Ecuador: Azuay-
Tarqui. Coll. Harling & Andersson no. 18205 bis! (AMES, S); Cuena,
Coll. Strobel s.n./ (AMES); without exact locality. Coll. Rodrigo
Escobar-R. no. 1380! (AMES).

Illustrations: Amer. Orch. Soc. Bull. 32: 905, 1963, as O. nubi-

genum.,

Oncidium tarquiense Stacy, sp. nov.

Type: Ecuador, Province Azuay, Portete del Tarqui,
Cuenca-Girén, 2700-2950 m. altitude. Coll. Harling &
Andersson no. 18205! (S).

Epiphytica erecta, usque ad 30 cm. alta; radicibus
flexuosis, glabris, pseudobulbis ascendentibus ovoideis
vel pyriformibus, 2-foliatis, basi vaginis chartaceis ob-
tectis, usque ad 5 cm. longis, 1.5 cm. latis: foliis satis
tenuibus, lineari-oblanceolatis, obtusis, usque ad 10 cm.
longis, 1.5 cm. latis: inflorescentiis gracilibus, paululo
arcuatis, laxe secundifloris: racemo laxifloro: bracteis
infundibuliformibus, oblique ovato-triangularibus, 5mm.
longis; sepalo postico elliptico, obtuso vel subacuto, 16
mm. longo, 6 mm. lato; sepalis lateralibus quarta parte
basali inter se connatis, supra liberis, anguste ellipticis,
acutis, 17 mm. longis, 5 mm. latis; petalis late ellipti-
cis, acutiusculis, 14 mm. longis, 8 mm. latis; labello
3-lobo, lobis lateralibus oblique ovatis, lobo intermedio
maximo, cuneato-flabellato, antice bilobulato, lobulis
rotundatis; disco basin callo minuto, subquadrato, sur-
sum curvato ornato; toto labello 24 mm. longo, 28 mm.
lato; columna humili; clinandrio marginato; ovario
pedicellato usque ad 85 mm. long.

Observation: The flowers in general appearance are larger than
those found in O. a/ticola. The transversely rectangular-oblong callus

which is distinetly curved upwards is unique in the whole alliance,
hence affords an excellent field character for identification.

[ 160 }

PLtatreE XXXV

O. alticola

O. tarquiense

LAG

Oncidium aequinoctiale Stacy, sp. nov.

Type: Ecuador, without precise locality. Coll. André
s.n.f (K).

I. piphytica, usque ad 80cm. alta; radicibus flexuosis,
glabris; pseudobulbis approximatis, anguste ovoideis,
unifoliatis, usque ad 5 cm. altis; foliis lineari-lanceolatis,
acutis vel subacuminatis, usque ad 17 cm. longis, 1.5
cm. latis; inflorescentils erectis, supra racemosis, laxe
paucifloris, usque ad 80 cm. longis; floribus satis tenui-
bus, secundis; bracteis cucullatis, acutis, ovariis pedicel-
latis multoties brevioribus, 4mm. longis; sepalo postico
valde concavo, elliptico, obtuso, 11 mm. longo, 6 mm.
lato: sepalis lateralibus ad tertiam partem apicalem inter
se connatis, valde concavis, obtusis, usque ad 12 mm.
longis, 6 mm. latis; petalis obovatis, obtusis, inconspicue
apiculatis, 11 mm. longis, 7 mm. latis; labello 3-lobo,
lobis lateralibus divergentibus, ovato-oblongis, lobo in-
termedio sessili, paululo cuneato, transverse elliptico,
antice bilobo, utrinque sinuoso; callo satis prominenti,
3-partito, partitione intermedia retrorse conica; toto la-
bello 18 mm. longo, basin 14 mm. lato, antice 24 mm.
lato: columna humili, clinandrio marginato.

Observation: In floral structures the plants of this new species re-
semble those of O. nubigenum. It differs from O, nubigenum especially in
having free lateral lobes forming open, acute to acuminate sinuses with
the midlobe of the lip; the midlobe is cuneate not cordate at the base.

Onsidium erosilabium Stacy, sp. nov.

Type: Ecuador, without precise locality. Cultivated
by Rodrigo Escobar-R. no. 1879! (AMES).

K.;piphytica erecta, usque ad 30 cm. alta; radicibus
flexuosis, glabris; pseudobulbis ovoideis, lateraliter com-
pressis, usque ad 6 ecm. altis; foliis linearibus, acutis,
18-25 cm. longis; inflorescentiis gracilibus, paululo arcu-
atis, paucifloris, usque ad 10 cm. longis; floribus albidis,
sepalis petalisque pallide brunneo maculatis striatisque,

[ 162 |

PiatE XXXVI

O. aequinoctiale

QO. erosilabium

LAG

[ 163 ]

labello basin purpureo maculato, callo aurantiaco; sepalo
postico elliptico, obtuso, 11 mm. longo, 6 mm. lato;
sepalis lateralibus tertia parte basilari inter se connatis,
oblongo-ellipticis, acutis, 14 mm. longis, 5 mm. latis;:
petalis late ellipticis, obtusis, 11 mm. longis, 6 mm.
latis: labello e cuneata basi 8-lobo, lobis lateralibus sub-
quadratis, truncatis, lobo intermedio sessill, transverse
reniformi, antice rectangulariter late eroso: disco venis
incrassatis ornato; callo cylindrico, antice 8-partito; toto
labello 18 mm. longo, 21 mm. lato; columna erecta;
clinandrio marginato; ovario pedicellato usque ad 24
min. longo.

Observation: In floral structure the plants of this species are simi-
lar to O. chimborazoénse, but the barred petals and the squarely erose
lip differentiates it immediately not only from that species but from
every member of the whole alliance. I have seen another specimen
referable to this species in the Kew Herbarium; it was cultivated
during the turn of the century under the name of O. cucul/atum var.

Oncidium chimborazoénse Stacy, sp. nov.

Type: Ecuador, Chimborazo, 3200-8400 m. altitude.
Coll. F.C. Lehmann no. 215! (AMES, Kk).

K.;piphytica, ascendenti, usque ad 25 cm. alta; radici-
bus flexuosis glabris ; pseudobulbis approximatis, oblique
ovoideis, unifoliatis, usque ad 25 mm. altis: foliis per-
gameneis, anguste lanceolato-ellipticis, subacutis vel ob-
tusis, usque ad 18 cm. longis, 1 cm. latis; inflorescentiis
erectis vel paululo arcuatis, supra laxe paucifloris; flori-
bus ut videtur secundis; bracteis infundibuliformibus,
ovato-triangularibus, acuminatis, 5 mm. longis; sepalo
postico obovato-elliptico vel elliptico-oblanceolato, acu-
to, vel abrupte subacuminato, usque ad 12 mm. longo, 6
mim. lato; sepalis lateralibus basin inter se breviter con-
natis, anguste ellipticis, acutis vel subacuminatis, usque
ad 18 mm. longis, 5 mm. latis; petalis late ellipticis,
obtusis, usque ad 11 mm. longis, 6 mm. latis; labello e
cuneata basi 8-lobo, lobis lateralibus obliquis, apice ro-

[ 164 |

PLATE AXXVIIT

O. chimborazoénse’

O. tunguraguense

LAG

tundatis; lobo intermedio maximo, transverse reniform],
antice valde bilobulato, basi subecordato, disco basin callo
tumido, 3-partito ornato, toto labello usque ad 23 mm.
longo, 28 mm, lato; columna humili, crassa, clinandrio
marginato: ovario pedicellato usque ad 25 mm. longo.

Observation: Oncidium chimborazoénse is larger than O. tunguragu-
ense, especially in the size of the flowers. It differs from the latter in
the plants having only one-leaved pseudobulbs and the lip is much
wider than long.

There is another specimen at Kew collected also in Chimborazo by
Cross no. 52! (K), who sent it to the Veitch establishment in 1882.

Illustration: Bot. Mag. 94: t. 5708, 1868, as O. cucullatum var.
nubigenum; Belg. Hortic. 19: t. 19, 1869, as O. nubigenum; Dict. Icon.
Orch. Oneid. t. 30, 1900, as O. nubigenum; Amer. Orch. Soc. Bull.
41: 1001, 1972, as O. nubigenum.

Oncidium tunguraguense Stacy, sp. mov.

Type: Keuador, Vulcan Tunguragua, 8000-8400 m.
alt. Aug.-Sept. Coll. F.C. Lehmann no. 8572! (Ix).

Epiphytica, erecta, usque ad 25 cm. alta: radicibus
flexuosis, glabris; pseudobulbis approximatis, pyriform-
ibus, lateraliter compressis, bifoliatis, usque ad 4 cm.
altis; foliis ellipticis vel lanceolato-ellipticis, acutis, usque
ad 13 cm. longis, 2.5 cm. latis; inflorescentils erectis,
sursum arcuatis, laxifloris, usque ad 25 cm. longis; flori-
bus ut videtur secundis; sepalo postico oblanceolato vel
obovato-oblanceolato, acuto vel obtusiusculo, usque ad
10 mm. long, 4.5 mm. lato; sepalis lateralibus supra
basin inter se breviter connatis, lineari-oblongis, acutis
vel obtusiusculis, 11 mm. longis, 8 mm. latis; petalis
ellipticis obtusis vel abrupte acutis, usque ad 9 mm.
longis, 5 mm. latis: labello in ambitu trapezoideo, 3-
lobo, lobis lateralibus e cuneata basi triangulis, lobo inter-
medio transverse reniformi, antice retuso vel subbilobulo :
disco venis incrassatis ornato; callo tumido, 8-partito:
toto labello usque ad 15 mm. longo, 16mm. lato; colum-
na humili; clinandrio marginato; ovario pedicellato usque

ad 2 cm. longo.
| 166 |

Observation: From the closely related O. chimborazoénse this new
species is readily distinguishable by the 2-leaved pseudobulbs and the
rhombic lip which is approximately as long as wide. The sepals and
petals are maroon inside, much paler and suffused outside; the lip is
white with heavy crimson blotches on the basal half especially on the
lateral lobes; the callus is orange.

ACKNOWLEDGMENTS

The steady cooperation of my colleagues at the Orchid
Herbarium of Oakes Ames, and especially the guidance
of Dr. Leslie A. Garay, have been a continuous inspira-
tion throughout the preparation of thisstudy. Their help
is deeply appreciated. My special thanks to Dr. Peter
Taylor, the Orchid Herbarium, Royal Botanic Gardens
and to Dr. Luciano Bernardi, Conservatoire et Jardin
Botaniques, Geneve, for their kind cooperation in pro-
viding specimens and drawings of types. For permission
to reproduce their color slides so essential to this study
I am indebted to Mr. Gilberto Escobar-R, Mr. Rodrigo
Escobar-R, Dr. George Kennedy and Mrs. Henry T.
Northen. The publication of color plates was made possi-
ble through the generosity of Mr. Maurice Tl’. Freeman,
Mr. Walter Hunnewell, Mr. and Mrs. Lee B. Kuhn,
the Lear Siegler Foundation, Mrs. Henry 'T. Northen,
Mr. Robert M. Scully, Sr., Mr. Richard Seifert and
Mr. Robert G. Stone. Their thoughtfulness and help
are sincerely acknowledged. ‘To Miss Gertrude Ahern
I am grateful for the dedicated attention given by her
to this paper.

[ 167 ]

PLATES

[169 |

Top left: Oncidium cucullatum. Photograph: R. Northen.

Top right: Oncidium cucullatum. Photograph: G. Escobar-R.
Bottom left: Oncidium olivaceum. Photograph: G. Escobar-R.
Bottom right: Oncidium rhodostictum. Photograph: G. Kennedy.

[170]

PLATE XXXVIIL.

Top left and right: Oncidium Kennedyi. Photographs: G. Kennedy.
Bottom: Floral details of Oncidium Kennedyi.

XXXIX.

PLATE

Top: Oncidium mimeticum. Photograph: R. Northen.
Bottom: Oncidium mimeticum. Photograph: G. Kennedy.

[174]

PLATE XL.

Top left: Oncidium mimeticum. Photograph: G. Kennedy.

Top right: Oncidium sanguinolentum. Photograph: G. Escobar-R.
Bottom left: Oncidium sanguinolentum, Photograph: R. Escobar-R.
Bottom right: Oncidium sanguinolentum. Photograph: G. Escobar-R.

[176 |

PLATE XLI.

Top left and right: Oncidium mimeticum. Photographs: G. Kennedy.
Bottom: Oncidium tripterygium. Photograph: R. Northen.

[178]

PLATE XLII.

[179]

Top: Oncidium Phalaenopsis. Photograph: R. Northen.
Bottom: Oncidium Dayanum. Photograph: R. Northen.

[180]

XLII.

PLATE

[181]

Top left and right: Oncidium Dayanum. Photographs: G. Kennedy.
Bottom left and right: Oncidium spathulatum. Photographs: G. Kennedy.

[182]

PLATE XLIV.

[183]

Top: Oncidium tarquiense. Photograph: J. Stacy.
Bottom: Oncidium alticola. Photograph: R. Escobar-R.

[ 184}

PLATE XLV.

[185 |

Top and bottom: Oncidium nubigenum. All photographs by G. Kennedy.

[186 |

PLATE XLVI.

[187]

Top left: Oncidium alticola. Photograph: G. Escobar-R.
Top right: Oncidium erosilabium, Photograph: R. Escobar-R.
Bottom: Oncidium chimborazoénse. Photograph: G. Kennedy.

[188]

PLATE XLVIL.

[189]

Top: Oncidium aequinoctiale. Photograph: R. Northen.
Bottom: Oncidium tunguraguense. Photograph: J. Stacy.

[190]

PLATE XLVIII.

[191]

APR 19 197

CAMBRIDGE, MassAacHusETTs, APRIL 9, 1976 VoL. 24, No. 8

PLANTAE COLOMBIANAE XIX
E PARTIBUS AMAZONICIS WITOTORUM PLANTAE
FRUCTUARIAE SATIVAE NOVAE
BY
RicHarp Evans SCHULTES

In studiis ethnobotanicis meis de plantis alimentis in-
diorum witotorum e Amazonia colombianae duas arbores
novas inveni: unam e genere leguminosarum, Inga; al-
teram e genere apocynacearum, Macoubea. In regionibus
Rio Karaparand et Rio Igaraparana utraeque arbores ab
indigenis coluntur.

Doctori Thomas S. Elias pro descriptione speciei In-
gae novae gratias ago. Similiterque erga Patrem Miguel
Junyent, Matrem Holga Edith Sajona et Sororem Ligia
Maria Ochoa, propagatores doctrinae christianae catholi-
cae in colonia San Rafael apud Rio Karaparana inferio-
rem, pro beneficentiis suis, praecipue mihi flores ex
arbore typica praebuisse, gratus sum.

APOCYNACEAE

Macoubea witotorum PR. FE. Schultes spec. nov.

Arbore usque ad quadraginta ped. alta, dense foliosa,
ex radice singulare truncis multis, in terranon inundabile
culta; cortice laeve, crasso, in maculis albido-griseo,
abundanter albo-lactifere. Folia valde chartacea, supra
atroviridia, infra pallide viridia, late elliptica, base ae-
qualia, apice obtusa vel rarenter subacuta, lamina 20-24

[ 193 ]
ECONOMIC BOTANY LIBRARY

OF OAKES AMES
HARVARD BOTANICAL MUSEUM

PuatTE XLIX

(fi Yi,

LA

\Ss

MACOUBEA t)
witotorum #.€. Schu/tes ki JBC

1, leafy branch, approximately } natural size. 2, inflorescence, ap-
proximately 5 natural size. 3, flower, approximately 14 natural size.
t, basal portion of flower dissected, approximately twice natural size.
5, stamen, approximately 4 times natural size. 6, corolla lobe, ap-
proximately 4/5 natural size. Drawn by Josuua B, Crank

PuatE L

MACOUBEA
witotorum
ae R.E. Schultes

B, dissected fruit, showing

A, fruit, approximately s natural size.
9 ’ pp . ri
placement of seeds, approximately 4 natural size. C, seed, approxi-

mately natural size, with detail of surface greatly enlarged.
Drawn by Josuvua B. Cirark

em. longa, 11-15 em. lata, utrinque glabra, nervis utrinque
prominentibus; petioli 1-2.5 cm. longi. Inflorescentiae
corymbosae, multiramosae, usque ad trigintiflorae, bracte-
atae, minutissime griseo-puberulae, axibus principalibus
robustis, usque ad 8 cm. longis, pedicellis brevibus, pler-
umque I cm. vel brevioribus, minute bracteolatis. Flores
albido flavi vel eburnei, non fragranti, plus minusve 25
mm. longi. Calyx persistens; corollae lobis vix contor-
tis, base glandulosis, ovatis, 15 mm. longis et basin 10
mm. latis: corollae tubo quam lobis multo breviori, in
fauce albo-piloso; antheris sessilibus, sagittatis, usque ad
4 mm. longis; ovario dense albo-piloso, stylo 0.8-1.2
mm. longo annulo crasso atrobrunneo coronato et crista
quinquelobata capitata in acumine 0.8 mm. producto.
Apocarpium magnum, usque ad 16 cm. longum, 11 ¢m.
<9 cm. in diametro, irregulariter reniforme vel stomach-
iforme, petioli basin versus submammiforme, extus in
maturitate brunneum vel subaureo-brunneum, sordide
granuloso-squammulosum, intus viridulum praeter sem-
inarum lectum atrobrunneum, usque ad 15 mm. crassum,
pulpa brunneola liquida dulce ; seminibus plurimis, rubro-
brunneis, curvatim elliptico-oblongis, superficie dense et
grosse vermiculato-foveolatis, inter foveas minutissime
eristatis, cristis magnopere minute indumento § albo-
spongioso indutis, 16 mm. longis, 6 mm. latis; pedunculo
lignoso, 10 mm. crasso, usque ad 13 cm. longo, calyce
valde persistente.

Cotompra: Comisaria del Amazonas, Rio Karaparana, San Rafael
(in vicinity of El Encanto), near confluence with Rio Putumayo. “Tree
up to 40 ft. Many trunks from root. Cultivated. Bark smooth, light
grey with whitish patches. Latex copious, white, sticky. Leaves
thick, coriaceous. Outside of fruit brown; inside light green except
where seeds lie (which is brown). Pulp liquid, light brown, sweet,
eaten. Seeds dark brown. Seeds and pulp from one fruit 250 ce.
Witoto name: 00-rroo’-soo-je (in Rio Karaparana), o0-koo-je (in Rio
Igaraparana)’’. March 80-April 2, 1970. Richard Evans Schultes 26072
(Tyveus in Herb. Gray: Typr Dupticati in Keon. Herb. Oakes Ames ;

| 196 |

PiLaTtTeE LI

Fruit of tree from which the type material of Macoubea witotorum
was taken. Photograph: R.E. ScHuites

[ 197 |

Herb. Nac. Colomb.).—Same locality. ‘““Flowers yellow-cream or
whitish yellow, without fragrance. Taken from type tree’’. August
1970. P. Miguel Junyent sine num. (Fioris Typrin Herb. Gray ; Econ.
Herb. Oakes Ames: Herb. Nac. Colomb. ).

99
.

Genus Macoubea, ut hodie comprehenditur, duas spe-
cies includit: MM. guianensis Aubl., quae totam terram
regionem guianarum atque Brasiliae amazonicae inhabi-
tat, sed varietas eius, MW. guianensis var. pubiflora Mona-
chino, non nisi ex loco typi prope civitatem Iquitos, e
regione Amazoniae peruvianae observatur: Macoubea
Spruce’ (Muell.-Arg.) Markgraf, cuius typus a Richard
Spruce apud Rio Uaupés juxta Brasiliae Colombiaeque
terminum collectus est, tantum in regione Rio Negro
Brasiliae amazonicae afferitur; distributio varietatis eius,
Macoubea Sprucei var. paucifiora (Spr. ex Muell.-Arg. )
Monachino, cum specie congruit.

Secundum Monachino (Lloydia 8 (1945) 296), ‘‘spe-
cies Macoubearum inter sese arcte affines. Differentiae
primae quae in elementis variis discernendae possint in
florium proportionibus magnitudinibusque fundantur,
sed nequaquam constantes sunt’.

Macoubea witotorum a speciebus adhoc cognotis non
solum florium characteres proportionesque differt, sed
etiam a JM. guianense fructus formae proportionesque
valde dissimilis est.

Praeterea haec species nova ab utraque specie florium
characteribus minoribus distingui potest. Macoubea wito-
torum corollae lobi 15 mm. longi atque 10 mm. lati
metiuntur, dum in MM. guianense 3-4 mm. et in M.
mim. lati sunt; corollae tubus quam lobus

Sprucet 5-7
multo brevior qui in JZ. guianense quam lobus dimidium
vel longior, et in JZ. Sprucet prope eadem aequilongus
est. Macoubea witotorum a M. guianense fructus forma
proportioneque facile distinguitur: in priore mericarpio
immaniter reniforme vel stomachiforme est, 16 ¢m. longo

[ 198 |

Piarye E11

a)
Macoubea C PUY ANNENELE

Illustration of Macoubea guianensis, published in Aublet: Plantes
de la Guiane Francaise 2 (1775) t: S78.

[ 199 ]

et 1] cm. X 9 cm. in diametro metitur; in posteriore,
globoso, circiter 6-8 cm. in diametro metitur.

Secundum exiguos libros, Macoubea arbores sunt sil-
vicola et semper sua sponte crescunt et silvas in terra
humida sed supra inundationem annuam_ preoptans.
Macoubea witotorum est arbor quae ut videtur solum in
cultura exsistit et semper in terra alta, bene siccata, nun-
quam inundata prosperat.

Indigenae tribus witotorum ut arbores solum in regione
Rios Karaparana et I garaparana coleri affirmant,et verum
est ame nunquam ipsis in multis meis annis explorationis
in regionibus amazonicis invenitis per quos annos cogita-
tiones meas plerumque versus plantas cultas dirigi. Si
naturaliter occurrit, ea est rara.

Indigenae fructus parte terminale acutiore praecidunt
vel in eo foramen faciunt et copiosam pulpam fulvam
ducunt quae in forma liquidi dulcis adest. Infeliciter,
studia nutrimenta huius liquidi non facta sunt, sed sine
dubio sacchari abundantia saturata est. Semina non
devoruntur.

Arbores mensibus martio aprileque fructus abundanter
ferunt. Unusquisque exempla quae in cultura vidi mul-
tas (usque ad quinque vel sex) truncos corpulentos ex
radice unico surgentes habuerunt. Utrum hoe proprium
incrementum mutilationibus trunci primari an ab indi-
genis eventus est non potuit, sed indigenae ipsae arborem
normaliter hoe modo crescere affirmant et se nullo modo
arborem ad conditionem faciendam truncare vel secare.
Certe, hae incrementi consuetudo mensuram = arbore
coronae Magnopere auget atque, propter hoc, abundan-
tem segetem fructuum.

Doctor C. Sastre (Museum d°‘Histoire Naturelle,
Paris) specimina florescentia quae ut videtur esse elus-
dem Macoubeae species in pago La Chorrera prope Rio

Igaraparana colligit.

[ 200 ]

Pyare LE!

DIAGONAL SHADING
ON INSET INDICATES
AREA SHOWN
IN DETAIL BELOW

lauitos f A
R 1) LETICIA ! 5

: XN
“Si = 4 ~ rd au
a Ao (CIARG :

-
Ya RETIRG ENTRE Rios Pa ieee

&

Re

i ; c
5 ( ws)
(@) tt aie
\ - y
KARAPARANA-IGARAPARANA =, L 6 ~~
REGION US \ | Om iin
of the ( ;
COLOMBIAN AMAZON © .

) t
AN fo * We AI) af if VA
wa (c 4 p ae OO

CotomBra: Comisaria del Amazonas, Rio Igaraparana, La Chorrera.
**Arbre, 10 m., latex blane abondant, petales jaunes, fruits comest-
ibles, pulpe liquide avee graines noires. Nom witoto: ukurai.’’ Sep-
tember 29, 1973. C. Sastre 2340.

LEGUMINOSAE

Inga chorrerana Elias spec. nov.

Arbores parvae; ramis teretibus, sparse ferrugineo-
tomentellis, lenticellis conspicuis. Ramuli ferrugineo-
tomentosi. Folia petiolis 4-5 cm. longis, dense ferrugineo-
tomentosis, partim inter Juga superiora foliolorum dila-
tatis; ala late ellipticis vel obovatis, 6-12 mm. latis.
Foliola 6-jugata, petiolulis 2-8 mm. longis, dense fer-
rugineo-tomentosis; lamina elliptica, apice apiculata,
basi rotundata, 5.5-15 em. longa, 2.8—5.5 cm. lata, sub-
coriacea, nervis lateralibus 16-20, subtus conspicuis;
glandulae rhachidis 38-8.5 mm. diam., orbiculares; stip-
ules non visae. Inflorescentiae multiflorae, spiciformes,
3-5 cm. longae, terminales aut in axillis superioribus
dispositae; pedunculis 1-2 cm. longis, dense ferrugineo-
tomentosis; bracteis late ovalis, apice acuminatis vel sub-
apiculatis, 5-6 mm. longis, dense ferrugineo-tomentosis ;
calyx cylindricus, sessilis, 7-9 mm. longus, sparse tomen-
tosus, lobis ovatis 1-1.5 mm. longis; corolla cylindrica,
apice vix dilatata, 1.7-1.9 em. longa, dense villosa, lobis
ovatis 2.5-8 mm. longis; stamina tubo vix exserto; fila-
menta alba, ad 8.5 mm. longa: antherae dorsifixae ;
pistillum sessile, teres, ca. 2 mm. longum, glabrum.
Legumen coriaceum, in maturitate areis ciliorum diffu-
sis ferrugineo-villosum, plano-compressum, ad 45 cm.
longum, 5-7 cm. latum, ca. 1 cm. crassum, marginibus
vix elevatis. Semina sine testa, cotyledonibus 8-8.2 cm.
longis, 1 em. latis.

CoLtombra: Comisaria del Amazonas, Rio Igaraparana, La Chorrera.
Small tree. Pulp of fruits edible. Cultivated. Fruit 1-14 ft., long,

slender, thickness of index finger. Flowers white, fragrant. June 6,
1942. Richard Evans Schultes 3896. (Tyrus in Herb. Gray).

( 202 |

PLatE LIV

Habit drawing, approximately | natural size. Pod, approximately
s natural size. Inflorescence, approximately natural size.

Drawn by Irene Brapvy

[ 203 ]

Hance speciem arborem fructiferem ob siliquaram mag-
narum pulpa albam dulcem semina cingens coleri ferunt.

Doctor Elias seripsit: ‘“‘Jnga chorrerana in Sectione
Inga, Series Spectabiles propter proprietates florales
fructuosas debet. Ab aliis speciebus affinibus folio sex-
jugati, folium apice apiculato atque fructu pubescente
distingueri potest. Ad Ingam spectabilem Willd. propin-
qua atque fortissime proxima ad J. Go/dmannu Pittier
ut videtur est.”

[ 204 ]

PACHYTENE CHROMOSOME MORPHOLOGY
AND ITS BEARING ON INTERSPECIFIC
AND INTERGENERIC RELATIONSHIPS

OF COIX*
BY
J. VENKATESWARLU', R.S.K. CoaGantr,
AND PanuGant! N. Rao’

The genus Coix is a member of the tribe Maydeae
(family Gramineae) which comprises seven genera. ‘Two
of these, Zea and Tripsacum, are native to the New
World, while the rest, Coix, Chionachne, Trilobachne,
Sclerachne and Polytoca, are native to southeast Asia
(Mangelsdorf and Reeves, 1939). The Maydeae is a taxo-
nomic rather than a natural assemblage of genera (Cha-
ganti, 1965), with monoecism as the feature in common to
both the New World and Asiatic members. However,
from time to time, the existence of possible phylogenetic
relationships between the two geographically separated
groups has been postulated (Anderson, 1945; W.C.
Galinat, personal communication).

Coix is the largest of the five maydeaceous genera na-
tive to Asia. Its systematics is confusing (Venkateswarlu

* Investigation supported by a grant from the Indian Council of
Agricultural Research, New Delhi.

‘Department of Botany, Andhra University, Waltair, Andhra
Pradesh, India.

*The New York Blood Center, 310East 67th Street, New York,
N.Y. 10021.

[ 205 |

and Chaganti, 1973). At least three highly polymorphic
species assemblages are recognized, namely C. aquatica
Roxb., €. Lachryma-Jobi L., and C. gigantea Iwoen.
C. aquatica and C. Lachryma-Joli have chromosome
numbers of 2n=10 and 20 respectively, while in the case
of C. gigantea populations with chromosome numbers
of 2n=20 and 40 have been reported (Venkateswarlu
and Chaganti, 1978).

Comparative chromosome morphology, especially at
the pachytene stage of meiosis, where the chromosomes
are more extended than at mitotic metaphase and often
present characteristic features, is a useful parameter in
understanding interspecific and intergeneric relation-
ships. In the case of Coix, such studies have been few
(Venkateswarlu and Chaganti, 1973). We have studied
and report here the pachytene chromosome morphology
in populations of Coix representing the three species.

Materials and Methods

Samples of seeds from populations representing the
three species were assembled from different regions of
India. Some were collected in the field by the authors,
while others were collected for them by colleagues in
different parts of the country. In addition, two samples
from Japan and one trom Brazil (countries in which Coix
is not native, but into which it has been introduced for
use as a fodder, cereal or ornamental) were also obtained.
The geographic origin and method of collection of all the
seed material is listed in Table |. Plants were raised from
these seeds in the experimental gardens of the Depart-
ment of Botany of the Andhra University. Young in-
florescences were fixed in 8:1 ethanol-acetic acid. After
24-48 hours of fixation, the fixative was replaced by 70%
ethanol. The material was stored in this fluid at 4° C,
until used for study.

r

[ 206 |

Acetocarmine squash preparations were made of young
anthers. Appropriate nuclei were photographed, and
camera lucida drawings were made of them from tempo-
rary preparations. Chromosomes were measured from
camera lucida drawings alone, and their lengths, which
include the centromeric regions, are expressed in micron
units. Numbers are assigned to chromosomes on the
basis of decreasing order of length: thus chromosome 1
is the longest of the complement. Individual chromo-
somes are identified on the basis of their length, arm-
ratio (length of long arm/length of short arm), and the
pattern of distribution of heteropyenotic and eupycnotic
region.

Observations

Coixv aquatica Roxb. The chromosomes of this species
are difficult to study at pachytene, because they are long
and often intertwined. In addition, the chromosomes
possess long heteropyenotic segments which often ex-
hibit non-homologous association (Plate LV), thereby
making it difficult to follow the chromosome from one
end to the other. Deep staining heteropycnotic regions
are present in the proximal regions of both arms. Addi-
tional heteropyenotic regions are present at other loca-
tions on the chromosome arms. ‘The chromosome arms
terminate in pronounced and dark staining chromomeres.
Exact measurements of the heteropycnotic regions were
not possible to obtain, because of variability in their ex-
tent (see Venkateswarlu and Chaganti, 1973). Camera
lucida drawings of the five chromosomes are represented
in Plate LVI, while the idiogram represented in Plate
ILX is based on mean values of length and arm ratio of
five measurements for each chromosome (see Table II).
The third longest chromosome of the complement has
the nucleolus organizer in the long arm in a sub-terminal
position (Plate LX).

‘uMOUYUN 3d1NOS AleUul

“lig ‘suapiey d1ue}0g A}ISI9AIU Ls)

dy} UL Sain}[Nd wWoIs Saadeosa se

pajiejs aaey pinoy ‘sndulesy AjisIaa

-1UQ) BIYPUY 9} UO PIM BSulMmois

_ sioyjyne 9y} Aq paqdaTjoo ~—s sjuryd Jo Auojoo ¥ ‘pyiMy sndwes

Sl-o a89[[0D [einqpnousy ‘jedioutsg :Aq patjddns BIpU] ‘d10]eBquIIOZ

(s]eyr) UIo}sey) eIpUy
Je[nsuluag Ula}seayynos jo asuel

3-5 (OMSY) stoyine vyj Jo suo Aq pa}da][OD UTeJUNOU! [B}SeOD ay} Ul LuIsejJUeUY 7
LD umou yun uMmouyuUN 4
Bipuy jo
Gp-- AVAING [BOIURIOg ‘jsIuRJOg [eUOIZoy :Aq parddns BIpuy ‘Wessy ry
eIpuy jo ‘WYRM VIIA] “eA
BP--Q AvAInS [eoIuUe}Og ‘JsIuRJOg [eUuOIsoy :Aq patiddns eIpu] ‘Wessy 1QO[-DUAALYIDT “D
8 WO (JMSY) Sssoyine ayi Jo auo Aq pa}daq[oo BIpu] ‘ysepeid e1ypuy “
uonerodi07
te ¥)D Spaas jeuONeN “WweeYyd “A°D “Aq :Aq parddns BIpu] ‘essliC “
g-Y'D AllsiaalugQ 1esnes ‘apueg “YS ‘Jorg :Aq patddns BIpU] ‘ySepetg eAyupRly pojonbv “>D
Jequiny uonisinboy jo apoyy UISLIQ sIydeIs0a5) xx S3100dS
uoyeindog

Io UOISSIDDY

, UONVSNSIAUI SIY} UI pasn ‘UZ0Y VazuDTIG ‘Dy pue ‘"] 1gof-DuAAYIDT “DQ “GxOY vIIDNbHY “D “ZILA ‘X10d JO

saidads 391Y} 9Y4] 0} Sulsuojeq suonjejndod snorea ay} JO [el1a}eW peas Jo uO}ISInboe Jo spoul pue UIsIIO D1ydeuso0ar)

I ATaAVL

[ 208 |

‘untieqiaH

AJISIOAIUL) RIYPUY JY} Ul paytsodap uveq aARY 9A0gP pajsl] SatjalieA puke satdads ay} [Je Jo SUsUIOAdS UINTIeqIOY x»
‘sojdures paas yim sn parddns oy ¢ uuIN]OO UI pajsl] ajdoad snorieA ay} 0} [NJa}e13 Iv OM y

eIpuy jo

I°9D = Aaarng [eotuejog ‘ory *S eI]OY “Iq :Aq partddns
99D (QYNd) szeyyne ay} Jo auo Aq pajoaT]oo
ojnedg ors

TD ‘orlewy ‘1S ‘lasneyyeeyos “AY “Aq :Aq paddns
erpuy jo

Ga) Avains |eo1uej}Og ‘ory “Ss BI[OY “Iq :Aq parjddns

oJOAY ‘AVISIOAIUL) OAYIES

‘ainj[nosy jo Jusuijiedaq ‘A10}e10qe’] Bul

€-d -peeig pue souaINs dolla ‘epeiey iq :Aq parjddns
Bipuy jo

2-2 AVAING [edIURIOg ‘jsIURJOg [eUOISay :Aq patjddns
ojoAY ‘AJISIOAIUS) OAYIES

‘gin}[Nosy jo Juawjiedaq ‘Ai1ojeloqe’y] Bul

Il-D -paeig pue aouatog doi ‘eperey ‘iq :Aq patjddns
eipuy jo

qOI-D AeAINS [edIuRIJOg ‘JsTUR}Og [BUOIsay :Aq patjddns

erpuy jo
ROT-D AVAING [edIURIOg ‘JsIuRlOg [eUOIsay :Aq patddns

(s}eys) U191S9M )
eIpu] JO Iejnsuluag U13}sOM

-yjnos jo asuel ulejUNOW j[e}seo7 DIJUDSIS ‘dD

“p91g]

pue 4afij1Mow “SIA

U99M}9q J}eIPIULI9}

-UL AT[BOIBOTOYCI0J

eIpU] ‘iseUeUYy 1g0[-DUUKAYIDT “ZD

[Izelq «

jdeis (Jaueuloy) uand-pu “IeA
eIpu] ‘ulessy 1g0[-DUKAYIDT “9D

uedef

jdeys vdups0uajs “1eA

eIpu] ‘Wessy 190[-DUKAYIDT “Dd

uedef

eIpuy ‘und “e1yaq “
YEM 4afipiuou “IBA

eIpu ‘ung elyeq 190[-DUAAYIDT “D

{ 209 ]

Cow Lachryma-Jobi lL. Chromosomes of all the popu-
lations examined, with the exception of one (C-8 in Table
I), show differential staining into eu- and heteropycnotic
regions as in the case of C. aquatica. Heteropycnotic
segments are always present in the proximal regions of
both arms. In this species also the chromosome arms
terminate in deep staining chromomeres. In addition,
terminal as well as intercalary knobs are present. The
sizes and numbers of knobs vary in the different popu-
lations. Thus, some populations are devoid of Knobs
(e.g., population C-4b), while others possess them (e.g.,
eight knobs are present in population C-4a). Detailed
studies were not performed of knob frequencies in the
various populations. A complete analysis of the pachy-
tene chromosome morphology was achieved in one popu-
lation (C-8), in which the chromosomes do not exhibit
differential staining (Plate LVII, a, b). An idiogram
(Plate I,X) was constructed on the basis of average value
of up to five measurements of lengths and arm ratios for
sach chromosome (see ‘Table II). The sixth longest
chromosome of the complement has the nucleolus organ-
izer inthe long arm in a nearly terminal position (Plate
ILX). Plate LVIIL illustrates the nucleolus organizing
chromosome in a population (campus wild) that shows
differential staining of chromosome regions.

C. gigantea Koen. Of this species, populations have
been reported by several investigators (Darlington and
Janaki Ammal, 1945; Nirodi, 1955; Venkateswarlu and
Chaganti, 1973) with chromosome numbers of 2n=20
and 2n=40. The population that we studied (C.G. 1,
which is native to the mountain range on the west coast
of peninsular India called the Western Ghats) exhibited
variation in chromosome numbers ranging from 2n=18
to 2n=22. One of the plants with 18 chromosomes
(which at meiosis forms nine bivalents) exhibited well

tL 21:

PLatE LV

Portion of pachytene nucleus of C. aquatica showing non-homologous

association of centromeres and heteropycnotic regions.

[ 211 ]

spread pachytene chromosomes (Plate LIX, a, b). All
the chromosomes of this species, Just as in the case of
those of C. aquatica and C. Lachryma-Joln, show differ-
ential staining into eu- and heteropycnotic regions, the
latter being present in the proximal as well as interstitial
regions of chromosome arms. Furthermore, the chromo-
some arms terminate in deep staining chromomeres. An
idiogram constructed on the basis of mean values of at
least ten measurements of lengths and calculation of arm
ratios for each chromosome is presented in Plate LN
(see Table I1). The eighth longest chromosome of the
complement has the nucleolus organizer situated almost
terminally on the entirely heteropyecnotic short arm

(Plate LN).
Discussion

Variation in chromosome morphology and knob numbers
in C. Lachryma-Jobi: The species of Coix are predomi-
nantly outbreeding. Four to six varieties have been
recognized in C. Lachryma-Jobi based on morphological
features of the fruit case (Mimeur, 1951; Bor, 1960).
The varieties interbreed readily if grown together and
allowed to open pollinate and their identity is lost within
a few generations. Hence, the varietal differences are
based on gene mutations, and the varieties arose through
geographic isolation of populations or through selection
by man. The variation observed at the population level
in the chromosome phenotype: namely, the degree of
heteropyenosis in chromosome arms and the sizes and
numbers of knobs; can also be considered part of the
same evolutionary processes that lead to varietal deline-
ation. A detailed study of the variation of the chromo-
some phenotype has not yet been accomplished, but
available information points to existence of a considerable
amount of it. The fact that, in one of the populations

[ 212]

PLatTe LVI

1\Op

Chromosome 2

Chromosome 1

Chromosomes 4(A) and 5(B) Chromosome 3. “a, -

Camera lucida drawings of the five chromosomes of C. aquatica at
pachytene.

| 2ts: |

(C-8), the chromosomes did not exhibit heteropycnosis
might indicate that this feature is in some way controlled
by the genotype. It is well known that several facets
of chromosome form and behavior are under genotypic
control (Rees, 1961; Chaganti, 1965). Knobs occur on
the pachytene chromosomes of other members of May-
deae, e.g., Zea and Tripsacum. In these genera, num-
bers and positions at which knobs occur are characteristic
features of populations. In C. Lachryma-Jobi var. typica
collected at Coimbatore, a city in southwestern India,
Nirodi (1955) observed five terminal and one interclary
knobs. Ina population (C-8) of the same variety collected
at a different location (Anatagiri in the coastal mountain
range in southern peninsular India called the Eastern
Ghats), we found only two terminal knobs. In yet an-
other population (C-4a) of the same variety originating
from a different locality in India (Assam, a hilly province
in eastern India), we found eight knobs. The relation-
ship of knob variation to geographic distribution of the
populations remains to be studied.

Non-homologous associations of centromeres and hetero-
pycnotic regions in C. aquatica: In C, aquatica, at pachy-
tene, non-homologous centromeres as well as non-
homologous heteropyenotic regions are frequently asso-
ciated (Plate LV). These associations, however, fall apart
before diakinesis. The chromosomes of C. Lachryma-
Jobi and C. gigantea exhibited none of these characteris-
tics. The significance of such association is not known;
however, it is likely that exchanges might take place in
the associated regions and lead to reciprocal transloca-
tions as suggested by Venkateswarlu (1958). The genome
of C. aquatica is characterized by a degree of instability ;
complex translocations are of frequent occurrence in
natural populations(Venkateswarlu and Chaganti, 19738).

Comparison of the genomes of the three species: A com-

[ 214 |

> LVII

‘E

Puat

‘sn[Oafonu = apolo UI4oig { a1aMIO1JUaD = fold Uado : We HuO[="T + We WoYsS=G “snsfonu sues 944
jo SULMBIP Bplony, Bisulvd Y (q) ‘}UaPlAa JOU SI SUOIHAI d1joud{dolazay pue -nd Jo SUIUIeYs [VIQUSIOHIC] “9U9}

-Ayord ye (] [qe], ul g

+s | -—_—_—H1

aN

ee
i. |

Q uorjeindod) vordh} IVs 1g0f-vuhiyov'y “2 JO 2aSd010dsosJo1W & Jo snaponu aq J, (®)

( we $6
7 4

parison of the idiograms of the three species (Plate LX)
reveals that, except for the nucleolus organizing chromo-
some, the chromosomes of the three species are quite
dissimilar in their morphology. The length of each
chromosome, its arm ratio, and the contribution its
length made to the total length of the genome are given
for the three species in Table II. C. aquatica has the
longest chromosomes of the complement, with a genome
which is 401.9% long (mean chromosome length of
80.38). The genomes of C. Lachryma-Jobi and C. gig-
antea respectively are 582.0% (mean chromosome length
of 58.24) and 488.84 (mean chromosome length of
54.31) long; thus C. Lachryma-Jobi and C. gigantea
have genomes of approximately equal length. They are
also shorter than that of C. aquatica by about 40%. The
nucleolus organizing chromosomes of the three species,
even though of widely differing lengths and arm-ratios,
are similar: they are median to submedian, and the nu-
cleolus organizer occupies a sub-terminal position. The
dissimilar morphology of chromosomes between the pre-
sumably basic diploid C. aquatica and the derived poly-
ploid species C. Lachryma-Jobi and C. gigantea must
reflect, on the one hand, the effects of hybridity and
chromosome doubling which presumably have been in-
volved in the emergence of the species with higher chro-
mosome numbers, and, on the other hand, the proneness
of the genome of C. aquatica to re-patterning due to
translocation (Venkateswarlu and Chaganti, 1973). Inter-
specific hybrids between C. aquatica and C. Lachryma-
Joli, C. Lachryma-Jobiand C. gigantea, and C. aquatica
and C. gigantea are feasible, and the chromosomes in the
I’, hybrids exhibit intergenomic pairing (unpublished ob-
servations of Rao). Thus, at least one genome is common
to all three species even though its presence cannot be de-
tected by comparison of chromosome morphology alone.

[ 216 ]

Piate LVIII

B

(a) Portion of a pachytene nucleus of C. Lachryma-Jobi var. typica
(population Campus wild) showing the nucleus organizer chromosome.
This population exhibits differentinl staining of eu- and heteropcynotic
regions of the chromosomes which is evident in this cell. (b) A camera
lucida drawing of the nucleolus organizing chromosome.

Comparison of the genomes of Cow, Sorghum and maize:
The suggestion made by Anderson (1945) that maize
originated in southeast Asia as an amphidiploid of a hy-
brid between species of Coix and Sorghum possessing a
diploid chromosome number of 10, though rejected on
a number of grounds (Mangelsdorf and Oliver, 1951;
Chaganti, 1965), makes interesting a comparison of the
morphology of the chromosomes of C. aquatica and spe-
cies of Sorghum with a diploid chromosome number of
10 and those of maize. Pachytene chromosome mor-
phology has been reported by other investigators in two
species of Sorghum with a diploid chromosome number
of 10: viz., S. intrans (Garber, 1947) and S. purpureo-
sericeum (Reddi, 1958). The chromosomes of JS. intrans
are uniformly dark staining, and the centromeres are
hard to locate in them. Three pairs are of one length;
the remaining two are shorter and participate in nucleolus
organization. The chromosomes of S. purpureo-sericeum
ure in the same size range as those of maize and possess
dark staining chromomeres in the proximal regions of
both arms, making the centromeres easy to locate.
Neither species possess knobs. The genome of C. aqua-
tica, at pachytene, is about 1.7 times longer than that of
maize (data from this study compared to measurements
of maize chromosomes at pachytene given in the publi-
‘ation of Rhoades, 1955), and the chromosomes them-
selves are quite dissimilar in morphology to those of
maize. Chromosomes of C. aquatica show differentially
stained eu- and heteropyecnotic regions. ‘They are also
devoid of knobs. Maize chromosomes do not exhibit
similar differential staining, and they possess knobs.
Morphological comparisons, thus, do not reveal common
features between the genomes of C. aquatica, S. intrans,
and S. purpureo-sericeum on the one hand and maize on
the other. This observation by itself, in the absence of

[ 218 ]

PLatE LIX

<<]

2s
St

e

5s
15
e
8s @ eae

-
a
os

-

4s

Pad
-—"

6
4%

15
~

7s
eee i
7
$i
3s bad am

ae |

B
(a) The nucleus of a microsporocyte of C. gigantea at pachytene.
Differential staining of eu- and heteropycnotic regions is clearly seen.

(b) A camera lucida drawing of the same nucleus. 5=short arm;

L=long arm; open circle =centromere; broken circle = nucleolus.

genome analysis based on hybridization, cannot be con-
sidered evidence against possible existence of true phylo-
genetic relationships between these species. As discussed
earlier, even in the case of the genus Coix, comparison
of the chromosome morphology of the three species does
not indicate the presence of the common genome demon-
strated by intergenomic pairing of chromosomes in in-
terspecific hybrids.

Chromosomes of C. Lachryma-Jobi and maize exhibit
a few similarities. They fall in approximately the same
length range at pachytene (data from this study com-
pared to measurements of maize chromosomes given in
the publication by Rhoades, 1955) as well as at diakine-
sis and metaphase I (Longley, 1941). Knobs, the sizes
of which and positions on chromosomes vary in different
populations, are present in both species. There is, how-
ever, no evidence based on genome analysis to indicate
that these similarities reflect true phylogenetic relation-
ships. The many attempts made at producing viable
hybrids between maize and Coix have so far met with
failure (Mangelsdorf and Reeves, 1989: Venkateswarlu,
1963).

Summary

Lengths, arm-ratios, and other morphological features
of chromosomes at the pachytene stage of meiosis are
reported for three species of Coix, namely, C. aquatica
(2n=10), C. Lachryma-Jobi (2n=20), and C. gigantea
(2n=18). The genome of C. aquatica is the longest of
the three species and is about 40% longer than that of
either of the other two species. When stained with
acetocarmine the chromosomes of C. aquatica show con-
spicuous differential staining into eu- and heteropyenotic
regions. They also exhibit non-homologous association of
centromeres and heteropyenotic regions. C. Lachryma-

[ 220 ]

is

PLATE

‘IT 2148], Ul paquasaid soijzei-Wiie pue SYySUI] Jo San[eA UBITI Jo SISeq 94} UO
paiedaid vajunsis “cy pure ‘igof-vuhsyovy *Q ‘voipnby *Zy Jo samosomoiyo auazAyord ay} Jo sureIDOIp]

[ 221 ]

TABLE II
Chromosome lengths*, arm ratios, and percent relative contribution of each chromosome to the total genome length
at pachytene in three species of Coix, viz., C. aquatica Roxb., C. Lachryma-Jobi L., and C. gigantea.

Chromosome No. Species
C. aquatica C. Lachryma-Jobi C. gigantea
BS Percent Percent - Percent _
contribution contribution contribution
Arm to the total Arm to the total Arm to the total
Length ratio genomelength Length ratio genome length Length ratio genome length

1 107.7 1.04 26.80 92.0 1.81 15.81 80.0 1.05 16.37

m4 2 101.3 1.44 25.20 66.0 2.62 11.34 71.2 1.54 14.57
S 3 84.6** 1.50 21.05 63.2 1.89 10.86 70.0 1.55 14.32
in 4 56.9 2.33 14.16 59.2 4.58 10.17 67.2 1.70 13.75
5 51.4 1.97 12.79 59.2 1.32 10.17 58.0 3.41 11.87

6 52.0** 1.10 8.93 38.0 3.09 7.77

7 50.0 2.08 8.59 37.6 2.07 7.69

8 50.0 2.08 8.59 36.0** = 2.22 7.36

9 45.6 1.04 7.84 30.8 1.39 6.30

10 44.8 2.44 7.70

*In microns. Mean value of up to five separate measurements in the case of C. aquatica and C. Lachryma-Jobi and ten
measurements in the case of C. gigantea.

** Nucleolus organizing chromosome.

Jobi exhibits variation in chromosome morphology be-
tween populations in two features: (a) the extent of
differential staining into eu- and heteropycnotic regions ;
and (b) the numbers and positions on chromosomes of
knobs which are present on the chromosomes of this
species. The chromosomes of C. gigantea show differen-
tial staining as in the case of C. aquatica and some popu-
lations of C. Lachryma-Jobi. But they are devoid of
knobs. Comparative chromosome morphology does not
reflect the presence of the common genome (or genomes)
detected by interspecific hybridization.

Comparison of chromosome morphology does not
show similarities between C. aquatica, species of Sorghum
with a diploid chromosome number of 10, and maize:
whereas some similarities are evident between the chro-
mosomes of C. Lachryma-Jobi and those of maize. Data
on genome analysis based on intergeneric hybridization
between Coix, Sorghum, and maize are not available in
order to interpret in genetic terms the morphological
comparisons made in this study.

[ 223 ]

LITERATURE CITED

Anderson, E. 1945. What is Zea Mays? A report of progress. Chron.
Bot. 9: 88-92.

Bor, N.L. 1960. Grasses of Burma, Ceylon, India and Pakistan.
Pergamon Press, New York.

Chaganti, R.S.K. 1965, Cytogenetic studies of Maize-Tripsacum hy-
brids and their derivatives. Bussey Inst. Harvard Univ. Publ.,
pp. 1-938.

Darlington, C.D. and E.K. Janaki Ammal. 1945. Chromosome At-
las of Cultivated Plants. George Allen and Unwin, London.

Garber, E.D. 1947. The pachytene chromosomes of Sorghum intrans.
Journ. Hered. 38: 251-252.

Longley, A.E. 1941. Chromosome morphology in maize and its rela-
tives. Bot. Rev. 7: 263-289.

Mangelsdorf, P.C. and D.L. Oliver. 1951. Whence came maize to
Asia? Bot. Mus. Leafl. Harvard Univ. 14: 263-291.

Mangelsdorf, P.C. and R.G. Reeves. 1939. The origin of Indian
Corn and Its Relatives. Texas Agric. Expt. Sta. Bull. No. 574:
1-315.

Mimeur, G. 1951. Systématique specifique du génre Coir et systé-
matique varietale de Coir lacryma-jobi.. Morphologie de cette petite
céréale et étude de sa plantule. Rev. Int. Bot. Appl. Agric. Trop.
81: 197-211.

Nirodi, N. 1955. Studies on Asiatic relatives of maize. Ann. Mo.
Bot. Gard. 42: 103-130.

Rees, H. 1961. Genotypic control of chromosome form and behavior.
Bot. Rev. 27: 288-318.

Reddi, V.R. 1958. On the differentiation of A and B chromosomes
of Sorghum purpureo-sericeum at pachytene. Jour. Ind. Bot. Soc.
387: 279-289.

Rhoades, M.M. 1955. The cytogenetics of maize. In Corn and Corn
Improvement. Ed. G.F. Sprague. Academic Press, New York,
pp. 123-219.

Venkateswarlu, J. 1958. Cytological observations on spontaneously
occurring ring and chain formation in Cotx aquatica. Jour. Ind.
Bot. Soc. 37: 329-333.

Venkateswarlu, J. 1963. Cytogenetic evolution in angiosperms-
Maydeae. Mem. Indian Bot. Soc. No. 4: 65-73.

Venkateswarlu, J. and R.S.K. Chaganti. 1973. Job’s Tears. Tech.
Bull. No. 44 Indian Council of Agricultural Research, New Delhi,
pp. 1-54.

[ 224 J

NEW CHIHUAHUAN UMBELLIFERAE
BY
LINcoLN ConsTaNCE' AND Ropertr A. Byes, Jr.’

Our knowledge of the Umbelliferae of the Mexican
Kstado de Chihuahua began apparently with Wislizenus’
pioneer journey to northern Mexico just before the mid-
dle of the 19th Century, which yielded Hryngium hetero-
phyllum Kngelm. The activities of the 1880s’ of those
two doughty American collectors, Palmer and Pringle,
led to the description of Arracacia edulis S. Wats.,

vryngium madrense S. Wats., Hulophus tenuifolius S.
Wats., LZ. ternatus S. Wats., Museniopsis ternata var.
Ailifolia C. & R., Prionosciadium madrense 8. Wats., and
P. Pringle: S. Wats. A spate of field investigations at
the turn of the century by Goldman, Nelson, M. E.
Jones, and ‘Townsend and Barber led to the discovery
of Contoselinum mexicanum C. & R., Eryngium fluitans
M.E. Jones, 7. Goldmanii Hemsl., 7. medium Hemsl.,
Ligusticum Goldmani C. & R., L. madrense C. & R.,
L. Nelsonu C. & R., Museniopsis pubescens C. & R., and
Prionosciadium Townsendi C. & R. Inthe 1980s’ there
was another wave of collecting by Gentry, LeSueur,
Muller and Pennell. Gentry’s intensive study of Rio
Mayo plants led to the description of Hryngium calaster
Standl. Up to the present, some thirty taxa of Umbel-

' Department of Botany, University of California, Berkeley, Cal.
* Botanical Museum, Harvard University, Cambridge, Mass.

225 ]

liferae are known to occur in the state, according to the
taxonomic concepts of Mathias and Constance (1944-
45). Since more than one-third of these taxa belong to
the difficult genus Mryngium, this total is susceptible to
change.

In his field studies’ of the ethnobotany of the Tara-
humara centered in the Sierra Madre of southwestern
Chihuahua, the junior author has given particular atten-
tion to Umbelliferae. As a result, he obtained material
of no fewer than twenty taxa (including cytological ma-
terial of many of them). Four of the collections repre-
sent apparently undescribed species. The purpose of this
paper is to put them adequately on record.

Donnellsmithia silvicola Constance & Bye sp. nov.

Plantae perennes caulescentes humiles condensatae ;
caules simplices vel pauciramosae sparsim foliati 8-25
em. alti e radice palari tumida omnino glabrae; folia
basalia deltoideo-orbiculata 1-8 cm. longa, 1.5-5 cm.
lata, trifoliata, foliolis 8 ovalibus orbiculatisve, obtusis
lobatis vel pinnatifidis diametro 1-2 cm.: petioli 1-8
cm. longi anguste scarioso-vaginantes; folia caulina ea
basalia simulantia sed sursum reducta vaginis parum dila-
tatis: pedunculi usque ad 2 em. longi pauci alternati
graciles umbellis plerumque lateralibus sessilibus sub-
sessilibusve; involucrum nullum:; involucellum nullum;
radii 2—5 filiformes patenti-adscendentes 6-25 mm. longi:
pedicelli fertiles 0-2, 3-4 mm. longi; flores flavo-virentes ;
stylopodium depressum, stylis brevissimis ; carpophorum
usque ad basim bipartitum; fructus immaturus ovoideo-
orbicularis 1.5 mm. longus, 2 mm. latus ad apicem ver-

* Financial support between 1971 and 1974 was generously extended
by the Botanical Museum and the Department of Biology, Harvard
University, the National Geographic Society and the National Science
Foundation Grant GB-35047 for Improvement of Doctoral Disserta-
tion Research.

[ 226 |

Tears LA

a

Donnellsmithia silvicola Const. & Bye. a, Habit, <4; b, basal leaf,
“2; ¢c, flowering umbellet, «4; d, immature fruit, “10; e, petal,
10. All from Bye 6376.

sus truncatus basi rotundatus, costis humilibus filiformi-
bus; vittae et seminum superficies non visae: chromoso-
matum numerus 7 = 20.

Small, compact perennials from a fleshy taproot, 8-25
em. high, the stem simple or few-branched, very sparsely
leafy, glabrous throughout; basal leaves deltoid-orbicu-
lar, 1-8 em. long, 1.5—5 cm. broad, trifoliate with 8 oval
to orbicular, obtuse, lobed to pinnatifid leaflets 1-2 cm.
in diameter; petiole 1-8 cm. long, narrowly scarious-
sheathing: cauline leaves like the basal but reduced up-
ward, the sheaths slightly dilated; peduncles few, alter-
nate, slender, up toe 2 cm. long, but most umbels lateral
and sessile or subsessile; involucre 0, or of a single trifid
bract; rays 2-5, filiform, spreading-ascending, 6-25 mm.
long; involucel 0; fertile pedicels 0-2 (many umbellets
staminate), 8-4 mm. long; flowers greenish-yellow; sty-
lopodium depressed, not evident: styles very short;
sarpophore 2-cleft to base; immature fruit ovoid-orbicu-
lar, 1.5 mm. long, 2 mm. broad, truncate at apex,
rounded at base, the ribs low, filiform: vittae and seed
face not seen; chromosome number r= 20.

Type. Chihuahua: Municipio de Guazapares, among Pinus ponderosa,
P. leiophylla, Quercus sp. and Arbutus sp., on recently logged ridge,
elevation 8100 feet, off road northwest of Estaci6n San Rafael to Las
Lagunitas, 21 July 1974, Robert A. Bye, Jr. 6376 (GH: holotype;
ECON, UC).

The plant was rather abundant but very local on slopes
that had recently been disturbed by logging operations.

Donnellsmithia silvicola appears to be most closely re-
lated to D. serrata (C. & R.) Math. & Const. and D.
dissecta (C. & R.) Math. & Const., both species of south-
ern Mexico. It is unlike both in its lower stature and
glabrous foliage. From the former, it differs also in its
shorter rays; from the latter, in its lack of an involucel.
The most striking disparity, however, is in its uniquely
distinctive leaf form.

[ 228 |

Tauschia bicolor Constance & Bye sp. nov.

Plantae brevi-caulescentes acaulescentesve glabrae,
‘caulibus gracilibus 10-40 cm. altis e caudice gracili: folia
basalia rosulata triangulari-ovata vel ovata 1.5-8 cm.
longa, 2-6 cm. lata, profunde trilobata vel ternata vel
imparipinnata foliolis 1—8-partibus late ovalibus vel an-
guste oblongis serratis lobatisque vel pinnatifidis, superne
cinereis, inferne purpureis; petioli anguste vaginantes
1.5-10(-25) em. longi; folia caulina pauca; pedunculi
graciles 7-20 cm. longi; involucrum plerumque nullum:
radii 8-12 eis exterioribus fertilibus 5-11 mm. longis:
involucellum dimidiatum bracteolis 8-5 linearibus acutis
1-3 mm. longis; pedicelli crassi 1-2 mm. longi fructo
centrali saepe subsesseli; flores purpurei, petalis intus
albis extus purpureis; stylopodium nullum, stylis valde
gracilibus; carpophorum bipartitum dimidiis_ bifidis;
fructus ovoideus apice truncatus basi rotundatus, glaber,
2.5-8 mm. longus, 2 mm. latus, costis filiformibus acutis
quam intervallis multo angustioribus; vittae in interval-
lis plures in commissuris 2; seminum superficies plana;
chromosomatum numerus 7 = 22.

Plants rather slender, short-caulescent to acaulescent,
the stems slender, 1 or 2, 10-40 cm. high from a slender
rootstock, glabrous; basal leaves rosulate, triangular-
ovate to ovate, 1.5-8 cm. long, 2-6 cm. broad, deeply
3-lobed or ternate to pinnate with 1-8 pairs of broadly
oval to narrowly oblong, serrate and lobed to pinnatifid
leaflets 8-30 mm. long, 8-30 mm. broad, ashy-pale above,
purplish beneath; petioles 1.5-10(—25) em. long, nar-
rowly sheathing; cauline leaves few, reduced upward:
peduncles terminal, slender, 7-20 cm. long: involucre
of a linear bract or usually 0; rays 8-12, 5-11 mm. long,
apparently only the outer fertile: involucel dimidiate,
the bractlets 3—5, linear, acute, 1-8 mm. long; pedicels
stout, 1-2 mm. long, the central fruit often subsessile ;

[ 229 ]

flowers purple, the petals white within, deep purple on
the back: styles very slender, 1.5-2 mm. long, spread-
ing, purple (contrasting with a white disc); carpophore
2-parted, the halves bifid; fruit ovoid, 2.5-8 mm. long,
” mm. broad, truncate at apex, rounded at base, the ribs
filiform, acute, much narrower than the intervals; vittae
several in the intervals, 2 on the commissure; seed face

plane; chromosome number n= 22.

Type: Chihuahua: near boundary of Municipio de Batopilas and
Municipio de Urique, between Quirare and Basigochic, elevation about
6900 feet, flowers red, 8 June 1973, Robert A. Bye, Jr. 4068 (GH:
holotype; ECON, UC). It was collected in the same locality on 31
May 1973, Bye 3863 (ECON, UC), and 6 October 1975, Bye 6965
(ECON, UC).

Tauschia bicolor, known to the Tarahumara as ‘‘masii-
wari” (or ‘‘masiéwari’’)‘, is considered a quelite or pot-
herb, although it is not commonly consumed today. The
young tender leaves can be collected during the fall as
well as during the spring. ‘To date, we know of only one
locality for this herb, which is eagerly devoured by graz-
ing goats and sheep. The plants are restricted to the
western slopes above Arroyo Basigochic near an isolated
stand of Abies durangensis. ‘They are common on moist
(not wet) eroded slopes of partially open pine and oak
forest, where they often grow with Chimaphila and
Galium, while scattered individuals can be found in
crevices of white volcanic rock or in thin soil of the mixed
wood forest. Due to heavy grazing pressure, few plants
are able to produce mature fruits. The rhizomes appear
to be important in the vegetative propagation of the

Pirate LXIL. Tauschia bicolor Const. & Bye. a, Habit, » 4: b, fruiting
umbel, “2; c, mature fruit, “8; d, fruit transection, “15; e, carpo-
phore, <8; f, petal, 15; g, first basal leaf, <4. a-d, f, from Bye
4068; e, from Bye 6965; g, from Bye 3868.

Ff a F ee x as
* One informant referred to this herb as ‘“kurisove’’.

[ 230 ]

Puiate LXII

population. Indeed, this characteristic may be the key
to the survival of the population in the immediate future,
because a road building program will destroy the site
within the next year.

Although plants relatively similar to T'auschia bicolor
in habit may be found in such genera as -fletes and Arra-
eacia as Well asin Tauschia, the combination of glabrous
herbage, ternate or pinnate leaves with serrate to pinnati-
fid leaflets, fertile outer rays, dimidiate involucel, stout
pedicels with an often subsessile central fruit, petals
purple/red externally and white internally, very slender
styles, and a bipartite carpophore with bifid halves ap-
pears to be unique.

Tauschia tarahumara (Constance & Bye sp. nov.

Plantae humiles breviter caulescentes glabrae, caulibus
pluribus 12-30 cm. altis e caudice elongata ramosa car-
nosa, praeter inflorescentiam puberulentam; folia ovata
3-5 cm. longa, 2-4 em. lata, 2~-8-pinnata, divisionibus
ultimis lineari-filiformibus breviter mucronatis 5-20 mm.
longis, ca. 0.5 mm. latis; petioli 1-6 cm. longi graciles
anguste scarioso-vaginantes; folia caulina reducta sessilia
vaginis conspicue scariosis praedita; pedunculi 6-12 cm.
longi plerumque terminales graciles sub umbella puberu-
lentes; involucrum nullum; involucellum nullum ; radii
6-12 usque ad 1 cm. longi puberulenti; pedicelli 1-38
mm. longi eis exterioribus fertilibus: flores rhodo-
purpurei; styli erecti ca. 0.5 mm. longi; carpophorum
bifidum; fructus ovoideus 5-7 mm. longus, 8-5 mm.
latus ad apicem versus truncatus basi rotundatus, costis
humilibus rotundatisque quam intervallis latioribus; vit-

Prare LXIITL. Vauschia tarahumara Const. & Bye. a, Habit, <4; b,
basal leaf, “1; c, seedling, ~1; d, carpophore, 5; e, fruit tran-
section, “7; f, mature fruit, » 4: g, flowering umbellet, 5; h,
petal, ~ 20, a-b, g-h, from Bye 6288; c, from Bye 7027 (Const.
1918); d-f, from 7027.

PiatreE LXIII

tae magnae in intervallis una in commisuris 2; seminum
superficiesanguste suleata; chrosomatum numerus7= 44.

Plants low, shortly caulescent, the stems several, 12-
30 cm. high from a very elongate, branched, fleshy cau-
dex, glabrous except for the puberulent inflorescence :
leaves ovate, 8-5 cm. long, 2-4 cm. broad, 2—8-pinnate,
the ultimate divisions linear-filiform, 5-20 mm. long,
about 0.5 mm. broad, shortly mucronate; petioles slen-
der, 1-6 cm. long, narrowly scarious-sheathing, the cau-
line leaves with conspicuous scarious sheaths and sessile
reduced blades; peduncles mostly terminal, slender, 6—
12 cm. long, puberulent beneath the umbel: involucre
and involucel 0; rays 6-12, up to 1 cm. long, puberulent ;
pedicels 1-8 mm. long, the outer fertile; flowers deep
red-purple; styles about 0.5 mm. long, mostly erect:
carpophore bifid: fruit ovoid, 5-7 mm. long, 8-5 mm.
broad, truncate at apex, rounded at base, the ribs low
and rounded, broader than intervals; vittae large, soli-
tary in the intervals, 2 on the commissure; seed face
narrowly sulcate: chromosome number 7 = 44.

Tyre: Chihuahua: Municipio de Bocoyna, in wettest portions of
grazed meadow with dark mucky soil, elevation ca 7300 feet, N of
San Ignacio Arareco, EF. of Creel, 11 July 1974, Robert 4. Bye, Jr.
6288 (GH: holotype; ECON, UC). It was collected also in similar
habitats in the general vicinity of San Ignacio Arareco, SE or E of
Creel, on 14 July 1971, Bye 1535 (ECON), 16 July 1972, Bye 2369
(ECON), 5 July 1973, Bye 4150 (GH, UC), and 19 July 1973, Bye
4272B (ECON, UC), 9 October 1975, Bye 7027 (UC) and 7034
(ECON).

The older Tarahumara of San Ignacio call this herb
“*huve™” and ascribe medicinal properties to the distine-
tive rhizome. It is ground, mixed with oil or fat, and
rubbed on the affected parts of the body to treat rheuma-
tism. Only a small piece of the rhizome is placed in a
cavity in order to alleviate toothache, because it is said
to be very strong. Some young Tarahumara in the San

| 284 |

Ignacio region claim it is an edible green and call it
“sene

This deep rooted herb is restricted to a narrow inter-
zone in the moist upland meadows, between the low,
wet area (often occupied by standing water during parts
of the rainy season) with Plantago and Ranunculus spe-
cies, and the slightly elevated and drier area with J'agetes
lucida. Even though this umbel has a deep, vertical root-
stock (over 50 cm. long with a relatively constant di-
ameter) which may be well adapted to environmental
changes, such human-related activities as grazing, culti-
vation, gully and sheet erosion, draining and soil com-
paction in these accessible meadows threaten to destroy
the few known populations of this curious plant.

In the published key to the species of Tauschia
(Mathias & Constance, 1944, pp. 81-82), the new taxon
would key to either 7". tenuwifolia (S. Wats.) Math. &
Const., a little known Chihuahuan plant, or to 7°. mari-
ana (S. Wats.) C. & R. ex Drude (= 7\. decumbens
(Benth.) C. & R. ex Drude), which occurs from México
to Michoacan. It differs from both, however, by its lack
of an involucel and its smaller fruit with broad, obtuse
costae. In addition, 7". tarahumara is unlike 7". tenui-
folia in lacking a densely fibrous stem base and in its more
numerous but shorter rays. From 7°. decumbens, 7.
tarahumara may be distinguished by its lower stature,
broader leaves with narrower divisions, and red-purple
flowers, in addition to its geographic range.

Eryngium Gentryi Constance & Bye sp. nov.

Plantae perennes graciles caulescentes haud ramosi
1.8-5 dm. altae ex caudice horizontali gracili; folia pauca
alternata disticha lineari-lanceolata 8-20 cm. longa, 2-5
mm. lata, transverse septata, margine serrulato inte-
grove, venis parallelis, vaginis amplexicaulibus latitudi-

[ 235 ]

nem laminae aequantibus vel longioribus; folia caulina
summa opposita lanceolati-acuminata: inflorescentia
cymosa reducta; capitula 1-8 globosa pedunculata sub-
‘caerulea diametro 5-10 mm. ; bracteae involucrales 8-12
lanceolatae acutae integrae vel pauci-dentatae, inferne
virides, superne argenteae, quam capitulum duplo longi-
ores: bracteae florales lineari-acuminatae integrae fruc-
tum multo excedentes; sepala ovata ca. 2 mm. longa;
petala lineari-oblonga 1—1.25 mm. longa apice fimbriata;
styh graciles 2.5-8 mm. longi quam calyces longiores;
fructus (immaturus) ovoideus diametro ca. 2 mm., super-
ficiebus mericarpiorum dense squamatis, squamis ovatis
subaequalibus complanatis; chromosomatum numerus
n=T.

Plants slender caulescent perennials 1.8-5 dm. high
from a slender rootstock, the stems solitary to several,
weakly erect, unbranched below inflorescence: lower
leaves few, alternate, distichous, linear-lanceolate, 3-20
em. long, 2-5 mm. broad, remotely serrulate or entire,
the venation parallel, transversely septate, amplexicaul-
sheathing, the sheath as broad or broader than the blade ;
uppermost cauline leaves opposite, lanceolate-acuminate ;
inflorescence a reduced cyme of 1-38 pedunculate heads:
heads globose, bluish, 5-10 mm. in diameter; involucral
bracts 8-12, spreading-ascending, lanceolate, 8-25 mm.
long, 2-5 mm. broad, acute, entire or with one or two
pairs of short teeth, green beneath, silvery-white above,
about twice as long as the head: bractlets linear-acumi-
nate, entire, much longer than fruit; sepals ovate, about
2 mm. long, mucronate; petals linear-oblong, 1—1.25
mm. long, with a narrower fimbriate tip; styles slender,

Prater LXIV. Eryngium Gentryi Const. & Bye. a, Habit, <4; b,
cauline leaf, «1; c, floral bractlet, «10; d, involucral bracts, 2;
e, immature head (two bracts removed), <2; f, lower, 10; g, petal,
<20. All from Bye 4766.

[ 236 ]

PuateE LXIV

2.5-8 mm. long, longer than sepals; fruit (immature )
ovoid, about 2 mm. long and broad, the squamae ovate,
subequal, flattened, the dorsal faces densely squamose ;
chromosome number ”=7.

Tyree: Chihuahua: Municipio de Ocampo, transition zone, pines,
meadow, infrequent, marginal to streams, elevation 7500 feet, Meme-
lichi, Rio Mayo, 16 September 1936, H.S. Gentry 2770 (UC: holo-
type: kK).

Oruer coLLections: Chihuahua: Municipio de Madera, lake near

Chuichupa, 23 August 1936, H. LeSueur 817 (UC); Chuichupa, 7000
feet, 21 September 19038, /.h. Diehl s.n. (POM): Municipio de Bo-
coyna, in moist meadow E of Gonogochi (on Continental Divide) E
of Creel, associated with low, scrubby Quercus, grasses, Ranunculus,
Cacalia and Tagetes lucida, elevation ca 7500 feet, 24 August 1973,

Robert 4. Bye, Jr. 4766 (ECON, UC).

Kryngium Gentryi represents a third member of Sec.
XN XI. Madrensia Wolff, comprising Mexican species of
wet habitats characterized by evidently transversely sep-
tate leaves, bluish or purplish ovoid-globose to ovoid-
cylindrical heads, and involucral bracts conspicuously
silvery above (Wolff 1918). The other two species, /.
madrense S. Wats. and 7. fluitans M.K. Jones (27. meai-
eanum sensu Wolff, 7. Wolffit Mathias), are very similar
in habit. ‘There is a resemblance also to Ht. phyteumae
Delar.f. (Sect. NNIV. Stellata Wolff), but this pos-
sesses well developed oblong to lanceolate, reticulately
veined leaf lamina. In Sect. Madrensia, the leaves are
usually bladeless or with a rudimentary lamina and not
only the petiole but the entire leaf tends to be septate.
The silvery upper surface of the involucral bracts sug-
gests an affinity to both Sects. NXNIV. Stellata Wolff
and NNV. Carliniformia Wolff. Wolff notes that /.
phyteumae ‘‘bildet einen deutlichen Ubergang zu den
Madrensia’” (Wolff 1913, 61: 186). From its two allies
in Sect. Madrensia, 47. Gentry: differs in its possession
of a rootstock, its fewer globose heads without a coma,

[ 288 |

and in its broader involucral bracts that much exceed
the head.

The type collection was associated by Standley with
the unpublished herbarium name ‘‘/. juncifolium M.E.
Jones’’, which Jones had applied to specimens of both
I. flutans and 7. Gentryt. Gentry reported the collec-
tion in his Rio Mayo flora (1942) as H. phyteumae, as
erroneously determined by Mathias and Constance. The
“H. juncifolium Jones” listed by LeSueur in his pioneer
study of Chihuahuan vegetation (1945) and attributed
to the Montane Forest, is also 47. Gentry. The invalid
name, ‘2’. juncifohum M.E. Jones’’, is not to be con-
fused with the South American /1. juncifolium (Urban)
Math. & Const.

It is with great pleasure that we name this species in
honor of Dr. Howard Scott Gentry, who has made such
significant contributions to our knowledge of the flora
of Chihuahua and other parts of Mexico.

SPECIMENS

Duplicates of the Bye collections will be deposited in
the Herbario Nacional del Instituto de Biologia, Uni-
versidad Nacional Aut6noma de México (MEXU),.

[ 239 ]

LITERATURE CITED

Gentry, H.S. 1942. Rio Mayo Plants. Carneg. Inst. Wash. Publ.
527, 328 pp.

LeSueur, H. 1945. The ecology of the vegetation of Chihuahua,
Mexico, north of Parallel Twenty-eight. Univ. Texas Publ. 4521,
92 pp.

Mathias, M.E., and L. Constance. 1944-45. Umbelliferae: Fl. N.
America 28B: 43-295.

Wolff, H. 1913. Umbelliferae; in Engler, Das Pflanzenreich 61
(4°"): 1-305,

[ 240 |

CHROMOSOME NUMBERS IN
CHIHUAHUAN UMBELLIFERAE
BY
LINCOLN ConstaNCcE!, Tsan [anc CHUANG?
AND Ropsertr A. Byer, Jr.®

In his ethnobotanical studies of the ‘Tarahumara cul-
ture in the Sierra Madre of southwestern Chihuahua,
Robert A. Bye, Jr., made a concerted and very success-
ful effort to obtain material of Umbelliferae for chromo-
some counts. The results of these counts (made by
Chuang) are shown in the accompanying table and Plate
LXV, and some brief comments (by Constance) follow

(Table LT and Plate LXV, a-k).

Donnellsmithia C. & R. The report on D. silvicola is
new: that of 7=20 for D. ternata disagrees with one of
n=21 for the same taxon (Constance, Chuang and Bell,
in press). Chromosome numbers in this genus are still
confusing. Counts made on eight specific taxa have
yielded the following haploid numbers: 11, 14, 20, 21,
22, 40, 42, 44. Thus, both D. silvicola and D. ternata
are probably tetraploid.

' Department of Botany, University of California, Berkeley, Cal.

* Department of Biological Sciences, Illinois State University,

Normal, II].

3 . ae . .
Botanical Museum, Harvard University, Cambridge, Mass.

C241 |

EXPLANATION OF THE ILLUSTRATION

Karyotypes of Chihuahuan umbelliferae

a, Donnellsmithia ternata (S.Wats.) Math. & Const., Diak., Bye 4393:
b, D. silvicola Const. & Bye, Diak., Bye 6376; ¢e, Tauschia bicolor
Const. & Bye, Diak., Bye 3863; d, 7. tarahumara Const. & Bye, MII,
Bye 6288; e, drracacia edulis S,Wats., Diak., Bye 6753; f, Eryngium
Jluitans M.E.. Jones, MI, Bye 4763; g, FE. Gentryi Const. & Bye,
Diak., Bye 4766; h, FE. heterophyllum Engelm., Diak., Bye 4334; i,
Ek. Lemmonii Coult. & Rose (EF. calaster Standl.), MI, Bye 4463; j,
EF. Lemmonii Coult. & Rose (FE. calaster Standl.), MII, Bye 4809; k,
KE, phyteumae Delar.t., MIL, Bye 5905 (Const. 1906). (Diak.—= Dia-
kinesis; MI, MII-Metaphases I and II. All ea. 1050).

[ 242 ]

PLATE LXV

op a
o 2 eo eC
Qha .S J 4
g ool o? ‘
6 go hy
tf i
> al o%
r) ®e@ oe &
oe oh ie Se
q ay seve
@ 5% °
Ne 4° -*
eo
Q eof
© - a? "6
e %°% Sf
fle whe Ff 6?
é 4
a
Po oj

Tauschia Schlecht. Both 7’ bicolor and T. tarahumara
were previously uncounted. Five other species of the
genus have revealed only n=11 and n=22, so the count
of n=44 for 7. tarahumara, a probable octoploid, rep-
resents a new polyploid level for the genus.

Arracacia Baner. There is no previous chromosome
count for A. edulis, a species which has been placed al-
ternately under T'auschia. The complement of 72=20 is
equally discordant in both genera. The fifteen specific
taxa of Arracacia examined heretofore have all shown a
haploid number of 22, except for one unexplained occur-
rence each of the numbers 14 and 82.

Ligusticum L. The count of n=11 for L. Porter,
which has been made before, agrees with findings on
some 16 species of this cireumboreal genus, all of which
showed n=11, n =22, or 2=33, save for an anomalous
n=12 and one n=27.

Pseudocymopterus C. & R. The count of 2=11 for
P. montanus coincides with earlier findings for this genus

and species.

Prionosciadium S$. Wats. Of eight taxa of this genus
counted, six have n=22, two have n=21. A count of
22 for P. madrense conforms to earlier information.

n=

gryngium L. Of the six counts reported here, only
that for #. Gentryi, n=7, is new. From some 100 spe-
cies that have been counted in this very distinctive cos-
mopolitan genus, it seems clear that the basic number is
n=8. However, there are descending aneuploid series
extending to n=7, 2 =6, or even nm=5 in both hemi-
spheres, sometimes in what appears to be the same spe-

[ 245 |

cies, as in £7. heterophyllum and EB. Lemmoni. Extensive
polyploidy, particularly in the Americas, has arisen on
the basic complement of 8 (7 =16, 24, 82, 40, 48), and
smaller series based apparently on 7 in Eurasia.

LITERATURE CITED
Constance, L., IT. 1. Chuang and C.R. Bell. Chromosome numbers in
Umbelliferae V. Amer. Jour. Bot., in press.

[ 246 |

TABLE I

Chromosome numbers of Chihuahuan Umbelliferae.

Name

Voucher
Specimen*

Municipio and Locality

Donnellsmithia
1. ternata(S.Wats.) C. & R.

2, stlvicola Const. & Bye
Tauschia

3. bicolor Const. & Bye

4. tarahumara Const. & Bye

Arracacia
5. edulis S.Wats.

Ligusticum
6. Porteri C. & R.

Pseudocymopterus

7. montanus (A.Gray)C.& R.

Prionoseiadium

8. madrense 5. Wats.
Eryngium

9. fluitans M.E. Jones

10. Gentryi Const. & Bye

11. heterophyllum Engelm.

12. Lemmonii C. & R.

13. phyteumae Delar.f.

44:

20

11

Bye 4393

Bye 6376

Bye 3863

Bye 6388

Bye 6753

Bye 6651

Bye 6637

Bye 6352

Bye 4763

Bye 4766

Bye 4334

Bye 4463
Bye 4809

Bye 5905

Guachochi, Cusdrare

Guazapares, San Rafael-
Las Lagunitas

Batopilas—Urique, between
Quirare and Basigochic

Bocoyna, San Ignacio
Arareco

Guachochi, Cusdrare

Bocoyna, W of Creel

Bocoyna, W of Creel

Batopilas, S of Quirare

Bocoyna, W of Gonogochi
Bocoyna, E of Gonogochi
Bocoyna, Gonogochi
Batopilas, N of Quirare

Bocoyna, Creel—San
Ignacio Arareco

Bocoyna, W of Creel

* Voucher specimens are preserved in ECON, GH and UC.

ECONOMIC BOTANY LIBRARY
OF OAKES AMES
HARVARD BOTANICAL MUSEUM

BOTANICAL MUSEUM LEAFLETS
HARVARD UNIVERSITY

CamBripGe, MassacnuusetTts, JuNE 30, 1976 VoL. 24, No. 9

THE GENUS OECEOCLADES LINDL.
BY
Lestige A. GARAY AND PETER TayLor*

In 1832 while describing Oeceoclades, as a genus differ-
ent from Angraecum, Lindley offered the following ob-
servation: ‘‘The genus Angraecum. . . is known by its
undivided lip, which is neither cucullate, nor articulate
with the column; by its spreading perianthium, which
never has the segments turned upwards as in Eulophia;
by its long taper-jointed spur, which is rarely enlarged
at the base; and finally, by all these characters being
connected with coriaceous leaves that are never ribbed
or plaited. Such being the definition of Angraecum, it
will be apparent that, . . . such as our Angraecum macu-
latum and a few of those of Du Petit Thouars, must be
excluded; these form a genus nearly related to Eulophia,
from which they are to be distinguished by their coria-
ceous leaves, by the perianthium never being secund, and
by the want of a crest upon the lip.”

With this introduction Lindley also provided a sepa-
rate generic description in Latin as well as a list of names
—five under Angraecum and four under Limodorum —
which he considered belonged to Oeceoclades, but with-
out making the proper nomenclatorial transfers. ‘The
following year, in 18338, in his Genera and Species of
Orchidaceous Plants, he treated Oeceoclades in a similar

* Orchid Herbarium, Royal Botanic Gardens, Kew, England.

[ 249 |

fashion, but with a somewhat different content. On both
occasions, however, he included the sympodial Kulophia-
like Angraecum maculatum with the remainder, all
monopodial orchids. At that time Lindley remarked:
‘‘Oeceoclades will probably comprehend all the Eulophia-
like epiphytes... ; it is very near Eulophia, from which
its coriaceous, not plaited leaves, distinguish it among
other things. ”’

Thus, from the very beginning Lindley considered
Oecceoclades to represent Eulophia-like plants. He em-
phasized his conviction once more in 1859 (Journ. Linn.
Soc. 83: 86) by saying ‘*... it is probably that O. macu-
lata is the only plant to which the generic name will
attach.’ Since that time the remaining species of Oeceo-
clades have all been transferred to various angraecoid
genera. Consequently the genus Oeceoclades must be
typified by O. maculata, the only Eulophia-like plant.

In 1887, Pfitzer in his Entwurf einer natiirlichen An-
ordnung der Orchideen p. 87-88 established the mono-
typic genus Mulophidium which he also based on Angrae-
cum maculatum. It is difficult to understand why he chose
to follow that particular course when he was fully aware
of Lindley’s suggestion that the name Oeceoclades be
applied only to O. maculata: ‘‘Ich muss Hulophidiwm
maculatum entsprechend Lindley’s Vermuthung als den
Typus einer besondren Gattung betrachten. . . “—

Since both genera, Oeceoclades Lindl. and Hulophidium
Pfitz. are based on the same type—Angraecum macula-
tum V.indl.—Oececeoclades must be reinstated because of
the rule of priority.

Summerhayes, in 1957, published a synopsis of the
genus Hulophidium (Bull. Jard. Bot. Bruxelles 27 (3):
391-403). In that study he argued that Pfitzer and sub-

** In keeping with Lindley’s supposition, I must regard Eulophidium
maculatum as the type of a special genus.”’

[ 250 ]

sequently Schlechter, both in their studies emphasized
the vegetative aspects of the plants without paying much
attention to floral details, which resulted in a rather poor
circumscription of the genus.

‘l'o augment this one-sided presentation, Summerhayes
provides the following observations:

‘*For some time now I have been struck by the marked
similarity in floral structure between typical members of
Eulophidium, such as F. maculatum (LINDL.) PFITZ.,
on the one hand, and species which have always been re-
tained in Hulophia such as I. saundersiana RCHB.F.
and the Asiatic 17. macrostachya LIN DL., on the other.
All these species have a marked quadrilobed labellum
with two short parallel or slightly divergent calli at the
base and no long keels or hair-like outgrowths such as
are so widely distributed in Mulophia. The side lobes
almost invariably have marked darker veins. . . Some-
time the two lower lobes, or lateral lobes if you prefer
to call them that, are much reduced, occasionally so
much that the labellum is almost bilobed. The spur is
relatively short and often swollen, sometimes it is more
or less shortly bilobed at the apex.

‘*As regards the vegetative structure the aerial pseudo-
bulbs may be heteroblastic with 1-8 leaves at the apex
or homoblastic (with several elongated internodes) with
one or more leaves at the apex. All intermediates can
be found between a clearly heteroblastic condition with
no cataphylls or leaves arising along the pseudobulbs,
through forms in which the lower swollen internodes are
quite short and other forms with only 2 elongated swol-
len internodes, to typical homoblastic conditions where
there are clearly several well-defined elongated swollen
internodes with cataphylls arising from the lower nodes
and leaves from the uppermost.

“The great majority of species have markedly petio-

[ 251 |

late leaves, but in a few cases the petiole is very short.
It is almost invariably articulate with a number of sharp
or blunt teeth at the joint, this articulation usually being
some distance above the base of the leaf, and sometimes
in the centre of the long slender petiole. Reichenbach
and Schlechter, when dealing with some species, have
treated the lower part of the petiole below the articula-
tion as being the rostrate upper internode or prolonga-
tion of the pseudobulb, but this is clearly not a possible
interpretation in plants like J”. sawndersiana where there
are two leaves, both with their petioles articulate some
distance above the base. The leaves are usually rather cori-
aceous and often banded or spotted with paler markings.

‘* As these variable characters are associated with very
similar floral structures I feel that too much emphasis
should not be placed on them and that all these species
should be placed in an enlarged Hulophidium.”

In no way diminishing the value of these important
observations, we are compelled to reexamine some of its
crucial points. We believe that the distinction between
homoblastic and heteroblastic pseudobulbs is not clearly
understood by many. Both types of pseudobulbs are de-
rived from the common sympodial stem which consists
of aset of nodes and internodes. In the case of the homo-
blastic pseudobulbs, each internode or most of the inter-
nodes are equally developed and enlarged throughout
the entire length of the sympodium, at the nodes leaves
or sheaths or cataphylls are produced. In the case of the
heteroblastic pseudobulbs only one internode is fully de-
veloped and enlarged throughout the entire length of the
sympodium. The nodes are closely approximate, often
so congested that no visible separation is apparent ; rarely
the leaf-bearing nodes are separated by obvious distances,
yet minimal in proportion to the internode that repre-
sent the true heteroblastic pseudobulb. The one to three

[ 252 |

leaves produced at the top of a heteroblastic pseudobulb
are originating independently from separate nodes, which
may or may not be closely approximate. The misunder-
standing of the nature of the heteroblastic pseudobulb
caused Summerhayes to enlarge and alter the generic
description of Mulophidium, which unfortunately now
encompasses several criteria applicable only to Hu/lophia.

Likewise the similarities in floral structure, mentioned
by him, we believe are produced through convergent
evolution rather than through the processes of speciation.

Although florally Hulophia macrostachya is very simi-
lar to those found in many plants of Oeceoclades, the
thin, plicate leaves and the homoblastic pseudobulbs
immediately exclude it from that relationship. If FH.
macrostachya is to be admitted to Oeceoclades, then FE.
graminea, I’. euglossa, E.. guineénsis, and other related
species would have to be included also. For additional
names see the list of Mulophia Sect. Pulchrae Kral. at
the end of this paper.

Both Lindley and Pfitzer were explicit about such
generic characters as the heteroblastic pseudobulbs, cori-
aceous, conduplicate leaves and Eulophia-like flowers.
Consequently we adhere to the original circumscription
of the genus in our assignment of the species.

Oeceoclades Lindl. in Bot. Reg. 18: sub t. 1522, Sept.
1, 18382.
Syn.: Aeceoclades Duch. in Orbigny, Dict. 9: 170,
1849.
Saccolabium Sect. Oecceoclades (Lindl.) Cor-
dem., Fl. Reunion 197, 1895.
Lectotype: Angraecum maculatum Lindl.—
Lindl]. in Journ. Linn. Soc. 3: 36, 1859.
Eulophidium Pfitz., Entw. Natur. Anordn.
Orch. 87-89, 1887.

[ 253 ]

Eulophia Sect. Hulophidium (Pfitz.) H. Perr.
in Bull. Soe. Bot. Fr. 82: 147, 19385.
Lissochilus Subgen. Hulophidium (Pfitz.) H.
Perr. in Humbert, Fl. Madag. Orch. 2: 17,
1941.

Type: Angraecum maculatum Lindl.—Pfitz. did.

Sepals and petals variously spreading; lip 8-lobed,
basally produced in a spur, midlobe commonly lobulate
or emarginate; disc either with a pair of approximate,
quadrate or triangular calli at the entrance to the spur
or with three variously thickened, parallel ridges which
together with the lateral veins are sparsely but distinctly
papillose or hirsute; column erect, rather short, oblique
at base; stigmata confluent; rostellum short; anther
cucullate to cristate; pollinia 2, on a short or rudimen-
tary stipe; viscidium large.

Pseudobulbs more or less approximate, heteroblastic,
one- to three-leaved at apex; leaves coriaceous, condu-
plicate, never plicate, commonly petiolate, rarely sessile,
articulate with colliferous apex of pseudobulbs; inflores-
cence lateral, racemose or paniculate; bracts inconspicu-
ous; flowers rather small and thin in texture, resupinate.

31 species native to tropics and subtropics of Sey-
chelles, Madagascar, the Mascarene Islands, Africa,
South America, West Indies and Bahamas.

Type of the genus: Angraecum maculatum Lindl.

Key to Species

1. Petals at most 4 to 4 the length of the sepals . . . . . . 2
la. Petals and sepals more or less equal in length . . . . . . 6
2, Pseudobulbs ovoid to cylindrical; leaves lorate [at

least 40 em. long]; inflorescence diffusely branched. . . . 3

2a. Pseudobulbs globose to pyriform; leaves linear [at
most 20 em. long]; inflorescence racemose or rarely
with few short branches . . . . . e 7 ee ee ee ee 4

[ 254 ]

3a,

8a.

9a,

Pseudobulbs 2-leaved; sepals spathulate to oblan-

ceolate, obtuse; petals elliptic, obtuse; spur of lip

forward projecting under lip... .. . . .. . O.calcarata
Pseudobulbs 1-leaved; sepals and petals lanceolate-

elliptic, acute to subacuminate; spur of lip project-

ing away fromlip. ......... .. .. O. Hebdingiana
Sepals spathulate, obtuse; spur globose . . . O. spathulifera
. Sepals obovate-oblanceolate, acute. . .......4.. 5

Leaves sessile; sepals at least 14 mm. long; petals

elliptic, acute; spurcylindric. . .... . . O. Decaryana
Leaves petiolate; sepals not more than 8 mm. long;

petals suborbicular, obtuse; spur subglobose, ven-

trally compressed. . ... . 2. . « . OJ angustifolia
Lip ecallose at base; basal Baise of 3 parallel veins
of disc somewhat carinate-thickened . ... . - ee ee

Lip with a bilobed callus, or bilamellate at or near Bee - » « 1S
Pseudobulbs elongate, slender, fusiform to cylindric,

2-3 leaved, approximately the length of the petiolate
feaves or longer 4-5. we ew: eB a ae wwe me ee a 8
Pseudobulbs short, ovoid, 2-leaved, much shorter

than the petiolate leaves. . . A =
Lip broadly elliptic; midlobe of lip senda: at oe
overlapping with lateral lobes without a sinus; disc

ecallose at junction of lateral and median lobes ..... 9
Lip narrowly ovate-oblong to elliptic-oblong; midlobe

of lip cuneate at base, forming a distinct sinus with

lateral lobes; disc with a pair of fleshy gibbosities at

junction of lateral and midlobes. . . . .. . . O.ugandae
Midlobe of lip suborbicular in outline, half as long

as the entire length of the lip; lateral lobes subfal-

cate, obtuse at apex. .. . . « . O, lanceata
Midlobe of lip reniform in eee one- fourth the

length of the entire lip; lateral lobes broadly

rounded atapex ........... . . O.seychellarum
Lip lobate from middle; lateral lobes of lip
teunente in ont «6 6 «ow a 4: 4 ee ee ee oe DT

Lip lobate one-fourth from apex; lateral lobes of lip

OUUduE 1 TORE 6s a eG Se we RS De HS SB AB
Lip wider than long; midlobe deeply emarginate to
divaricately bilobulate in front . .. . . . .O. Lubbersiana
Lip longer than wide; midlobe cuneate or with a

distinct claw, at most retuse to indented in front O. pandurata
Inflorescence laxly racemose; flowers greenish-

[ 255

16.

16a.

18a,

19a.

20.

20a.

21.

yellow with maroon dots; sepals and petals

ovate-oblong, acute... . ; . 2 2. « « O. atrovirens

Inflorescence with short ecnelieds ncn white,

sepals and petals linear-oblong, obtuse . . . . . O. latifolia

Blades of leaves narrow, linear . . . . ..... =... 14
a. Blades of leaves broad, ovate to elliptic. . . ..... .17

Leaves long-acuminate, lorate, 50 cm. or more long,
gradually tapering to pseudobulb without a distinct

Wellgle 4. 9 40% % « & & &'% ce eae oo & OD Perse
Leaves acute or obtuse, linear- sieve: 30 cm. or less

long, petiolate or subpetiolate .. . a ¢ aa oe
Spur vesicular, longer than lip; median aes of lip

reflexed; disc in front of callus inormate . . . O. quadriloba

. Spur cylindrical, shorter than lip; median lobe of

lip not reflexed; disc in front of callus 3-carinate . . . . 16
Pseudobulbs 2-leaved ; leaves subpetiolate, 20-30

cm. long; inflorescence racemose; lateral lobes of

lip rounded, larger than median lobe . . . . O. sclerophylla
Pseudobulbs 1-leaved; leaves petiolate, less than

15 cm. long; inflorescence branched ; lateral lobes

of lip subquadrate with obtuse angles, equal to, or

somewhat smaller than median lobe . . . . QO. analavelensis
Plants caespitose ; pseudobulbs aggregate, ecol-

liferous or with hardly any projections; leaves

sessile or with conduplicate, short, petiole-like

base ... ‘ ee eo hie 2 ow ee ole
Plants diicsdutouxt : spendouaihe ners.

prominently colliferous, leaves distinctly petiolate . . . . 25
Pseudobulbs 2-leaved . ......... ++ 4. 419
Pseudobulbs l-leaved . . . . 1... 1 ee ee ee ee 20

Inflorescence profusely paniculate; lateral lobes of

lip faleate, when expanded parallel with, and as long

as midlobe; midlobe of lip wider than long; spur

conical, acuminate ..... . 2. ee «OJ gracillima
Inflorescence racemose or rarely with one or few

short branches; lateral lobes of lip triangular to

oblong, when expanded rectangular with midlobe ;

midlobe of lip as wide as long, subquadrate; spur

vesicular, obtuse. . . . .-. . . +. . . QO. roseovariegata
Leaves ovate, subcordate at base... ....... + 21
Leaves elliptic to oblong-elliptic, cuneate at base . . . . 22

Sepals and petals similar, elliptic to ovate-elliptic,
obtuse; lateral lobes of lip much larger than mid-

[ 256 |

Qla.

22.

22a,

28a.

29a.

30.

30a.

lobe; column short, erect . . . .. . . . . . O. boinensis
Sepals and petals dissimilar, linear-lanceolate,

acute; lateral lobes of lip as large as midlobe;

column elongate, arcuate ......... . . O. Rauhii
Lip as long as or shorter than wide; terminal lobe

separated from the lateral lobes by acute, indented

or rounded sinuses .... . a a ae ee ee
Lip as long as or longer than wide: terminal lobe

separated from the lateral lobes by a distinct

isthmus... . . . « O. monophylla
Spur eapelonece: ee of lip bilobed . . . . O. ambongensis
. Spur clavate to cylindric; callus of lip bilamellate . . . . 24

Sepals and petals acute: lip as long as wide; sinuses

in middle of lip; inflorescence always racemose . O. maculata
Sepals and petals obtuse; lip shorter than wide;

sinuses one-third from apex of lip; inflorescence

racemose or subpaniculate .. . » . . « « OF Mackeni
Pseudobulbs narrowly cylindrical, often stem-like,

terete, two-leaved; lateral sepals longer than dor-

sal sepal; lip equally 4-lobed . . . . . . . O. Saundersiana
Pseudobulbs ovoid to conical, 1-leaved; sepals of

same length; lip flabellate or unequally 4-lobed . . . . . 26
Lateral lobes of lip larger than median lobe. . . . . . . 27
Lateral lobes of lip much smaller than median

lobe, ear-like . ... . & ; : ; ee 4 ee
Leaves lanceolate, joni-abeenuaie, si bachametes

sepals and petals apiculate . . . fu ea 4 0. sansibarica

Leaves lanceolate-elliptic or Sacco elliptic,

acute to subobtuse; sepals and petals obtuse O. alismatophylla
Callus of lip and the disc in front of it puberulent . . . . 29
Callus of lip and dise glabrous. . ‘5 - A moo
Leaves broadly elliptic, large, 7- caened. with esti

30-50 em. long; column-wing ciliolate-hirsute along

margins; lateral lobes of lip rounded. . . O. cordylinophylla
Leaves narrowly ovate-lanceolate, 1-3 nerved, with

petiole 13-20 cm. long; column-wings glabrous along
margins; lateral lobes of lip obliquely triangular-

faleate ... . 2. 2. . O. analamerensis
Lip donuheleted 4 in aaidale. sccicniate: basal part

suborbicular, apical part divergingly bilobed with

rounded lobes; disc bilamellate at base in front

of which are 8 thickened veins . ..... . . O. pettolata
Lip not pandurate, 3-lobed; lateral lobes obtusely

[ 257 |

angular, semiovate, terminal lobe divergingly

bilobulate with rectangular-oblong lobules;

disc without thickened veins in front of basal

lamellae . ........... . . . . 0. lonchophylla

ENUMERATION OF SPECIES

Oeceoclades alismatophylla (Rchb.f.) Garay &
Taylor, comb. nov.
Basionym: HMulophia alismatophylla Rehb.f. in Flora
68: 5438, 1885.
Syn.: Mulophidium alismatophyllum (Rehb.f.) Sum-
merh. in Bull. Jard. Bot. Bruxelles 27: 894,
1957.
Type: Madagascar, Forét d’Ankaye. Coll. Humblot
s.n.f (W).
Distribution: Madagascar.

Vegetatively the plants of this species are very similar to O. anala-
merensis, O. petiolata and O. lonchophylla, but readily distinguishable
from them in the floral structure, especially in the shape of the lip
with the lateral lobes being larger than the median Jobe. We have
seen only the type specimen.

Oeceoclades ambongensis (Schltr.) Garay & Tay-
lor, comb. nov.
Basionym: HMulophidium ambongense Schltr. in Ann.
Mus. Col. Marseille, ser. 3, 1: 182, t. 17,
1913.
Syn.: Hulophia Schlechtert H. Perr. in Bull. Soc. Bot.
Fr. 82: 154, 1935.
Lissochilus Schlechteri (H. Perr.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 27, 1941.
Type: Madagascar, Manongarivo (Ambongo). Coll.
Perrier no. 1684! (P).
Distribution: Madagascar.

From the related species of the O. maculata alliance, the plants of
this species are readily identifiable by the comparatively larger flow-
ers, subglobose spur and the bilobed callus of the lip.

[ 258 ]

Oeceoclades analamerensis (H. Perr.) Garay &
Taylor, comb. nov.
Basionym: Lissochilus analamerensis H. Perr. in Not.
Syst. 8: 42, 1939.
Syn. : Hulophidium analamerense (H. Perr.) Summerh.
in Bull. Jard. Bot. Bruxelles 27: 894, 1957.
Type: Madagascar, Province de Diégo-Suarez, Anal-
amera, rive droite de la riviére Analabe, afflu-
ent du Rodo. Coll. Humbert no. 19247! (P, Kk).
Distribution: Madagascar.

The two small approximate lamellae near the base of the lip and
the hirsute dise of the lip amply separate the plants of this species
from the vegetatively similar O. alismatophylla.

Perrier gives Humbert no. 19020 as the type number, but the speci-
mens in Paris as well as at Kew bear the number 19247!

Oeceoclades analavelensis (H. Perr.) Garay & Tay-
lor, comb. nov.
Basionym: Lissochilus analavelensis H. Perr. in Not.
Syst. 8: 41, 1939.
Syn.: Hulophidium analavelense (H. Perr.) Summerh.
in Bull. Jard. Bot. Bruxelles 27: 895, 1957.
Type: Madagascar, Forét d’Analavelona, au N. du
Fiherenana. Coll. Humbert no. 14218! (P,K).
Distribution: Madagascar.

Florally the plants of this species are very similar to those of O.
sclerophylla, both having three thickened ridges in front of the callus,
but they are very dissimilar vegetatively.

Oeceoclades angustifolia (Sengh.) Garay & Taylor,
comb. nov.
Basionym: Hulophidium angustifolium Sengh. in
Adansonia ser. 2, 6: 558, 1967.

Type: Madagascar: near Diégo-Suarez. Coll. Rauh
& Buchloch no. 7987 (HEID).

Syn.: Hulophidium angustifolium ssp. diphyllum
Sengh. in Adansonia ser. 2, 6: 561, 1967.

[ 259 |

Type: Madagascar, near Sakaraha, river Fiherenana.
Coll. Rauh no. 10423 (HEID).

Distribution: Madagascar.

From the related O. Decaryana, the plants of this species are dis-

tinguished in having petiolate leaves as well as differently propor-
tioned lips.

Oeceoclades atrovirens (Lindl.) Garay & Taylor,
comb. nov.
Basionym: HMulophia atrovirens Lindl., Gen. and Sp.
Orch. Pl. 184, 1833.

Syn.: Graphorchis atrovirens (Lindl.) O. Ktze., Rev.
Gen. Pl. 2: 662, 1891.

Type: India, without proper locality. Coll. Wallich
s.n.! (KX).

Distribution: India orientalis.

This peculiar plant is known from a colored drawing prepared by
Wallich in 1828 for the East India Co., now in the Kew Herbarium.
Ironically, Lindley originally bas written on the drawing ‘‘Oeceo-
clades’’, then at a later time he crossed it out. So far no specimens
are known to exist of this plant. Judging from the drawing, especially
the floral details, it is near O. latifolia from which it differs in having
a simple raceme, rather pointed sepals and petals and a somewhat
different lip.

Oeceoclades boinensis (Schlitr.) Garay & Taylor,
comb, nov.
Basionym: Mulophidium boinense Schltr. in’ Ann.
Mus. Col. Marseille ser. 8, 1: 182, t. 17,
19138.
Syn. : Lissochilus boinensis (Schlitr.) H. Perr. in Hum-
bert, Fl. Madag. Orch. 2: 26, 1941.
Type: Madagascar, bords de la riviére Andranofasy
(Boina). Coll. Perrier no. 1884! (P).
Distribution: Madagascar.
This species has its only relative in O. Rauhi, both having a more
or less cordate base to the leaves. In floral structures they are, how-

ever, very different. The type number of O. boinensis is Perrier 1834!
and not 1384 as given by Schlechter.

' 260 ]

Oeceoclades calcarata (Schltr.) Garay & Taylor,
comb. nov.
Basionym: Cymbidium calcaratum Schltr. in Ann.
Mus. Col. Marseille ser. 8, 1: 181, t. 16,
19153.
Syn.: Hulophia calcarata (Schltr.) Schltr. in Fedde
Rep. Beih. 32: 262, 1925.
Type: Madagascar, Manongarivo (Ambongo). Coll.
Perrier no. 1681! (P).
Eulophia paniculata Rolfe in Gard. Chron. ser.
3, 88: 197, 1905, not O8cceoclades paniculata
Lindl.
Lissochilus paniculatus (Rolfe) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 29, 1941.
Eulophidium paniculatum (Rolfe) Summerh.
in Bull. Jard. Bot. Bruxelles 27: 399, 1957.
Type: Madagascar, without precise locality. Collec-
tor unknown. Flowered in cultivation at the
Royal Botanic Garden, Glasnevin in June
1904! (K).
Distribution: Madagascar.

The forward-projecting spur under the lip is unique in the genus.

Oeceoclades cordylinophylla (Rchb.f.) Garay &

Taylor, comb. nov.

Basionym: Eulophia cordylinophylla Rehb.f. in Flora

68: 541, 1885.

Syn.: Lissochilus cordylinophyllus (Rehb.f.) H. Perr.
in Humbert, Fl. Madag. Orch. 2: 20, 1941.
Eulophidium cordylinophyllum (Rcehb.f.) Sum-
merh. in Bull. Jard. Bot. Bruxelles 27: 3895,
1957.

Type: Comoro Islands, without proper locality. Coll.
Humblot s.n. (W).

Eulophia lokobensis H. Perr. in Bull. Soc. Bot.
Fr. 82: 158, 1985.

[ 261 }

Lissochilus lokobensis (H. Perr.) H. Perr. in
Humbert, Fl. Madag, Orch. 2: 22, 1941.
Eulophidium lokobense (H. Perr.) Summerh.
in Bull. Jard. Bot. Bruxelles 27: 8396, 1957.
Type: Madagascar, Forét de Lokobe dans I|’Ile de
Nossi-Bé. Coll. Perrier no. 19013! (P).
Distribution: Comoro Islands, Madagascar.

The actual specimen of Eulophia cordylinophylla is missing. How-
ever, there is a sheet among Humblot’s collections from the Comoro
Islands with an unpublished name by Reichenbach, which agrees well
with the original description. This specimen also is identical with the
type of FE. lokobensis from He de Nossi-Bé, near the Comoro chain.
Perrier’s description of the column-wing as being toothed is incorrect.
The margins of the column are ciliolate-hirsute.

Oeceoclades Decaryana (H. Perr.) Garay & Taylor,
comb, nov.

Basionym: Hulophia Decaryana H. Perr. in Bull. Soe.

Bot. Fr. 82: 154, 1935.

Syn.: Lissochilus Decaryanus (H. Perr.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 32, 1941.
Vulophidium Decaryanum(H.Perr.)Summerh.
in Bull. Jard. Bot. Bruxelles 27: 395, 1957.

Type: Madagascar, without proper locality. Coll.

Decary s.n. (P). Type is cultivated in Serres
du Museum d’Histoire Naturelle de Paris

(no. K 467).
Distribution: Madagascar, Mozambique, Rhodesia,

Kenya.

Judging from the assortment of specimens we have examined, the
length of the sepals varies with age. It may be separated easily from
the related O. spathulifera by the shape of the spur. Because of the

cylindrical spur, the illustrations published in Die Orchidee 18: 246,
1967, as Eulophidium spatuljferum are referable here.

Oeceoclades gracillima (Schltr.) Garay & Taylor,
comb, nov.

[ 262 ]

Basionym: Hulophia gracillima Schltr. in Ann. Mus.
Col. Marseille ser. 3, 1: 170, t. 14, 1913,
not Ridl. 1886.
Syn.: Mulophidium gracillimum Schltr. in Fedde Rep.
Beih. 88: 255, 1925, nom. nov.

Lissochilus gracilimus (Schltr.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 28, 1941.
Type: Madagascar, bassin du Besafotra, affluent de

Menavava (Boina). Coll. Perrier no. 1059! (P).
Distribution: Madagascar.

This species is closely related to O. roseovariegata, but the diffusely
paniculate inflorescence and the shape of the lip and spur readily keep
them apart.

Oeceoclades Hebdingiana (Guillaum.) Garay &
Taylor, comb. nov.
Basionym: Lissochilus Hebdingianus Guillaum. in
Bull. Mus. Nat. Hist. Nat. ser. 2, 35:
521, 1968.
Type: Madagascar, sous bois d’Anipanihy, Prove-
nance Montagnac. Flowered in cultivation in
Jardin Botanique ‘‘ Les Cedres’’. Collector
unknown! (P).
Distribution: Madagascar.

Related to O. calcarata from which it differs primarily in the shape
of the lip and not having a forward-projecting spur.

Oeceoclades lanceata (H. Perr.) Garay & Taylor,
comb. nov.
Basionym: Mulophia lanceata H. Perr. in Bull. Soe.
Bot. Fr. 82: 156, 19385.
Type: Madagascar, bois des pentes occidentales a
Manerinerina sur le Tampoketsa, entre |’ [kopa
et la Betsiboka. Coll. Perrier no. 168438! (P).
Distribution: Madagascar.

Perrier has reduced this species to a synonym of Eulophia pandu-
rata Rolfe, but the two are amply distinct from one another in floral

[ 263 ]

details, especially in the shape of the lip. Vegetatively it is reminis-
cent of O. seychellarum. The flowers are rose-colored.

Oeceoclades latifolia (Rolfe) Garay & Taylor, comb.
nov.
Basionym: Mulophia latifolia Rolfe in Bol. Soc. Broter.
9: 189, 1891.
Syn.: Mulophidium latifolium (Rolfe) Summerh. in
Bull. Jard. Bot. Bruxelles 27: 896, 1957.
Type: Island of Sio Tomé. Coll. Quintas s.n./ (KX).
Distribution: Africa—Island of Sao Tomé.
Florally the plants referable to this species are rather similar to
O. ugandae, but vegetatively they are very different and more remi-
niscent of O. atrovirens. The lateral veins of the lip are papillose-

ciliolate in this alliance to which O. pandurata, O. seychellarum and
O. lanceata also belong.

Oeceoclades lonchophylla (Rchb.f.) Garay & Tay-
lor, comb. nov.
Basionym: IZulophia lonchophylla Rchb.f. in Flora 68 :
542, 1885.
Syn.: HMulophidium lonchophyllum (Rehb.f.) Schltr. in
Fedde Rep. Beih. 83: 256, 1925.
Lissochilus lonchophyllus (Rehb.f.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 26, 1941.
Type: Comoro Islands, without proper locality. Coll.
Humblot no. 433! (P,W).
Kulophia tainioides Schltr. in Engl., Bot. Jahrb.
26: 889, 1899.
Kulophidium tainioides (Schltr.) Summerh. in
Bull. Jard. Bot. Bruxelles 27: 408, 1957.
Type: Mozambique, between Morumben and Mass-
inga, region of Inhambane. Coll. Schlechter
no. 12106! (K,Z).
Eulophia dissimilis Dyer in FI. Pl. Afr. 27: t.
1066, 1949.
Eulophidium dissimile Dyer in Fl. Pl. Afr. 27:
t. 1066, 1949, nom. altern. in obs.

[ 264 ]

Type: Mozambique, Lourenco Marques District, Le-
bombo Mountains. Coll. Daintree s.n. (PRE).
Distribution: Mozambique, Comoro Islands.

We cannot find enough distinction between F. lonchophylla and
E. tainioides to maintain them separately.

Humblot no, 433 is a mixture, containing material also referable
to O. cordylinophylla and to O. Perrieri. In the Reichenbach Her-
barium no. 6531 the original description and drawings by Reichen-
bach are mounted with a specimen with long, lorate leaves. This
specimen does not belong to O. lonchophylla. We have chosen Her-
barium Reichenbach no. 5902 as the holotype of Humblot no. 433,
O. lonchophylla, for it agrees with the specimens under that name in
Paris.

Oeceoclades Lubbersiana (De Wildem. & Laurent)
Garay & Taylor, comb. nov.
Basionym: Hulophia Lubbersiana De Wildem. &
Laurent in Rev. Hort. Belg. 26: 4, 1900.
Syn.: HMulophidium Lubbersianum (De Wildem. &
Laurent) Summerh. in Bull. Jard. Bot. Brux-
elles 27: 397, 1957.
Type: Zaire, Sankur. Coll. Laurent s.n./ (BR).
Distribution: Zaire, Uganda.
Vegetatively the plants of this species are very similar to O. atro-
virens. The shape of the lip which is wider than long, however, is

unique in the relationship to which also O. (latifolia and O. pandurata
belong.

Oeceoclades Mackenii (Rolfe ex Hemsl.) Garay &
Taylor, comb. nov.
Basionym: Hulophia Macken Rolfe ex Hemsl. in
Gard. Chron. ser. 8, 12: 5838, 1892.
Syn.: HMulophidium Macken (Rolfe ex Hemsl.)
Schltr. in Ann. Mus. Col. Marseille ser. 3, 1:
188, 19138.
Type: Natal, near Verulam. Coll. McKen no. 11! (K).
Distribution: Natal, Mozambique, Rhodesia.

Superficially the plants of this species are rather similar to O. macu-
lata. However, the lip which is shorter than wide and the branched in-
florescence afford easy recognition in both the field and the herbarium,

[ 265 ]

Oeceoclades maculata (Lind].) Lind]., Gen. and Sp.
Orch. Pl. 237, 18388.
Basionym: Angraecum maculatum Lindl., Collect.

Syn.:

Type:

Bot. t. 15, May 1821.
Limodorum maculatum WLodd., Bot. Cab. 5: t.
496, June 1821.
Aerobion maculatum (Lindl.) Spreng., Syst.
Veg. 3: 718, 1826.
Eulophia maculata (Lindl.) Rehb.f. in Walp.
Ann. 6: 647, 1868.
Eulophidium maculatum (Lindl.) Pfitz., Entw.
Nat. Anordn. Orch. 88, 1887.
Graphorchis maculata (Lind].) O. Ktze., Rev.
Gen. Pl. 2: 662, 1891.
Brazil, without proper locality. Introduced and
cultivated by Loddiges no. 84.10.16! (BM).
Geodorum pictum Link & Otto, Ic. Pl. Sel. pt.
3: 85, t. 14, July 1821.

Lectotype: Brazil, without precise locality. Received

Type:

Type:

from British Gardens and cultivated in
Berlin. (Probably part of the original in-
troduction by Loddiges). Holotype was
destroyed during World War II. The
published plate is designated here as the
Lectotype.

Eulophia Ledienit Stein ex N.E. Br. in Kew

Bull. 90, 1899.

Eulophidium Ledienit (Stein ex N.E. Br.) De

Wildem. in Ann. Mus. Congo ser. 5, 1: 115,

1904.

Zaire, without precise locality. Coll. Ledien

s.n. (WRSL).

Julophidium Warneckeanum Kral. in Engl.,

Bot. Jahrb. 33: 70, 1902.

West Africa, Togo, near Lome. Coll. Warnecke

no. 196! (KX).

[ 266 ]

Eulophidium nyassanum Schltr. in Engl., Bot.
Jahrb. 53: 598, 1915.
Lectotype: Tanzania, near Mbaka Kilambo. Coll.
Stolz no. 1909! (BM,K).

Distribution: U.S. A.—Florida, Venezuela, Colombia,
Guyana, Peru, Bolivia, Argentina, Paraguay, Brazil,
Trinidad, Bahamas, Dominican Republic, W.1., Sene-
gal, Guinee Bissau, Sierra Leone, Liberia, Ghana, Togo,
Nigeria, Sio ‘Tomé, Gabon, Zaire, Congo- Brazzaville,
Burundi, Sudan, Uganda, Tanganyika, Zanzibar, Pemba,
Zambia, Rhodesia, Angola.

The plants of all three species, O. maculata, O. monophylla, and O.
Mackenii are very similar in appearance. Yet the proportions of the
lip in all are sufficiently distinct to allow easy recognition. The lip
of O. monophylla always has a distinct elongate isthmus.

Oeceoclades maculata var. pterocarpa (Hauman)

Garay & Taylor, comb. nov.

Basionym: Hulophidium maculatum var. pterocarpum

Hauman in Anal. Mus. Hist. Nat. B.
Aires 29: 881, 1917.

Type: Argentina, Formosa, Pilaya. Coll. Kermes
no. 507. (BA).

Syn.: Mpidendrum connivens Vell., Fl. Flumin. 9: t.
44, 1851.

Type: Brazil, Santa Cruz. Coll. Vellozos.n. No speci-
men is known to exist. Vellozo’s original draw-
ing in Flora Fluminensis is regarded here as
representing the holotype.

Distribution: Argentina, Paraguay, Brazil.

This variety differs from the typical form in having winged cap-

sules. It is possible that when more material is at hand, especially
fresh flowers, this variety may prove torepresent a species sufficiently

distinct from O. maculata. There is also a certain difference in the
outline of the lip.

Oeceoclades monophylla (A. Rich.) Garay & Tay-

lor, comb. nov.

[ 267 |

Basionym: Angraecum monophyllum A. Rich, in
Mem. Soc. Hist. Nat. Paris 4: 58, t. 9,
1828.
Syn. : Hulophidium monophyllum (A. Rich.) Schltr. in
Ann. Mus. Col. Marseille, ser. 3, 1: 183, 1913.
Type: Mauritius Island, without proper locality. Coll.
Commerson s.n./ (P).

Distribution: Mascarene Islands.

The distinction between the plants of this species and those of O.
maculata have already been stated above. If the distance of articula-
tion of leaves and pseudobulb is of specific importance, as Summer-
hayes seems to have thought, then Kulophia Ledientt now included in
O. maculata, will probably be recognized on its own as a close relative

of O. monophylla.

Oeceoclades pandurata (Rolfe) Garay & Taylor,
comb, nov.

Basionym: Mulophia pandurata Rolfe in Journ. Linn.

Soc. London 29: 52, 1891.

Syn. : Lissochilus panduratus (Rolfe) H. Perr.in Hum-
bert, Fl. Madag. Orch. 2: 29, 1841.
Kulophidium panduratum (Rolfe) Summerh., in
Bull. Jard. Bot. Bruxelles 27: 899, 1957.

Type: Madagascar, near Fort Dauphin. Coll. Elhot

no. 2546! (Ix).
Distribution: Madagascar, Rhodesia.

The free, truncate lateral lobes of the lip are very characteristic
for this species. The lateral veins on the dise are papillose-ciliolate.

Oeceoclades Perrieri (Schltr. ) Garay & Taylor, comb.
nov.

Basionym: MMulophidium Perriert Schltr. in Fedde
Rep. Beih. 33: 256, 1925, not Mulo-
phidium ambongense Schltr.

Syn.: Mulophia ambongensis Schltr. in Ann. Col.
Mus. Marseille ser. 8, 1: 169, t. 18, 19138.
Lissochilus ambongensis (Schltr.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 19, 1941.

[ 268 ]

Type: Madagascar, Manongarivo (Ambongo). Coll.
Perrier no. 1654! (P).

Distribution: Madagascar, Mozambique.

The long-acuminate, lorate leaves without a distinct petiole dis-

tinguishes the plants of this species from the related O. quadriloba,
O. sclerophylla and O. analavelensis.

Oeceoclades petiolata (Schltr.) Garay & Taylor,
comb. nov.
Basionym: Mulophia petiolata Schltr. in Ann. Mus.
Col. Marseille ser. 3, 1: 175, t. 18, 1913.
Syn.: Mulophidium petiolatum (Schltr.) Schltr. in
Fedde Rep. Beih. 33: 256, 1925.
Lissochilus petiolatus (Schltr.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 25, 1941.
Type: Madagascar, Manongarivo (Ambongo). Coll.
Perrier no. 478 bis! (P).
Distribution: Madagascar.

Vegetatively the plants of this species are practically identical with
those of O. alismatophylia, but the shape of the lip is very different
proportionately. From the related O. lonchephylila it differs in having
a pandurate lip with three thickened veins in front of the calli.

Oeceoclades quadriloba (Schltr.) Garay & Taylor,
comb, nov.
Basionym: MWulophia quadriloba Schltr. in Ann. Mus.
Col. Marseille ser. 8, 1: 176, t. 12, 1918.
Syn.: Hulophidium quadrilobum (Schltr.) Schltr. in
Fedde Rep. Beih. 33: 256, 1925.
Lissochilus quadrilobus (Schlitr.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 30, 1941.
Type: Madagascar, Manongarivo (Ambongo). Coll.
Perrier no. 1696! (P).
Distribution: Madagascar, Rhodesia.

The unique structure of the lip in these plants is reminiscent of
those belonging to the Asiatic genus Grosourdya. The vesicular, hang-
ing spur is longer than the reflexed midlobe of the lip.

[ 269 |

Oeceoclades Rauhii (Sengh.) Garay & Taylor, comb.
nov.
Basionym: Mulophidium Rauhiu Sengh. in Die Orch-
idee 24: 61, 1973.
Type: Madagascar, south of Anivorano. Coll. Rauh
& Senghas no. 22865 (HEID).
Distribution: Madagascar.
Closely related to O. boinensis from which it differs in having liner-
lanceolate sepals and petals and an equally four-lobed lip.

Oeceoclades roseovariegata (Sengh.) Garay & Tay-
lor, comb. nov.
Basionym: Mulophidium roseovariegatum Sengh. in
Adansonia ser. 2, 6: 561, 1967.

Type: Madagascar, near Diégo-Suarez, ‘‘Montagne
des Francais’’. Coll. Rauh & Buchloch no.
7985 (HELD).

Distribution: Madagascar.

As it was stated above, this species is closely allied to O. gracillima.
As a matter of fact, the plants were already known to Schlechter
through a collection by Perrier no. 16224! (P), and was regarded by
him as an undescribed species. The Perrier specimen has Schlechter’s
original drawings of the floral parts attached to the sheet. It was also
collected on ‘‘Montagne des Francais’’. Perrier identified it as Eulo-
phidium gracillimum var., but cited it without a varietal name in his
Orchids of Madagascar.

The photographs of O. roseovariegata in Die Orchidee 18: 24, 1967
show the spur as being distinctly bilobed. This is apparently due to
the angle in photographing, revealing the ventrally compressed and
slightly grooved tip.

Oeceoclades Saundersiana (Rchb.f.) Garay & Tay-
lor, comb. nov.
Basionym: Hulophia Saundersiana Rehb.f. in Bot.
Zeit. 24: 378, 1866.
Syn.: Graphorchis Saundersiana (Rehb.f.) O. Ktze.,
Rev. Gen. Pl. 2: 662, 1891.
Mulophidium Saundersianum (Rehb.f.) Sum-

merh. in Bull. Jard. Bot. Bruxelles 27: 401,
1957.

Type: Africa, West Coast, without precise locality.
Coll. Mann s.n./ (W).

Lissochilus barombensis Kral. in Engl., Bot.
Jahrb. 17: 52, 1898.

Type: Cameroun, Barombi. Coll. Preuss no. 546! (Kx).
Kulophia Bierleri De Wildem., Not. Pl. Util.
Congo 1: 311, 1904.

Type: Zaire, Coquilhatville. Coll. Bierler s.n. (BR).
Hulophia Mildbraedi Wrz). in Engl., Bot.
Jahrb. 43: 339, 1909.

Type: Zaire, Ruwenzori Range, Semliki Plains. Coll.
Mildbraed no. 275! (Kx).

Distribution: Sierra Leone, Liberia, Ivory Coast,

Ghana, Nigeria, Cameroun, Gabon,
Zaire, Uganda, Kenya, Tanzania, Zam-
bia, Angola.

The long cylindrical pseudobulbs with two leaves and the equally
four-lobed lip of rather large flowers easily identify the plants of this
species.

Oeceoclades sclerophylla (Rchb.f.) Garay & Tay-
lor, comb. nov.
Basionym: MWulophia sclerophylla Rchb.f. in Flora 68:
542, 1885.
Syn. : Mulophidium sclerophyllum (Rehb. f.) Summerh.
in Bull. Jard. Bot. Bruxelles 27: 402, 1957.
Type: Madagascar, Forét d’Ankaye. Coll. Humblot
s.n.! (W).
Hulophia Elhoti Rolfe in Journ. Linn. Soc.
London 29: 52, 1891.
Lissochilus Elliot (Rolfe) H. Perr. in Hum-
bert, Fl. Madag. Orch. 2: 47, 1941.
Type: Madagascar, near Fort Dauphin. Coll. Elliot
no. 2424! (Kx).

[ 271 ]

Among the plants with long, linear leaves O. sclerophylla may be
compared with O. analavelensis, but the two-leaved pseudobulbs and
the differently proportioned lips readily separate the two. Vegeta-
tively it also resembles O. quadriloba.

Oeceoclades seychellarum (Rolfe ex Summerh.)
Garay & Taylor, comb. nov.
Basionym: Mulophia seychellarum Rolfe ex Summerh.
in Bull. Mise. Inf. Kew 3638, 1928.
Syn.: Mulophidium seychellarum (Rolfe ex Summerh. )
Summerh. in Bull. Jard. Bot. Bruxelles 27:
402, 1957.
Type: Seychelles, Mahé, Cascade Estate. Coll.
Thomasset no. 88! (K).

Distribution: Seychelles.

Vegetatively the plants of O. seychellarum are identical with those
of O. lanceata. The two may be kept apart on account of the differ-
ences in the floral structures, especially in the shape and proportion
of the lip as shown in the key.

Oeceoclades spathulifera (H. Perr.) Garay & Tay-
lor, comb. nov.

Basionym: Hulophia spathulifera H. Perr. in Bull.
Soc. Bot. Fr. 82: 157, 1935, as J. spatu-
lifera, sphalm.

Syn.: Lissochilus spathulifer (H. Perr.) H. Perr. in
Humbert, Fl. Madag. Orch. 2: 33, 1941.
Hulophidium spathuliferum (H. Perr.) Sum-
merh. in Bull. Jard. Bot. Bruxelles 27: 403,
1957.

Type: Madagascar, Ambongo-Boina. Coll. Perrier
no. 15930 (P).

Distribution: Madagascar.

The distinctly spathulate sepals and petals combined with a glo-
bose spur easily separates the plants of this species from the related
O. calcarata and O. Hebdingiana.

[ 272 ]

Oeceoclades ugandae (Rolfe) Garay & Taylor, comb.
nov.
Basionym: Hulophia ugandae Rolfe in Bull. Mise.
Inf. Kew 839, 1913.
Type: Uganda, Mabira Forest. Coll. Brown no. 443!
(Kx).
Distribution: Zaire, Ghana, Uganda.

Summerhayes considered this species to be synonymous with O.
latifolia. As a matter of fact the shape of the lip superficially looks
very much alike in both. Yet vegetatively the plants of these two
species are closer to other members than to one another. The lip of
O. ugandae is also provided with a pair of gibbosities at the junction
of the lateral and midlobes, this character is absent in O. latifolia.

Oeceoclades zanzibarica (Summerh.) Garay & Tay-
lor, comb. nov.
Basionym: Mulophidium zanzibaricum Summerh. in
Bull. Mise. Inf. Kew 417, 1927.
Type: Zanzibar, without precise locality. Coll. Last
s.n.! (KX).
Distribution: Zanzibar, Pemba.

The lanceolate, long-attenuate leaf is rather unique in the genus.
Florally the plants of this species are closest to those of O. alismato-
phylla from Madagascar, but differ from one another in the shape of
the sepals and petals and in the proportions of the lip.

[ 273 ]

Excluded Species
. faleata (Thunb. ) Lindl. = Neofinetia falcata (Thunb.) Hu

. flexuosa Lindl. =Cleisostoma ramosum (Lindl. ) Hook.f.

. funalis (Sw.) Lindl. = Dendrophylax funalis (Sw.) Benth. ex Rolfe
. gracilis (Thou.) Lind]. =Chamaeangis gracilis (Thou. ) Schltr.
javanica Teijsm. & Binn. = Hymenorchis javanica (T. & B) Schltr.
Lindleyana Regel. = Neofinetia falcata (Thunb. ) Hu

. Lindleyi Regel = Neofinetia falcata (Thunb. ) Hu

. paniculata Lindl. = Robiquetia succisa (Lindl. ) Seidenf. & Garay
. parviflora (Thou.) Lind], = Angraecopsis parviflora (Thou. ) Schltr.
. pusilla Lind]. =Saccolabiopsis pusilla (Lindl.) Seidenf. & Garay
. Retzii Lindl. =Chiloschista pusilla (Retz) Schltr.

. tenera Lindl. = Trichoglottis tenera (Lindl.) Rchb.f.

SOS on ee O46 5 5

The following list of names constitutes the Section Pulchrae Krz]l.
of the genus Eulophia (Gard. Chron. ser. 3, 22: 262, 1897). Some
of these names have been referred to the genus Eulophidium ( = Oeceo-
clades) previously.

Eulophia gracilis Lindl. in Bot. Reg. 9: t. 742, 1823.

Eulophia emarginata Bl., Fl. Java, n.s. 1: 152, 1858.

Eulophia guamensis Ames in Philipp. Journ. Sci. Bot. 9: 12, 1814.
Eulophia macrostachya Lind|,, Gen. and Sp. Orch. Pl. 183, 1833.
Eulophia megistophylla Rchb.f. in Flora 68: 379, 1885.

Eulophia minimiflora Kral. in Not. Syst. 4: 137, 1928.

Eulophia novo-ebudae Krzl. in Bull. Soe. Bot. Fr. 76: 301, 1929.
Eulophia pulchra (Thou.) Lind]., Gen. and Sp. Orch. Pl. 182, 1833.
Eulophia Rourii Kral. in Sarasin & Roux, Nova Caled. 1: 82, 1914.
Eulophia silvatica Schltr. in Engl., Bot. Jahrb. 53: 5

Eulophia striata Rolfe in Journ. Linn. Soc. London 29: 53, 1891.
Eulophia Wendlandiana Krzl. in Gard. Chron. ser. 3, 22: 262, 1897.
Lissochilus ambrensis H. Perr. in Not. Syst. 14: 159, 1951.

BOTANICAL MUSEUM LEAFLETS

HARVARD UNIVERSITY

CAMBRIDGE, MASSACHUSETTS, NOvEMBER 30, 1976 VoL. 24, No. 10

MANICARIA SACCIFERA AND ITS CULTURAL
SIGNIFICANCE AMONG THE WARAO INDIANS
OF VENEZUELA

BY
JOHANNES WILBERT

Within the tropical and subtropical belt that circles the
earth, palms can truly be counted among the best friends of
man. In several regions, the various parts of the palm are so
thoroughly exploited for purposes of food, drink, basic ma-
terials, tools, and utensils that it assumes a pivotal position
in the cultural life of the people. In South America outstand-
ing examples are Astrocaryum, Euterpe, Guilielma, Jessenia,
Mauritia, Maximiliana, Oenocarpus, Orbignya, and Syagrus.
In addition, the Indians are known to utilize the products of
at least a dozen other genera of palms.’

This paper focuses on Manicaria saccifera, the temiche
palm, and its cultural importance among the Warao Indians
of Venezuela (PLATE LXVI). Ethnobotanical information
on this genus, except for descriptions in floras, is slight, and
the fact that Manicaria saccifera is used as a source of sago
has, heretofore, gone unrecorded in the scientific literature.”

As an ancient food-quest activity of man, the recovery of
palm starch “appears to be a pantropical phenomenon that is
most highly developed on the mainland of Southeast Asia and
in the West Pacific. In that region starch extraction involves,
principally, palms of the genus M etroxylon, Sago Palms
(Burkill 1935: 1460-1462; and Barrau 1959: 151-159). Palms

[ 275 ]

PLATE LXVI

Manicaria saccifera in the forest of the Intermediate zone of the
Orinoco Delta.

[ 276 ]

of other genera are also used to a lesser extent” (Heinen and
Ruddle 1974: 116). :

In South America, the Indians recover starch from at least
four genera of palms: Syagrus, Copernicia, Mauritia, and
Manicaria, the most important being Syagrus and Mauri-
tia.” Syagrus Romanzoffianum and Copernicia cerifera are
exploited for sago by the Tupi-speaking Guayaki Indians of
Paraguay and by several Guaranian and non-Guaranian
tribes of the Gran Chaco. Mauritia flexuosa and Manicaria
saccifera are utilized for sago recovery by the Warao Indians
of the Orinoco Delta in Venezuela and in the adjacent east-
erly regions of Guyana.

Syagrus Romanzoffianum, the Paraguayan coconut or
queen palm, has a southeasterly distribution in South Amer-
ica, where it is known by its Guarani name pindo. The Gua-
yaki call the palm toi (tdi, tdi) and refer to its sago as kraku
(Cadogan 1960). Because of the many uses to which its
different parts are put, Syagrus Romanzoffianum plays an
outstanding role in the lives of the autochthonous and rural
populations of this part of the New World. In fact, for the
different Indian tribes of the area it represents probably the
most important economic plant of their environment.

The extraction of sago from Syagrus Romanzoffianum
and Copernicia cerifera is known to be practiced in South
America by such tribes as the Guayaki, Mbya (Caingua),
Kaingan, Toba, Lengua, and Chamacoco. But it is highly
probable that this practice was diffused much more widely
in earlier, pre-agricultural times. Writes Vellard (1934-35:
240-241) concerning the Guayakti:

. it is the flour extracted from the pindo palm (Cocos Roman-
zoffiana) which, along with the game and honey, constitutes their
basic diet. . . . To prepare the flour . . . the Guayaki cut open the
trunk with an axe. The fibers are pounded and crushed on the
spot with the help of an old piece of bow or with the back of
the axe, then they are roughly strained on a square frame with
loose straw or lamellas of bamboo bark. After being moistened

with water the flour is used for making cakes which are eaten raw
or dried near the fire.

[ 277 ]

Vellard (ibid., p. 240) points out that krakw starch is pre-
pared not only by the Guayaki but also by the Mbya, a sub-
group of the Guarani-speaking Caingua. They fall back on it
as an emergency food. “To obtain the starchy pith of palm
trees, the Mbya extracted the long fibers imbedded in starch
from the lower part of the trunk. They either pounded them
in a mortar and sucked them or else dried them on a platform
in the sun or over the fire, pounded them, sifted them through
a net, and then made them into loaves or cakes” (Métraux
1946a: 262). Even the Paraguayans resorted to eating kraki
“after the disastrous war against the Triple-Alliance (1866-
1870)” (Vellard ibid. p. 240).

For the Kaingin (Caingang), a non-Guarani-speaking
tribe of southern Brazil, the sago of Syagrus Romanzoffi-
anum was an important food before it was replaced by man-
ioc flour. (The Indians crushed the pith in a mortar and
sifted the flour before roasting it in a pan, just as is done now
with manioc flour.) (Métraux 1946b: 445-453. )

Métraux (1946a: 248, 261) also describes the Toba, the
Lengua, and the Chamacoco tribes of the Gran Chaco as
recovering the palm starch of Copernicia cerifera. The Toba
pound the pith in a mortar and boil it as a mush, whereas the
Lengua grate it into flour for cakes. Carandaipe starch is a
principal vegetable food for the Chamacoco,.

The best documented case by far of palm-starch extraction
for any South American Indian tribe comes from the region
of the Warao, where chroniclers, missionaries, travellers, and
anthropologists have become aware of its existence and
where, on numerous occasions over the past twenty years of
intermittent field work, I witnessed at first hand the process
of recovery of sago from Mauritia.”. The ethnobotanical data
available on this palm are too abundant to be treated here.
Suffice it to say that Mauritia sago, ohidu aru, has been the
staple food for most Warao until very recently, when it was
supplanted by ocumo (Xanthosoma sagittifolium) and, to a
lesser extent, by manioc (Manihot esculenta).

The practice of extracting sago from Manicaria saccifera

[ 278 ]

(yahuhi aru) came to my attention only recently, and I
was able to witness the procedure for the first time in the
summer of 1975.

In a process that may have taken more than a thousand
years, the Warao have adapted their life and culture to the
rather difficult outer world of the Orinoco Delta, situated in
eastern Venezuela between 8° and 10° north latitude and
between 59° and 62° west longitude. The Delta is a large
fan of alluvial deposits occupying 17,000 square kilometers
and bounded on the south by the Orinoco River proper and
on the west by the Manamo which branches off the main
river at Barrancas, where the apex of the Delta is situated.
Most of this low-lying triangle is a tidal swamp lacking in
dry ground and stone and extending inland from 50 to 100
kilometers (Liddle 1928: 20-24).

Manicaria does not grow throughout this vast area but is
restricted mainly to regions immediately within the Inter-
mediate Delta zone behind the Lower Delta, most of which
is covered by a belt of mangrove. Typical of such environ-
ments, the soil is almost always inundated, thus providing
along the perimeter of the swamp the ideal condition for the
pioneering Rhizophora Mangle. The feature characteristic
of mangrove forests of growing on the periphery while dying
at the core is of primary importance to the Warao. The clear-
ing that results from the decomposing inner parts of the
mangrove forest is invaded by many kinds of trees important
to the Indians, among them useful palms such as the Mau-
ritia, Euterpe, and Manicaria. Thus, within and behind the
coastal mangrove belt of the Lower and Intermediate Delta
zones there developed the echonich of a palmetum that has
amply served the Warao as an abundant food basket and as a
secure home,

The average mean temperature of the Delta is 26° C, the
humidity 60 to 80 percent. The rainy season lasts roughly
from May to October and the dry season from November to
April, but rainy days with more or less intensive showers oc-
cur throughout the year. The annual precipitation ranges

[ 279 ]

from 100 to 200 centimeters (Heinen and Lavandero 1973:
4-11),

Twice daily the tide washes over the palmetum, encourag-
ing the growth of the trees. The annual flooding of the Ori-
noco is felt only indirectly in the Intermediate Delta; the
flooding of the palm groves during the wet season is actually
due to rainfall. In the dry season, when the waters of the
Orinoco recede, sea water penetrates into the Intermediate
Delta, salinates the rivers, and temporarily causes a potable
water problem. Within the palmetum, Manicaria appears to
seek out not only places under the influence of the tides but
also those exposed to the northeasterly trade winds that
sweep over the Delta almost incessantly.

Warao culture is particularly adapted to life in this palm-
rich environment. For the pre-agricultural Warao, the sago
primarily of Mauritia and secondarily of Manicaria provided
the staple food as well as a superabundance of edible fruits
during much of the year. The Euterpe (E. edulis), too
yielded fruit and an especially tasty and rich palmito. Mani-
caria milk and Mauritia (unfermented) wine helped solve
any drinking water problem, and the fat grubs of the palm
borer (Ryncophorus palmetum), collected in overwhelming
quantities from fallen Mauritia and Manicaria, was an impor-
tant added source of protein to their diet. That the palms
are a blessing is clearly recognized by the Warao, and these
plants permeate their entire culture: its technological, socio-
economic, and religious systems. The Indians refer to Mau-
ritia sago, especially in combination with fish, as the “true
food” of man; Manicaria, to a lesser degree, falls into the
same category. Sago was more than a vital source of human
sustenance; it came to be elevated to a position of ritual
significance which has helped the Warao to cope psycho-
logically with the hardships imposed upon them by a refuge
environment little amenable to human life and culture.

Other palms found in the Orinoco Delta, but mostly out-
side the palmetum of the Lower and Intermediate zones, are
Astrocaryum, Jessenia, Maximiliana, Socratea and others.

[ 280 ]

The study of Manicaria began in 1791, when the German
botanist Joseph Gaertner (1791: 468, Pl. 2) founded the
genus on the basis of the spathe, spadix, rachilla, and flowers
of the plant (PLare LXVII). Apparently, its trunk, foliage,
and fruit remained largely unknown to the scientific world
until sixty-two years later, when Wallace (1853: Plates II
and XXVI) furnished his illustration of this “unique and
handsome palm” (PLate LXVIII). Wallace and other field
workers have added much information to the original de-
scription,’ although the genus, its speciation and distribution
remain relatively little understood even today.’

Standley and Steyermark (1958: 271) have published a
concise summary of our present knowledge on the genus:

MANIcARIA Gaertner

Reference: Burret, Notizbl. Bot. Gart. Berlin 11: 389. 1928.

Plants very robust, tall, or low, unarmed, the caudex stout,
annulate, often curved or flexuous, covered with old leaf sheaths;
leaves terminal, very large, suberect, lanceolate, acute, plicate-
nerved, serrate at first and finally pinnatisect, the costa thick and
stout, the petiole slender, the sheath cleft, its margins with many
coarse fibers; spadices several, erect-spreading, tomentose, the
branches strict, rather thick, foveolate; spathes 2, the upper fusi-
form, terete, mucronate, fibrous, tardily rupturing; bracts subulate;
flowers monoecious, borne in the same spadix, this inserted among
the leaves, simply branched, the flowers immersed in pits in the
branches, the upper ones staminate, crowded, the lower ones
scattered, pistillate; staminate flowers oblong trigonous, the sepals
ovate-rounded, coriaceous, with scarious margins, imbricate, the
petals thick-coriaceous, obovate-oblong, valvate; stamens 24-30,
the filaments filiform, connate at the base, the anthers narrowly
linear, erect, bifid at the base, emarginate; pistillate flowers larger,
ovoid, the perianth little enlarged after anthesis, the sepals
rounded, their margins finally lacerate, broadly imbricate, the
petals longer, covolute-imbricate at the base, acute and valvate
at the apex; ovary sulcate, 3-celled, the stigmas 3, sessile; fruit
large, globose, 1-seeded, or depressed-globose and 2-3-seeded,
the stigmas terminal, the pericarp corticate, the cortex corky, an-
gulate-echinate, the endocarp vitreous-crustaceous, fibrous within;

[ 281 ]

PLATE LXVII

Spathe, spadix, rachilla, and flowers that served Gaertner to estab-
lish the genus Manicaria, 1791.

[ 282 ]

PLATE LXVIII

Teod A Vest ng

Waa

MANICARIA SACCIFERA Ht 40F

1853.

>

First depiction by Wallace of Manicaria

[ 283 ]

seeds globose, erect, the hilum oblou.g, the testa very hard, the
branches of the raphe closely reticulate and involving the seed,
the endosperm corneous, uniform, the embryo basal.

The genus has become known from Central America
(Bailey 1943: 392-393; Standley and Steyermark 1958: 271—
273), Trinidad (Bailey 1933: 409-413) and northern South
America: from Colombia, the Orinoco Delta, the Guianas
( Wessels Boer 1965a, b) to the mouth of the Amazon ( Burret
1928: 389). It also occurs on the Rio Negro and the Upper
Amazon ( Wallace 1853: 70).

A very complete description of the species Manicaria
saccifera was written by Wessels Boer (1965a: 21) PLATE
LXIX:

Trunk solitary, up to 6 m. tall, about 3 dm. in diameter, in the
upper part covered with dead leaf-bases, at base with prominent
leaf-scars. About 10 contemporaneous suberect leaves; dead
leaves persistent for some time and hanging down on the trunk;
sheath with fibrous ventral part enclosing young leaves about 7
dm. long; petiole stout, about 12 dm. long and 8 cm. in diameter,
grooved, leaf-blades very large, simple or irregularly pinnatisect
through the action of the wind, also in leaves of juvenile plants,
up to 72 m. long and 23 dm. wide, bifid at apex, margin serrate;
about 120 primary veins, 3-4 cm. distant at the middle of the
blade, secondary veins inconspicuous; petiole, costa, and the
lower surface of the blade at first more or less brown-tomentose,
soon glabrescent.

Spadices almost erect, about 17 dm. long, with 2 spathes; outer
spathe about 7 dm. long, flattened, invisible between leaf-sheaths,
inner spathe about 11 dm. long, fusiform, mucronate, consisting
of densely interwoven fibers without any suture, enclosing the in-
florescence completely till long after anthesis; peduncle about 10
dm. long, rachis about 6 dm. long with up to 45 simple rachillas
or rarely a few rachillas bifurcate, several large bracts along the
peduncle within the inner spathe, smaller bracts at the base of
the rachillas. Male flowers densely crowded in the upper part of
the rachillas (and 2 laterally adjacent to each female flower),
sunken in small pits and subtended by bracts 7-12 mm. long;
sepals ovate, imbricate, 3-4 mm. long, petals lanceolate, valvate,
ligneous-incrassate, 6-7 mm. long; stamens many (20-34),
densely congested, filaments about 1% mm. long, anthers 3 mm.
long, the central ones usually misshapen. Female flowers few,

[ 284 ]

PLATE LXIX

Depiction in Standley and Steyermark (1958: Fig. 45) of details of
leaf, flower, and spathe of Manicaria saccifera.
(Courtesy Cuicaco Natura History Museum)

[ 285 ]

near the base of the rachillas between 2 male flowers, subtended
by bracts; sepals ovate, imbricate, 7 mm. long, 8 mm. wide, petals
ovate, valvate, acute at apex, ligneous-incrassate, 10 mm. long,
6 mm. wide; pistil globose, 4 mm. in diameter, stigma 3-fid, sessile,
erect, 3-4 mm. long, strongly papillose. Fruit depressed-globose,
3-seeded or by abortion 1- or 2-seeded; exocarp corky, angulate-
echinate, tubercles rather hard and not easily rubbed off; seed
globose, about 4 cm, in diameter; embryo basal . . .

In this account of the genus, the most recent, Wessels Boer
goes on to identify four species of Manicaria, very con-
spicuous and abundant swamp plants. The Manicaria occur
in forests interspersed with other trees but also form colonies,
or temichales, of great density. Manicaria saccifera is known
in Venezuela as temiche (timiche, timiti), an Arawakan
noun; in Guyana as truli, a Cariban term; and in Brazil as
ubussti, a Tupian word meaning “big leaves” (Civrieux 1957:
195-232). The Warao refer to it poetically as yahuhi
(yawihi) meaning “plumes of the sun,” descriptive of the
leaves that look like giant bird feathers.

The Origin Myth of the Temiche Palm

Long ago there was an old woman who followed her
husband to live in the lowlands of the Delta. Life was easier
for the old couple there than where they had come from,
and they greatly enjoyed the cool water of the bogland and
the sea breeze that incessantly fanned their new home. The
name of the old woman was Yahuhi.

As time went by, the woman felt a strange transformation
taking place all over her body. First her eyes began to clear
up so that she could see well again. Then the wrinkles in
her face disappeared, her body firmed up like that of a girl,
and her complexion became healthy and youthful looking.
Even the voice of the old woman changed back to that of a
maiden, and she began to sing with happiness over her re-
gained youth.

The husband of the woman was equally taken by surprise

[ 286 ]

and wondered what the cause of this miraculous transforma-
tion might have been.

“I have changed because my body was exposed to the
cool northwind,” said the woman. “Let’s remain here for-
ever.”

One day, the man told his youthful wife that he had to
leave the house for a short while. He wanted to go to the
field and do some gardening.

“Fine,” said the woman. “I shall cook dinner and wait for
you. But don't be later than you said.”

Time passed, and the man failed to return within the set
period of time. While his wife was waiting, she suddenly
saw a handsome young man approaching the house from the
North. He greeted her kindly and wanted to know where
her husband had gone. “He has gone to the field and is
overdue.”

So then the visitor took advantage of the man’s absence
and seduced the youthful woman.

When the husband finally got home, the suitor had long
since left the house, but the husband noticed that something
was wrong. Questioning his wife, she finally confessed that
a young man had visited and embraced her during his
absence. It was his fault, she said. He had left her alone
for so long.

This made the old man very angry. He prepared a rope
and whipped the poor woman so mercilessly that marks
began covering her body from head to foot. Days went by,
and the woman became very ill. Because of the pain, she
could neither sleep nor eat and, finally, she died.

Since there were no people living in the neighborhood who
might have helped the old man bury his wife, the husband
tied the corpse to a pole in an upright position. It looked as
if the woman was just standing there, alive. And even after
a full moon had passed, the woman’s body still looked youth-
ful and uncorrupted.

After that, however, the dead woman began to change into
a tree. The husband left her, and upon looking back a final

[ 287 ]

time, saw that his wife had become a temiche palm. He said
to himself: “Once the Warao will come to be on this earth,
they will have to call this palm yawuhi, because that was
her name as a woman.”

The etiological intent of this simple narrative is clear: a
prolific palm bearing fruit practic sally continuously is iden-
tified with a fertile young woman. Her fertility is mirac-
ulously caused by the iad of the north, whence also her
youthful lover puts in an appearance.

For a Warao listener, the introduction of the northwind
heralds doom. While the cool nortes in the Delta are certainly
invigorating winds, they also bring catarrh and other res-
piratory ailments to the Warao. Furthermore, they blow
from the direction where Haburi, the culture hero, lives in
a world mountain-tree. As a youth, he had unwittingly
seduced his own mother. The association of Haburi with
the woman’s paramour from the north is most certainly not
lost on a Warao listener. In addition, Haburi himself had
been made miraculously youthful and mature through the
agent of an old frog-woman, who wanted him for a lover.
Both the youthful temiche woman and Haburi were placed
into a fateful triangle by virtue of their newly acquired
sexual prowess. Adultery and incest were the inevitable
consequences, since the partners of both the temiche-maiden
and Haburi were infertile and no match. Consequently, the
woman transformed into a palm as did Haburi into a world
tree,

This sacrifice of metamorphosis results in enormous bene-
fits for mankind: from the transformed woman, the ever
producing temiche palm; and from the transformed hero,
the dugout canoe, a sine qua non of human life in the Delta.
So while the Warao listener can predict the tragedy that will
inevitably result from the action of the dramatis personae,
he can also anticipate the great benefits that will accrue to
him from this primordial drama of the temiche-maiden, the
old husband, and her paramour from the North.

The myth also explains why the temiche favors coastal

[ 288 ]

swamps with their tidewaters and sea breeze. Besides ex-
plaining the remarkable fertility of the palm, subject not to
an annual flowering and fruiting season but to continuous
yield and the swamp habitat of the temiche, the myth also
offers an explanation for the exceptionally prominent leaf
scars (yd esoara) that cover the entire stem of the palm.
Similar explanations are given in Warao mythology for the
rings of Euterpe and of trees like Calophyllum. Finally, as
we shall see, different parts of the palm are used as med cue
against respiratory illnesses. This blissful property of the
plant may possibly find its explanation in the love of the
tree-maiden for the nortes that commonly cause such ills.

Utilization of the Palm

The leaves. It is well to commence the discussion of the
cultural significance of Manicaria for the Warao Indians with
the plant’s most outstanding characteristic, its leaves. The
“plumes of the sun,” yahuhi, as the Warao call them, are
the largest entire leaf among palms and the lar gest in the
plant kingdom. The palm seudied for purposes of this paper
had seventeen contemporaneous sub-erect leaves with two
persistent dead leaves hanging down. The informant hap-
pened to know that the palm was between 30 and 35 years
old. Along their entire axis, large leaves of the Manicaria
saccifera measured 5-8 m. long ae 1.5-1.8 m. wide, with
petioles measuring 1.20-2 m. Spiaaa (1968: 111) reports
leaves 9 m. in length, and leaves of 10 m. are frequently
mentioned in the literature. Through exposure to wind, the
blades of large outer leaves eal to become irregularly
pinnatisect, but younger inner blades remain undivided.

It is precisely this latter quality coupled with their size
that make Manicaria leaves especially suited for house thatch
wherever the plant grows (PLare LXX). As Im Thurn
(1967 [1883]: 209) observed, “each gigantic undivided leaf
of the troolie palm ( Manicaria saccifera) is really a shelter in
itself; and a few of these laid, without further preparation,

[ 289 ]

PLATE LXX

Warao houses with Manicaria thatch. Mauritia flexuosa in the
background. (Courtesy P. T. Furst)

[ 290 ]

so as to overlap like tiles, make a most perfect roof. Indeed,
before corrugated zinc was introduced for the purpose, a
large trade was carried on between the Indians and the
planters on the coast in these troolie-leaves, with which most
of the buildings on the sugar estates were thatched.”

For weather protection, the Warao implant single leaves
or a whole line of them into the soft ground near their work-
ing areas. They also cover their heads with leaf segments
when traveling on foot or by boat, calling these makeshift
umbrellas aroko a kuasimara, leaf capes.

To thatch their houses, the Indians fold the leaves in half
along the rachis and lash them in overlapping fashion, each
rib 25 cm. from the next, vertically onto the infra-structure
of the roof, “so that each frond forms a long tile reaching
from ridge to eaves” (Spruce 1908, 1: 59) (PLarE LXXI1).
The house of my main informant had been covered in this
way in 1969 and did not begin to leak until 1975, testimony
of the durability of temiche thatch. To keep the rain from
drifting in with the sea breeze, the Warao install a screen of
temiche (dara yawihi) on the windward side of their other-
wise wall-less houses * (PLare LXXII). Sometimes they
construct a tunnel-like roof over the midsection of their dug-
out canoes to protect themselves from the weather during
long journeys.

An ingenious naval invention is making sails (yawihi
wera) from large Manicaria leaves. On the open windswept
canos of the Delta, Warao canoes go by at high speed under
full sail. Two or three crew members each hold up a
Manicaria leaf for sails, bracing it at the bottom against the
foot and holding it with one arm (PLATE LXXIII). A helms-
man keeps the course by means of a paddle held vertically
as a rudder. I have clocked canoes 6 m. long with two
paddlers but no sails going 3 kph, their full speed. Canoes
with temiche “sails” go that fast, or faster, and, of course,
for a longer period of time. Consequently, in terms of
primitive navigation, the yawihi wera of the Warao repre-
sents critical navigational tackle. Despite its Spanish-derived

[ 291 |

[ 26a |

Detail of roofing of a Warao house. (Courtesy P. T. Furst)

IXX'] 4LVId

PLATE LXXII

Detail of weather screen on the windward side of a Warao house.
(Courtesy P. T. Fursr)

[ 293 ]

| £62 |

Three canoes with some fourteen passengers and a heavy load of fire wood are being propelled by

holding up three leaves of Manicaria saccifera. Warao Indians of the Intermediate zone ‘of the Orinoco
— Ce Oe ee > i pe en |

HIXX'] 2LV 1d

name (wera=vela, sail), I wonder whether, in view of its
simplicity, sailing by yawihi wera does not antedate the
arrival of the white man in northeastern South America.
Since Manicaria is ubiquitous in the Lower and Intermediate
Delta zones, a “sail” can be picked up at practically any point
of departure and simply tossed away upon arrival at the
destination. In other words, Manicaria leaf-sails are handy,
free, and uncomplicated.

Manicaria leaves, or certain parts, are put to other uses
by the Warao Indians. The tips of the leaves are improvised
for use as fans (yami); several, 2 m. long pieces of rachises
of leaves are tied together in the form of a Venetian blind to
serve as fish weirs (noba); sections of midrib are rubbed
together to produce fire by rotation (Im Thurn 1967: 257).
The “plumes of the sun” represent a materia prima of great
importance to Warao technology.

A final comment on the etymology of the Warao term
“plumes of the sun.” It is derived from far more than the
shape of leaves that flicker in the sun (as some authors seem
to suggest). The Manicaria leaves obviously resemble over-
sized bird feathers; but why they are linked to the sun is
less evident. Manicaria, like many other palms, are closely
connected with a symbolism of light and darkness, day and
night, and I discuss this aspect below.

The spathe. Next in importance are the pouch-like spathes
that cover the entire inflorescence and the large pendant
infructescence (PLATE LXXIV). The brown spathes from
which the genus and the species derive their names ° are from
40 to 60 cm. long, “of fine, closely woven texture, and are
used by the natives to make soft brown caps without seams
or joinings” (McCurrah 1960: 129). The Warao refer to
these hood-like caps as yasi nona. After carefully peeling
the spathe off the fruit cluster, the Indians wet and stretch
it on the head to give it the desired fit (Appun 1871: 479;
Wilbert 1963: 9). By pleating them and decorating them
with bast ribbons the long-peaked caps sometimes acquire
a bizarre appearance (PLates LXXV-LXXVII). They are

[ 295 ]

LXXIV

PLATE

[ 296 ]

PuaTtE LXXV

Hats made from spathe of Manicaria saccifera.
Drawn by Hexica Ap1BI

[ 297 ]

PLATE LXXVI

Warao Indian with hat made from spathe of Manicaria saccifera.
(Courtesy P. T. Furst)

[ 298 ]

PLATE LXXVII

[ 299 ]

Warao Indian (on the left) with hat made from spathe of Manicaria saccifera.

(Courtesy P. T. Furst)

worn in many shapes and forms by natives, Creoles, and
tourists in Mesoamerica (Standley and Steyermark 1958:
273) and in Brazil (Wallace 1853: 70). I have not seen
them used by the Warao as bags and wrapping cloth (ibid. )
or as loin cloths (Braun 1968: 111).

The fruit. The fruits (yawihi aukwaha) serve the Warao
as food and drink. One specimen of infructescence that I
examined weighed 12 kg; the clear liquid inside the nuts
accounted for approximately one-third, i.e., 3.75 liters (PLATE
LXXVIII). The entire bunch was 87.5 cm. long and 25 cm.
wide and was composed of 67 fruits: 22 one-seeded, 25 two-
seeded, and 20 three-seeded ones.'° The single fruits mea-
sured between 7 and 8 cm. in diameter, twins and triplets
9.5 cm. All a Warao needs do to satisfy his thirst, whether
in the jungle, along the coast, or in the field, is to cut just
one bunch of temiche fruit and drink his fill. In times of
non-potable water, the Indians take along a load of fruit
clusters in their dugouts (Turrado Moreno 1945: 92). To
get at the water, the Indian bites into the corky exocarp
(nakoro ahoro) or cuts a hole in it and drinks the water
from a calabash or from its natural cup. The fleshy homo-
geneous endosperm of the as yet immature fruit (nohi) is
greatly relished by the Indians; the fruit is opened with a
blow of a heavy stick and the jelly-like substance scraped
out with the thumbnail. Fully matured fruits, which re-
semble miniature coconuts (ya umo) fall to the ground.
They get buried in the detritus or are washed out to the
rivers and the sea. In this state, they are hard and inedible,
but find use as whorls in the manufacture of toy tops for
boys (PLare LXXIX).

The seedlings. Seedlings (emukohoko) are sought after
by the Indians, young and old. If the germination results
obtained by Braun (1968: 54) for cultivated Manicaria
saccifera are any indication, the seeds of this palm germinate
in four months, more or less. The Indians searah for seedlings
that are developed enough to have grown their characteristic
bifid eophylls up to 30 cm. above the ground.'' They pull

[ 300 ]

LXXVIII

4
vi

PLATE

LYaaTI AA SANNVHO[ hq ydvssojoyg
me)

OUIDS

ayonijur pue oyyeds Suni

cI

Pa

[ 301 ]

PLATE LXXIX

Top made of Manicaria seed. Drawn by Heica AprBI

[ 302 ]

them out and crack the hard shell of the seed with a heavy
stick, exposing the white spongy haustorium. It has a mildly
sweet taste, and as many as twenty at a time can be con-
sumed without fear of digestive complications. Only children
who eat primarily a diet of ocumo and fish are said to suffer
occasionally from diarrhea.

Ethnomedicine. Several parts are believed to have me-
dicinal properties and are used as remedies against catarrh:
that is, cough and fever; and against symptoms of other
respiratory problems. An Indian woman prepared some
medicines in my presence, so that I can verify the following
recipes — if not attest to their effectiveness.

To alleviate coughing and to repress fever, the water of
Manicaria fruit is carefully strained through a cloth to
eliminate impurities. The patient drinks a cup three times
a day.

Fever and cough are alleviated with a potion made of
the green juices from the fresh eophylls of Mauritia mixed
with Manicaria water. Added to the slimy chlorophyllous
liquid is the urine of a child of opposite sex of the patient.
The preparation is set aside for forty-eight hours and then
applied to the patient’s body, especially his temples and
forehead. The treatment is repeated three times a day. I
found the remedy to have a definite cooling, hence soothing
effect. Small children are washed in the liquid from head to
foot to stop diarrhea accompanied by fever. Occasionally,
they are also given a small quantity to imbibe.

The anti-asthmatic and anti-catarrhal effect of the liquid
endosperm of Manicaria has been mentioned by previous
investigators. Since asthma does not seem to afflict the
Warao, I could not verify this assertion, although all in-
formants assured me that the fruit water of the palm
facilitates breathing in congested patients. Most effective
in this respect is a concoction made of grated Manicaria
palmito mixed with fruit water. The remedy is imbibed.

Indirect benefits. An indirect benefit accruing to the
Indians from Manicaria relates to the fact that the ripe,

[ 303 ]

fallen fruit attract certain mammals like deer, peccary, paca,
monkey, and agouti; turkey birds, like the pauji, are also
fond of them. Traditionally, however, the Warao avoid the
large species of mammal, like deer and peccary, and refrain
from eating monkey altogether; but they like agouti, paca,
and pauji. The agouti and the paca are ambushed at sunrise
by the hunter ice hides in the vicinity of the palm. The
birds are best shot in the early afternoon hours, when they
give themselves away by the crashing noises that they pro-
duce moving around among the leaves of the plant.

I have mentioned the important protein supplement to
the diet of these Indians provided by the rich supply of
fatty larvae that are laid in Mauritia and in Manicaria by the
palm borer

amen of sago. In the summer of 1974, while collecting
data on Warao ethnomedicine in the forests of the Orinoco
Delta, one of my informants volunteered that, in addition to
providing the Indians with a remedy against fever, catarrh,
and diarrhea, Manicaria also produced a starch similar to
that from the Mauritia. He added that the extraction of
temiche sago had fallen out of practice, but that it had
formerly hece common, Realizing that when a Warao speaks
of palm starch he knows what he is talking about, I asked
him to produce some temiche starch for me when I returned
the following summer. What follows is a report on the
process of sago recovery from Manicaria saccifera as wit-
nessed by me in August 1975. PLatres LXXX-XCVIII
illustrate the process.

Throughout an area of dense pluvial forest scattered in-
dividuals of Manicaria grew at distances from each other of
4to6m. Their trunks (akabaho) were mostly erect and of
varying heights, from 2 m. to 8 m. They were dark brown
with very prominent circular leaf scars 2 to 3 cm. thick.
Most of the palms seemed to be growing on small mounds 30
to 50 cm. high, but their roots (ya ahokonamu) were not
exposed, or only minimally so.

After felling the palm, the Indian established how. far

[ 304 |

PLATE XX.

Felling the Manicaria saccifera. Photograph by JOHANNES WILBERT

[ 305 ]

PLATE LXXXI

[ 306 ]

Removing the bark for sago extraction.

Photograph by JoHANNES WILBERT

[ 208 ]

Making a trough from a section of Mauritia flexuosa.

Photograph by JOHANNES WILBERT

IIXXX'T ALVId

down from the crown it contained sago. The testing was
done by driving an axe into the trunk at various intervals.
When the axe was withdrawn, with starch sticking to its
blade, it signified contact with starchy pith.

According to the Warao, Manicaria, unlike Mauritia, has
no annual flowering and fruiting period and contains starch
the year round. The specimen used for the experiment con-
tained sago in the upper 3 m. of the 6 m. long and 30 cm.
thick trunk. To obtain for me a unit measurement of the
volume of sago in one palm, the Indian removed with his
axe the bark of a 1 m. long section below the crown, exposing
in the opening a beige to light brown fibrous interior, not
pithy like Mauritia but somewhat ligneous.

The Indian then stood on top of the trunk and shredded
the pith with an adze or hoe (nahuru).’ The hoe is a
composite tool which the Warao claim to have adopted, in
remote times, from cannibalistic neighbors, called Siawani.
It consists of three basic parts — blade, handle, and binding.
The blade (nahuru ateho) is carved from the bark of a
mature Mauritia and is 3 cm. thick. Its length varies between
40 and 60 cm. according to the height of its user. The work-
ing end (ahi) of the wooden blade is about 15 cm. wide and
grooved to form a double cutting edge. Laterally, the blade
is carved concavely and provided with notched shoulders
(arokuaha) near the end opposite the cutting edge to facil-
itate securing the blade to the cleft end of the handle.

The handle (aka) is a round piece of wood 3 cm. thick.
Any hard wood will do, and the length of the handle is
roughly equal to the length of the blade. A cleft is made in
one end, into which the non-cutting end of the blade is
firmly wedged; the junction is lashed together with two-ply
cordage made of Mauritia bast. A second string of this kind
(ahutu) connects the blade with the handle like the crossbar
of the letter A. To prevent this binding from slipping, two
notches (iwiri) are made on the sides of the blade about
20 cm. below the cutting edge (PLates LXXXIV-LXXXVI ).

The Indian had made a new hoe the day before. As it

[ 308 ]

XXXII

~
4
wv)

PLATI

ro)

INAGTI AA SANNVHO[ hq ydvss0j0Yg

‘psonxayf DiyunDW

fO YUNA oy} jo UOT{IIS wolf ogperul ysno.y pode

ys

A

[ 309 ]

PLATE LXXXIV

Using a hoe to crush the starch. Photograph by JOHANNES WILBERT

[ 310 ]

PLATE LXXXV

The opened Manicaria saccifera is lying on top of leaves to prevent
the shredded pith from falling on the swampy soil. Close-up of the hoe
used in the process of crushing the pith.

Photograph by JOHANNES WILBERT

[311 ]

turned out, inexperience with extracting sago from Manicaria
prompted him to make a cutting edge as wide as that used
for extracting starch from Mauritia, which has a much larger
trunk. The space between the hard bark on either side of
the opening measured only 26 cm., much less than the trunk
of a Mauritia customarily utilized for starch recovery. Thus,
the edge proved to be too wide to be efficiently used. The
Warao hoe bears close resemblance to sago hoes used in
southeast New Guinea for the same purpose (Stohr, 1972,
Fig. 35).

After the pith was crushed, a woman washed it in a trough
made from a piece of the trunk of a Mauritia 1.25 m. long.
About 25 cm. at either end of the trough (canoa arua) was
left untouched and the centre section excavated by means
of an axe, so that a cross section was V-shaped (PLATE
LXXXIII).

After the men had placed the trough in a north/south
direction (it must never be in the direction of the course of
the sun), the woman drove four 1.20 m. long petiole sections
of Mauritia (namoru) halfway into the water-logged ground
next to the hollowed-out Manicaria (PLareE LXXXVIIT).
Two uprights 75cm. apart stood on either side of the trough,
and on top of these uprights the woman placed a dish-like
strainer (bihi) made of strips of Ischnosiphon. Immediately
below the strainer she positioned, at a steep angle, the fleshy
end of a Mauritia leaf stem (wate buaka) for the purpose of
collecting the washed pith below the strainer and channeling
it smoothly into the trough without splashing.

The woman collected the crushed pith from the trough,
transferred it onto the strainer by means of a calabash
(Cresentia Cujete), poured water over the pith, and began
kneading it. From time to time, she scooped out some water
from the trough and poured it over her hands and the pith.
The water turned milky in the process, and the sago began
settling on the bottom of the trough (PLAtes LXXXIX-—
XCIII).

The moment the woman started kneading the pith, she

[312 |

PrATE: LAXKV!

Warao hoe used in palm starch extraction from both Mauritia
flexuosa and Manicaria sacciferd. Drawn by HELGA ADIBI

[ 313 ]

[ FI8 |

Trough with sieve resting on four uprights ready for the washing of Manicaria sac-
cifera pith extracted from the palm in the foreground of the picture. Notice fleshy
end of a Mauritia leaf stem below the strainer to channel the water into the trough
without splashing. Photograph by JoHANNES WILBERT

ALVId

IAXXX'I

[ cT¢ ]

pl

Trough filled with starch containing water after the woman finished washing the
th. Photogrph by JOHANNES WILBERT

ITAXXX'T S3LVTd

[ 9T¢ ]

Woman collecting shredded Manicaria saccifera pith into her calabash.
Photograph by JOHANNES WILBERT

LLVIg

r.
»
e

TXXX'T

PLATE XC

Calabash filled with shredded Manicaria saccifera pith resting on
palm from which it has been extracted.
Photograph by JOHANNES WILBERT

[ 317 ]

observed to me that it was rich in sago which, she said, one
is able to determine by the viscid quality of the pith (PLATE
XCIV). She observed, too, that it felt exactly like processing
sago-rich Mauritia pith, except that she had to exert more
pressure kneading the pith of Manicaria, because it was
more ligneous than the other. Another difference between
processing the two starches was that she had to wait longer
for the Manicaria starch to settle at the bottom of the trough.
She said that the resulting meal was of a lighter quality.
After waiting for ten minutes in the shade of a windscreen
(made of two temiche leaves stuck in the ground), the
woman began to drain the water carefully by ladling it out
of the trough with her calabash (PLate XCV). She then
shaped the meal into a ball of light brown sago and, in doing
so, proved conclusively that Manicaria saccifera must be
counted among the sago-producing genera of the palms
(PLATE XCVI).

The total process of crushing, washing, and collecting the
sago took approximately thirty working minutes.

Upon completion of her work, the woman painstakingly
washed her strainer to remove all particles from between
the basketry strands (PLare XCVII). This act would pre-
vent the strainer from rotting. Next, she pulled the four
uprights out of the ground and tossed them to the side
(PLate XCVIII). This would prevent malignant shamans
from blowing on them, causing her arms to hurt. Finally,
the man picked up the heavy trough and carried it a short
distance away. This would make it more difficult for
youngsters to poke inside with their machetes. People after
mischief might do this to afflict the sago washer with ailing
arms and shoulders.

From the opening in the trunk, which was | m. long, the
woman collected two calabashes packed to the brim with
pith. The calabash used was 25 cm. long, 17.5 cm. wide, and
12 cm. deep. From that much pith, the woman washed out
750 gr. of sago. It had been ascertained that the trunk of the
felled Manicaria contained starch within 3 m. of its upper

[ 318 ]

half, the total sago yield from that plant would amount to
2.250 kg. I suggest adding at least another 750 gr. to account
for the fact that the wide blade of the hoe prevented the
Indian from extracting all of the available pith. A close
approximation of the potential yield of the tested Manicaria
is 3 kg. Even at that, Manicaria must be considered a low
yield sago palm for the purpose to which the Warao put it:
emergency food."

In order to assess the nutritional value of Manicaria starch,
I had the procured sample analyzed at laboratories in Cara-
cas. The results of these tests are summarized in Table I.

TABLE I
Composition of Manicaria saccifera Starch

pH (sol. 2% ) 5.75
Humidity 63.51% 63.51%
Fat * 0.55%
Dextrose * 5.07%
Protein °* 1.62%
Starch * 4.57%
Fiber ** 24.68%

36.49% 36.49%

100.00%

* Calculation based on dry material
°* By balance

Ethnobotanical lore. The process of producing starch from
Mauritia is a highly ritualized affair when done in prepara-
tion for the annual harvest festival nahanamu or in connec-
tion with any other propitiatory offering. The implements
used in Manicaria starch extraction are also subject to certain
taboos, but, for several reasons, Manicaria sago is considered
less appropriate for these sacrificial purposes than the other.

In the first place, the production yield of Mauritia is far
greater (Heinen and Ruddle 1974). For instance, the amount
of sago needed for a well prepared nahanamu festival varies

[319 ]

[ oz |

Transporting a calabash full of shredded Maniacaria saccifera pith to

the washing
Photograph by JOHANNES WILBERT

IOX @LV1g

PLATE XCII

[ 321 ]

Woman pours the Manicaria saccifera pith onto the sieve.

y JOHANNES WILBERT

Photograph b

from 500 to 1,500 kg., allowing from 2 kg. to 4 kg. per partici-
pant. Families are engaged from six to ten weeks in its pro-
duction. To produce this amount of sago from Manicaria
would be almost impossible in terms of time and labor. The
Warao insist, however, that Manicaria sago is acceptable to
the Supernaturals and that it can be offered to them in pro-
pitiation.

A decisive factor in making palm starch acceptable to the
gods is that, in its fresh state, it is practically odorless. Odors
play an important role in communicating with the Super-
naturals who accept as pleasing only the smell of tobacco
smoke and that of carana (Protium heptaphyllum). Both
types of sago available to the Warao fulfill this condition, but,
besides relative abundance, Mauritia outdoes Manicaria
on one other important score — it is colorless. When fresh,
Mauritia sago is as white as plaster of Paris; whereas Mani-
caria starch is beige to light brown. White is the color of
predilection of the directional world gods who supposedly
feel much less attracted by the color of Manicaria starch,
which turns a dark brown several days after production.

The color makes it much more appropriate for the Mani-
caria to serve a secondary god, that of the dark Underworld.
This spirit is known as Kanishabarao and is believed to dwell
below the earth in company with his people, right next to
the abode of Ya ahuba, the Temiche Master Snake.

All palms, major trees, and most animals depend for their
existence upon a master snake. The snake of Manicaria is a
night spirit, appearing on earth only around midnight, when
it comes to move softly the beautiful leaves of the palm.
Thus, Manicaria pertains to the midnight sun as Mauritia
belongs to the gods of the zenith and the world mountains
at the cardinal and intercardinal points of the universe (Wil-
bert 1973).

This aspect recalls the day and/or night association of
palms in other cultures. As Schultes (1974:7) pointed out,
similar beliefs prevail among the Kuripako of Colombia in
connection with Leopoldinia Piassaba. Fiber gatherers “are
often bitten by poisonous snakes that infest the thick clumps

[ 322 |

of hanging fiber —a danger that probably underlies in part
the natives’ belief that the evil spirit, the curupira, inhabits
piassaba groves and wanders around at night.” The same
curupira, we should add, is the central Master-of-Animals
figure in South America who functions as the patron spirit
of the trees and the forest (Zerries 1954: 18). Kanishabarao
of the Warao is obviously a cousin of the Amazonian curupira.
Furthermore, there exists also a sun-palm relationship among
the Yukuna Indians of Colombia, whose kai-ya-tee festival
resembles the annual sago festival, nahanamu, of the Warao,
and “basically celebrates the harvest of the pupunha palm
which was given to the Yukuna people by the ‘Sister of the
Sun’ as one of their major cultivated foods” (Schultes 1974:
16).

That the Manicaria belongs to the midnight sun rather than
to the day sun becomes apparent in the belief that, while its
sago may be unsuitable for a nahanmu festival in honor of
the cardinal gods, it is, nevertheless, used for this purpose by
Kanishabarao and his people of the underworld. Annually
they prepare a feast of temiche sago, and that is why the
Indians may come across hollowed-out trunks of the palm in
the forest. Kanishabarao and his people also eat the nuts and
all the parts of the palm that humans enjoy.

Finally, Manicaria sago is believed to be extracted by the
monkey (naku) for purposes of celebrating an animal na-
hanamu palm festival. Monkeys are supposedly very fond of
the starch and are believed to be yet other ‘ ‘people” that
hollow out the palms one happens across in the forest.
Instead of fish or crab, monkeys are said to prefer large
spiders (abunamoko) with their starch.

The custom of recovering palm starch from Manicaria
is essentially a feature of the past with the Warao. It came
to my attention when an elderly informant remembered hear-
ing his uncle and other elders send workers out to prepare
Manicaria sago. It was done occasionally, he said, when the
people were away from the Mauritia groves fishing on a ma-
jor river. To supplement their diet, they relied on Manicaria
sago which was always close at hand in the Intermediate

[ 323 ]

PLATE XCIII

Beginning the process of washing starch out of the shredded pith
of the Manicaria saccifera. Highly contaminated water is taken di-

rectly from the swamp.
Photograph by JOHANNES WILBERT

PLATE ACV

ams

Woman kneading the pith with both hands,

Photograph by JoHANNES WILBERT

[ 9z¢ |

After the starch has settled down at the bottom of the V-shaped trough the woman

starts ladling out the water. Photograph by JOHANNES WILBERT

AQX SLVIg

PLATE XCVI

Starch is scooped with hands from the bottom of the trough into a
calabash. Photograph by JOHANNES WILBERT

[ 327 ]

LHAGTI AA SANNVHO[ fig ydvisojoyg ‘TeUa}eul Aayseq ay} Jo
BuQ}OL ploae 0} Ay[Nyorvo AIOA JOUTeIYS OY} seyseM ULWIOM Yd ayy SuIssad0id J9z;y

ee

[ 328 |

PLATE XCVIII

Woman dismantles the washing stand after extracting enough starch.
Photograph by JOHANNES WILBERT

[ 329 ]

zone of the Delta. The wife of my informant, who comes
from a different region than her husband, said she also dis-
tinctly remembered having seen people collect and eat
temiche starch. Those who still know about it are in their
fifties and older. The younger people in the village had never
seen or even heard about it.

One need not invoke prophetic insight to predict the dis-
appearance, in the near future, of palm-starch production
from all of South America. The majority of southern tribes
referred to in the introduction have already either given it
up or are so reduced in number that the practice of starch
recovery will soon come to an end, Among the Warao, even
Mauritia extraction is declining rapidly and will soon be-
come a rarity. Yet, in South America, as elsewhere, recovery
of starch from different genera of palms represents the sur-
vival of a once more common tradition that reaches back
into the remote, prehistoric past. It survived the Neo-Indian
revolution of subsistence agriculture only among a small
number of marginal tribes, but it is giving way increasingly to
agricultural staples even among them. In any event, for the
Warao this form of traditional aboriculture provided a mea-
sure of economic stability not unlike that achieved by other
indigenous societies through agriculture. The entire man-
palm relationship as it developed among these people over
many centuries can be fully understood only in the context of
Manicaria, Mauritia, and Euterpe ethnobotany. Realizing
this and going a little beyond the scope and evidence of the
present paper, may I suggest in concluding that economic
stability is not the only benefit that the Indians have derived
from gathering around these regal plants. As their ethnobo-
tanical lore reveals, the palms have nurtured among the
Warao an exquisite partnership between man and nature —
a symbiosis that, in addition to a viable socio-economic blue-
print, generated an ideological matrix that gave meaning to
the world and purpose to life. Surely, achievements of this
sort must rank among the finest that mankind has made
anywhere.

[ 330 ]

pea

10.

11.

12.

ENDNOTES

. Cf. Lévi-Strauss 1950: 469-472; Fuerst 1970: 114-122.
. The only previous mention that I could find was Barral (1949:

150), who said: “Pero el “aru” propiamente tal, el “arw” guaraino
[Warao], es la torta hecha con la fécula extraida de la “Mauritia
flexuosa”, 0 también del temiche (Phitalephas temiche), no tan
ponderada como la de moriche, pero tan real.”

Among the secondary palms, the authors (ibid., 116) list: Acro-
comia, Caryota, Coelococcus, Corphyra, Eugeissona, Phoenix, and
Pholidocarpus.

. Iam using the so called ethnographic present, although the prac-

tice of palm-starch production may very well have been discon-
tinued now among some of the tribes of southeastern South Amer-
ica.

Gumilla 1791, 1: 145; Schomburgk 1848: 49; Turrado Moreno
1945: 73-83; Suarez 1966; Wilbert 1972: 81-82: Heinen and
Ruddle 1974.

. Cf. Dahlgren 1936: 202 and references since then (1936): Bailey

1943: 392-393 Dugand 1940: 43; McCurrach 1960: 129-131;
Wessels Boer 1965a, b; Braun 1968: 111.

. If it is any consolation to botanists, Manicaria is ethnobotanically

all but unknown; the snatches of information, frequently repeated,
are restricted largely to the usefulness of the leaves for thatch
and of the spathe for “monkey-caps.” In estimating migration and
settlement patterns, for instance, archaeologists and ethnologists
may have underestimated the effect of palm exploitation on the
relative stability of autochthonous peoples in South America and
elsewhere. I would like to thank Mr. August Braun and Drs.
Harold E. Moore, Jr., Richard Evans Schultes, and Julian A.
Steyermark for advice and assistance with the botanical aspects
of this paper.

Von Spix and von Martius (1823-1831; 3, 989) observed the same
custom practiced on Marajo Island in the mouth of the Amazon.
Lat. manica = sleeve; saccifera = sack-bearing.

The ratio of single to multilobed fruits seems to vary markedly.
In a second sample examined, I counted a total of 72 fruits with
42 one-seeded, 27 two-seeded, and three 3-seeded ones.

The apical incision in Manicaria eophylls results in two opposite
terminal leaflets, each segment with an acute apex and smooth
margins.

The hoe and all the other tools and techniques employed in the
recovery of Manicaria starch are the same ones that the Warao
use for the extraction of sago from Mauritia.

[ 331 ]

13. I am taking the Indians’ word for the fact that Manicaria has no
annual flowering period and that the palm carries starch the year
round. If incorrect, the result of a single test ought not to be
generalized, since it is known that the starch content in Mauritia
of the Orinoco Delta, for instance, may vary seasonally from 2 kg.
to 60 kg. per palm (Heinen and Ruddle 1974: Peae

[ 332 |

REFERENCES

Appun, Karl Ferdinand. 1871. Unter den Tropen: Wanderungen in

Venezuela, am Orinoco, durch Britisch Guayana und am

Amazonenstrome in den Jahren 1849-1868. Jena: Hermann

Costenoble.

1961. En los tropicos. Universidad Central de Venezuela.
Ediciones de la Biblioteca, Vol. 2. Caracas. Translation of
1871 original ed. Frederica de Ritter, Translator.

Bailey, Liberty Hyde, 1933, “Palmae indigenae Trinitenses et To-

bagenses.” Gentes Herbarum, Vol. 7, pp. 353-445.

. 1943. “Flora of Panama: Palmaceae.” Annals of the Missouri

Botanical Garden, Vol. 30, pp. 327-396.

1947. “Indigenous Palms of Trinidad and Tobago.” Gentes
Herbarum, Vol. 7, Fasc. 4, pp. 251-445,

Barral, Basilio Maria de. 1949, “Cantares Guaratinos.” Venezuela
Misionera, Ano 9, No. 124, pp. 148-150.

Barrau, Jacques. 1959. “The Sago Palms and Other Food Plants of
Marsh Dwellers in the South Pacific Islands.” Economic Bot-
any, Vol. 13, pp. 151-162.

Braun, August. 1968. Cultivated Palms of Venezuela.” Principes:
Journal of the Palm Society, Vol. 12, Nos. 2-3, pp. 39-103;
No. 4, pp. 111-136.

Burkill, Isaac Henry. 1935, A Dictionary of the Economic Products
of the Malay Peninsula. Vols. I and II. London: Crown
Agents for the Colonies.

Cadogan, Leon. 1960. “Algunos textos del Guayaki del Yfaro.”
Boletin de la Sociedad Cientifica del Paraguay y del Museo
Dr. Andrés Barbero Etnogréfico e Histérico Natural, Vol. 4,
Etnografia, No. 4. Asuncién.

Civrieux, Jean Mare. 1957. “Nombres folkléricos e indigenas de al-
gunas palmeras Amazonico-Guayanesas con apuntes etnobo-
tanicos.” Boletin de la Sociedad Venezolana de Ciencias Natu-
rales, Vol. 18, No. 89, pp. 195-233.

Dahlgren, Bror Erik. 1936. Index of American Palms. Chicago Field
Museum of Natural History, Botanical Series, Vol. 14, pp. 1-
438.

Dugand, Armando. 1940. “Palmas de Colombia.” Caldasia, Vol. 1,
pp. 442-458.

Fuerst, René. 1970. “Une civilisation du palmier.” Zeitschrift fiir
Ethnologie, Vol. 95:1, pp. 114-122.

Gaertner, Joseph, 1791. De fructibus et seminibus plantarum. 2 Vols.
Stuttgart.

[ 333 ]

Gumilla, Joseph. 1944 (1791). El Orinocco ilustrado. Biblioteca
Popular de Cultura Colombiana. Bogota.

Heinen, H. Dieter, and Julio Lavandero. 1973. “Computacion del
tiempo en dos subtribus Warao.” Antropologica, No. 35, pp.
3-24, Instituto Caribe de Antropologia y Sociologia. Funda-
cién La Salle de Ciencias Naturales. Caracas.

Heinen, H. Dieter, and Kenneth Ruddle. 1974. “Ecology, Ritual, and
Economic Organization in the Distribution of Palm Starch
among the Warao Indians of the Orinoco Delta.” Journal of
Anthropological Research, Vol. 30, No. 2, pp. 116-138.

Im Thurn, Everard F. 1967. Among the Indians of Guiana. New
York: Dover.

Lévi-Strauss, Claude. 1950. “The Use of Wild Plants in Tropical
South America.” Handbook of South American Indians, Vol.
6, pp. 465-586. New York: Cooper Square Publishers, Inc.

Liddle, Ralph Alexander. 1928, The Geology of Venezuela and Trini-
dad. Fort Worth.

McCurrach, James C. 1960. Palms of the World. New York: Harper
and Brothers.

Métraux, Alfred. 1946a, “Ethnography of the Gran Chaco.” Hand-

book of South American Indians, J. H. Steward, ed., Bureau

of American Ethnology, Bulletin 143, pp. 197-370. Washing-
ton.

1946b, “The Caingang.” Handbook of South American In-
dians, J. H. Steward, ed., Bureau of American Ethnology,

Bulletin 143, pp. 445-475. Washington.

Schomburgk, Robert H. 1848. Annotations to the Hakluyt Society's
edition of Sir Walter Raleigh, The Discovery of the Large,
Rich, and Beautiful Empire of Guiana. London.

Schultes, Richard Evans. 1974. “Palms and Religion in the Northwest
Amazon.” Principes: Journal of the Palm Society, Vol. 18, pp.
3-21.

Spix, Johann Baptist von, and Carl Friedrich Phil. von Martius. 1823-
1831. Reisen in Brasilien in den Jahren 1817-1820. 3 Vols.
Stuttgart: F, A. Brockhaus.

Spruce, Richard. 1908. Notes of a Botanist on the Amazon and Andes.
2 Vols. London: Macmillan and Company.

Standley, Paul C., and Julian A. Steyermark. 1958. Flora of Guate-
mala. Fieldiana: Botany, Vol. 24, Part I, pp. 271-273. Chi-
cago Natural History Museum.

Stohr, Waldemar, ed. 1972. Melanesien: Schwarze Inseln der Siidsee.
Rautenstrauch-Joest Museum fiir Volkerkunde. Cologne.

Suarez, Maria Matilde. 1966. “Les utilisations du palmier “moriche”
(Mauritia flexuosa L.f.) chez les Warao du Delta de L’Oré-

[ 334 |

noque, territoire Delta Amacuro, Vénézuela.” Journal d Agri-
culture Tropicale et de Botanique Appliquée, Vol. 13, pp. 33-
42.

Tomlinson, P. B. 1959. “Seedling Leaves in Palms and their Morpho-
logical Significance.” Journal of the Arnold Arboretum, Vol.
41, pp. 414-428.

Turrado Moreno, F. Angel. 1945. Etnografia de los indios Guaratinos.
Conferencia Internacional de Agricultura, No. 15. Caracas.

Vellard, Jehan. 1934-35. “Les indiens Guayaki.” Journal de la Société

des Américanistes, n.s., Vol. 26, 1934, pp. 223-922; Vol. 27,

1935, pp. 175-244. Paris.

1939. Une civilisation du miel: Les indiens Guayakis du

Paraguay. Paris: Gallimard.

Wallace, Alfred Russel. 1853. Palm Trees of the Amazon. London:

John Van Vorst.

Wessels Boer, Jan Gerard. 19654. The Indigenous Palms of Surinam.
Leiden: E. J. Brill.

. 1965b. Flora of Surinam, Leiden: E. J. Brill.

Wilbert, Johannes. 1963. “Vestidos y adornos de los indios Warao.”
Antropoldgica, No. 12, pp. 6-26.

. 1972. Survivors of Eldorado. New York: Praeger.

. 1973. “Eschatology in a Participatory Universe: Destinies of
the Soul among the Warao Indians of Venezuela.” In, Eliza-
beth Benson, ed., Death and Afterlife in Pre-Columbian
America, pp. 163-189. Washington, D.C.: Dumbarton Oaks.

Zerries, Otto. 1954. Wild- und Buschgeister in Siidamerika. Wies-

baden: Franz Steiner Verlag.

[ 335 ]