Citation
A monograph of the genus Clitoria (Leguminosae: Glycineae)

Material Information

Title:
A monograph of the genus Clitoria (Leguminosae: Glycineae)
Creator:
Fantz, Paul R., 1941-
Publication Date:
Language:
English
Physical Description:
3 v. (xii, 1052 leaves) : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Calyx ( jstor )
Claws ( jstor )
Flowers ( jstor )
Inflorescences ( jstor )
Leaves ( jstor )
Legumes ( jstor )
Nerves ( jstor )
Species ( jstor )
Stipules ( jstor )
Trichomes ( jstor )
Clitoris ( lcsh )
City of Sebring ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Includes bibliographical references (leaves 1026-1050).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Paul R. Fantz.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
029028128 ( ALEPH )
AAB3963 ( NOTIS )
03403316 ( OCLC )

Downloads

This item has the following downloads:


Full Text












A .OGRPi OF 1HE GENUS U1 l 0il i (- GU I,' S p !'


GLYCi i i -


PAUL R. FA%!VZ















A DISS-RTATICG'; PRE~',TED TO THE GRA JATE COUNCIL OF
TFH LUIV'ERSIT 0- F'LORi0,l
iN PARTIAL FULF L;:,T O. T!lE RJIRM1M:L.. FOP TiHE
DEGREE CF G[OC ;- c; PHIi .? !Y






UNIVERSE IT 'Y OF FLORkDF,


1I9771





















COPYRIGHT


By


PAUL R. FANTZ


1977
















ACKNOWLEDGEMENTS


I would like to express my appreciation for the interest, patience,

guidance, and helpful criticisms of Dr. Daniel B. Ward, under whose

supervision this monograph was accomplished. I also wish to thank the

other members of my committee, Dr. Dana G. Griffin III, Dr. Willard

Payne, Dr. Terry Lucansky, Dr. Jack Ewel, and Dr. John Kaufmann, for

their assistance during this study and for reading and providing

helpful editorial evaluations of the manuscript.

I am very grateful to the curators and staff of the herbaria

listed in the chapter on "Systematic Criteria and Taxonomic Notes" for

lending their material, including type-specimens, for use in this study.

Special thanks are extended to Gus Kovalick and the Staff of Hume

Library and Library West for their diligence and extensive aid in

searching for titles and institution of deposit, and obtaining xerox

copies of many older botanical publications; to Dr. Armstrong for her

assistance in providing appropriate geographical indexes, maps, and

atlases during my search for the obscure localities noted by the plant

collectors; and to Dr. Armstrong, Dr. Zimmermann, and Dr. Jones for

providing me with special library privileges to borrow needed references

for extensive periods during this study.

Thanks are extended to Dr. Daniel B. Ward and his staff at the

Herbarium of the Agricultural Experiment Station, for their assistance









in the necessary correspondence and the securing and returning of

borrowed herbaria material used in this study, and for providing the

needed facilities and space for the storage and examination of this

material.

I am indebted to Dr. Robert H. Mohlenbrock of Southern Illinois

University, who made this monograph possible by introducing me to the

world of plants, and whose enthusiasm stimulated me to change careers

and pursue a lifetime study in Vascular Plant Systematics.

I am grateful to Barbara Smerage and her assistant, Linda

Nordstedt, for their time, patience, and cooperation in the typing of

this manuscript.

A special thanks is extended to my wife, Janet, for her patience,

understanding, and the sacrifices she has made during the five years

it took to complete this study and for her assistance in portions of

this study.



















TABLE OF CONTENTS


Page


CHAPTER


ACKNOWLEDGEMENTS .


ABSTRACT .

INTRODUCTION .

ECONOMIC IMPORTANCE .

Medicinal Properties .
Leaves .
Seeds .
Flowers .
Additional Notes .
Aphrodisiacs .
Ornamental Properties .
Poisons and Pesticides. .
Dyes and Fibers .
Food for Man .
Fodder and Green Manure .

HISTORY OF CLITORIA .


Origin of the Generic Name .
General History . .
Historical Chronology .
Pre-Linnaean (1650-1753: Polynominal Period).
Post-Linnaean (1753-1824: Descriptive Period)
Post-Linnaean (1825-1858: Revisionary Period)
Post-Linnaean (1859-1932: Floristic Period) .
Post-Linnaean (1932- present: Research Period).

PRIOR RESEARCH ON CLITORIA .


Anatomical Research .
Chemical Research .
Cytological Research .
Developmental Research. .
Morphological Research. .
Agronomic Research .
Palynological Research. .


. . i i i


xi


. 1


. 2 1


. 79

. 79
. 82
. 84
. 86
. 90
S91
S94


. .
. .
. .
. .
. .
. .
. .













MORPHOLOGY .


Habit .
Stems and Bran
Pubescence. .
Leaves .
Leaflets
Stipules
Petiole a
Petiolule
Observati
Inflorescence
Chasmogam
Cleistoga
Chasmocle
Bracts
Pedicels
Bracteole
Observati
Chasmogamous F
Calyx.
Vexillum
Alae .
Carina
Androeciu
Observati
Cleistogamous
Calyx. .
Corolla.
Androeciu
Gynoeciun
Observati
Legume. .
Stipe.
Valves
Observati
Seeds .
Seed Gern
Observati
Summary .
Species [
Character


ches. .

. .



and Stipels .
nd Rachis .

ons .

ous Inflorescences. .
mous Inflorescences .
istogamous Inflorescences .
istogamous Inflorescences
. .
. .

ons. ............
s owers. .

ons .
lowers .




Flowers .
. .
. .
. .
m .




ons .


Flowers .


ion .
ons .
m .
. .
ons .
. .





nation .
ons .
. .


descriptionss
's Occurring


in Relatively


.



. .

. .







. .



. .

.



Few


. .
. .
.

. ., .
. o


.


. .

. .

.
. .

. .
.
Species


DISTRIBUTION OF CLITORIA . .

SYSTEMATIC CRITERIA AND TAXONOMIC NOTES .

SYSTEMATIC TREATMENT . .


95
97
98
100
101
101
102
102
102
103
105
109
109
110
113
113
114
114
118
121
123
124
124
127
133
133
138
138
139
139
140
141
145
148
148
151
151
152
152
153

159

173

180


Page


. 95









Page
Subgenus Bractearia . ... 193
Section Bractearia . 198
1. Clitoria arborea. ... 201
a. var. arborea. ... 211
b. var. longiramosa. ... 215
c. var. pseudoamazonica. .. 216
2. Clitoria fairchildiana. ... 218
3. Clitoria andrei .. 230
4. Clitoria juinensis. .. 234
5. Clitoria moyobambensis .. 242
6. Clitoria amazonum ... 247
a. f. amazonum ... .255
b. f. vulgaris 256
7. Clitoria nervosa. ... 258
Section Flexuosa .... 265
8. Clitoria woytkowskii. .. 267
9. Clitoria flexuosa ... .275
a. var. flexuosa ... .282
b. var. brevibracteola .... .283
10. Clitoria pozuzoensis. .284
a. var. pozuzoensis .............. 292
aa. f. pozuzoensis .... .292
ab. f. subpalmata. ... 292
b. var. schunkei ... 293
Section Brachycalyx. ........ ........... .295
11. Clitoria brachystegia ... 299
12. Clitoria hermannii. 307
13. Clitoria glaberrima ... 311
14. Clitoria canescens. .. 321
15. Clitoria brachycalyx. .. 326
16. Clitoria dendrina 332
17. Clitoria froesii. 344
S. Section Cauliflorae. . ... 349
18. Clitoria arborescens. ... 357
19. Clitoria javitensis 372
a. var. javitensis .. 398
aa. f. javitensis. ... 399
ab. f. bracteosubtenda ... 401
b. var. portobellensis 403
ba. f. portobellensis .. 404
bb. f. pilosa .. 407
bc. f. truncata .. 409
c. var. longiloba. 409
d. var. grandifolia.. ............ .410
e. var. klugii ... .412
20. Clitoria cavalcantei. ... 413
21. Clitoria coriacea ... 418
22. Clitoria tunuhiensis. ... 425
23. Clitoria sagotii. ... 429
a. var. sagotii. 439
b. var. caniculata ... .442
c. var. sprucei. ... 444









Page
24. Clitoria kaieteurensis. ... 446
25. Clitoria pendens. ... 451
26. Clitoria leptostachya 457
a. var. leptostachya 465
b. var. fruticosa .. 466
27. Clitoria selloi ... .466
28. Clitoria obidensis. ... 474
29. Clitoria plumosa. 480
Subgenus Clitoria . ... 486
30. Clitoria lasciva. ... 494
31. Clitoria ternatea ... .504
a. var. ternatea ... .536
aa. f. ternatea .. ... .. 536
ab. f. pauciflora. ............ 571
ac. f. fasciculata .. .. 571
ad. f. albiflora .. 574
b. var. angustifolia 584
c. var. pleniflora 591
ca. f. pleniflora. .... 596
cb. f. leucopetala 600
cc. f. subpolyadelpha. .. ... 601
32. Clitoria heterophylla 602
a. var. heterophylla .... .609
b. var. pedunculata .. 610
33. Clitoria biflora .. 612
34. Clitoria kaessneri. 619
Subgenus Neurocarpum. . 628
Section Mexicana .. .640
35. Clitoria polystachya. ... 645
a. var. polystachya .. 654
b. var. pringlei ... 659
c. var. congesta .. 659
36. Clitoria monticola. ... 660
37. Clitoria triflora ... .665
38. Clitoria mexicana. ... 671
39. Clitoria humilus. ... 682
40. Clitoria cordobensis. ... 686
41. Clitoria fragrans ... .696
42. Clitoria mariana. 705
a. var. mariana. ... 718
aa. f. mariana ... .. 722
ab. f. pedunculata 743
ac. f. pubescentia .. 746
b. var. orientalis 749
Section Tanystyloba. . ... 758
43. Clitoria macrophylla. .. 762
a. var. macrophylla ............. 774
b. var. sericea. .... .. 777
44. Clitoria javanica 779
45. Clitoria cordiformis. .... .... 784


viii









Page

46. Clitoria linearis ... .791
47. Clitoria hanceana 795
a. var. hanceana ... 802
b. var. thailanensis 805
c. var. laureola 805
d. var. latifolia. 809
e. var. petiolata. .............. 810
48. Clitoria australis. ... 811
Section Neurocarpum .. 820
49. Clitoria stipularis 826
a. var. stipularis ... 833
b. var. latifolia .. 836
50. Clitoria densiflora ... .837
51. Clitoria irwinii. .. 846
52. Clitoria laurifolia ... .852
a. f. laurifolia ... 863
b. f. glabrior ... .873
c. f. fasciculata. ... 879
d. f. parvifolia ... 879
e. f. petiolata. ... 880
53. Clitoria guianensis ... 881
a. var. guianensis 891
aa. f. uianensis .. 892
ab. f. macrofructa .... .903
ac. f. imperfecta. ... 903
ad. f. unifoliata. ... 905
b. var. macrocleistogama .. 905
c. var. chapadensis ............. 906
54. Clitoria epetiolata 910
a. var. epetiolata ... .916
b. var. angustissima 919
c. var. latiuscula ... 920
55. Clitoria simplicifolia. 921
56. Clitoria flagellaris. .. 927
57. Clitoria falcata. .. 932
a. var. falcata. ... 943
aa. f. falcata 943
ab. f. heteromorpha ........... 960
ac. f. stipulacea. ... 964
ad. f. longirachis ... .965
b. var. latifolia .. 966
c. var. aurantiaca .... .. 966
d. var. glabrescens. .. 971
58. Clitoria nana ... .972
a. var. nana .. ... ..... 978
b. var. caaguazuensis .. 979

NUMERICAL LIST OF TAXA .... ... .. 980

NOMINA ET SYNONYMA CLITORIA COMPLECTENS. ... 986









Page
PHYLOGENETIC CONSIDERATIONS. . ... 1007

Evolutionary Trends in Clitoria .... 1008
The General Plant Structure. ... 1009
Vegetative Structures. ... 1013
The Inflorescence Structures .... 1014
The Flowers . 1016
The Fruits and Gynoecium . 1018
Pubescence . ... 1020
Nonmorphological Data. ... 1021
Phylogenetic Pathways in Clitoria .. 1022

BIBLIOGRAPHY . 1026

BIOGRAPHICAL SKETCH. . ... ..... 1051










Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy


A MONOGRAPH OF THE GENUS Clitoria (LEGUMINOSAE: GLYCINEAE)


by

Paul R. Fantz

August 1977

Chairman: Daniel B. Ward
Major Department: Botany

This monographic treatment of the genus Clitoria L. (Leguminosae:

Glycineae) is the first since that of George Bentham in 1858. The

genus is treated on a world-wide basis. The study is based upon

examination of approximately 7000 specimens from thirty-three lending

institutions, and from field studies of United States species. Treated

in synonymy are Clitorius Petiv. ex Dill., Clytoria Presl, Macrotrullion

Klotsch ex Schomb., Martia Leandr. Scar., Martiusia Schult., Nauchea

Desc., Neurocarpon Hamilton, Neurocarpum Desv., Neurocarpus Hassk.,

Rhombifolium Rich. ex DC., Rhombolobium Rich. ex H.B.K., Ternatea Tourn.,

Ternatea Tourn. ex Mill., Vexillaria Eaton, and Vexillaria Raf.

Clitoria includes fifty-eight species and fifty-one nontypical

subspecific taxa. Forty-six species are native to the neotropics and

subtropics. One paleotropical species has been introduced and

naturalized in many regions of the neotropics. One species is native

to temperate North America and to tropical Southeast Asia. Eleven

species are native to the paleotropics in Africa and Madagascar, the

Indian subcontinent, Indonesia, and Arnhem Territory of northern

Australia. Two neotropical species have been introduced and naturalized








in portions of Africa and Indonesia. One species is introduced and

naturalized in many Pacific islands.

A history of Clitoria is provided in which the origin of the

generic name is traced to Breyne in 1678. A detailed morphology of the

genus is presented. Cleistogamous flowers, as well as chasmogamous

flowers, are noted to occur in thirteen species. Species of Clitoria

are recorded to have economic uses as medicinal plants, ornamentals,

aphrodisiacs, poisons, pesticides, dyes, fibers, food, fodder, and

green manure. Prior research on species of Clitoria is summarized in

the areas of agronomy, anatomy, chemistry, cytology, developmental

botany, morphology, and palynology.

A systematic treatment is given which recognizes three subgenera

and eight sections. These include sections Bractearia (7 spp.),

Flexuosa (3 spp.), Brachycalyx (7 spp.), and Cauliflorae (12 spp.) of

subgenus Bractearia; subgenus Clitoria (5 spp.); sections Mexicana

(8 spp.), Tanystyloba (6 spp.), and Neurocarpum (10 spp.) of subgenus

Neurocarpum. Each species is described and illustrated. A complete

synonymy is given. A nomenclatural discussion of type collections,

phenology, distribution, subspecific taxa, and citation of examined

specimens is provided under each species. Economic importance and

vernacular names are included for a species when known. Each taxon has

a mapped distribution. Keys are provided to each taxonomic level.

Phylogenetic considerations are given based upon the interpretation

of morphological data presented in this study. Subgenus Bractearia is

regarded as a primitive subgenus with subgenus Neurocarpum as advanced.















INTRODUCTION


The genus Clitoria contains fifty-eight species located mostly

within the tropical belt with a few species distributed within the

temperate zone. The bulk of the genus (ca. eighty percent of the

genera), is neotropical. The genus is absent from Europe, most of

Asia (except the Indian subcontinent), and most of Australia.

Although there is no universal name by which the genus is commonly

known, the most often used vernacular name for the genus is "butterfly

pea." Centrosema, a close morphologically and historically related

genus, bears the same vernacular name or is sometimes distinguished

as the "spurred-butterfly pea." When horticulturalists attempted to

standardize plant names (Kelsey and Dayton, 1942), they adopted

"butterfly pea" for the genus Centrosema and adopted "pigeon wings"

for the genus Clitoria, the latter name rarely used by botanists.

The Papilionaceous genus Clitoria is included by most botanists

within the tribe Phaseoleae, subtribe Glycineae. In the few

traditional tribes, a number of genera are usually included in a

list of exceptions to the typical tribal descriptions. In order to

reduce the number of exceptions, Hutchinson (1964) elevated a number

of subtribes to the tribal level. Thus, he placed Clitoria within the

tribe Glycineae.

Flowering plants of Clitoria are easily determined in the field

by a few unusual characteristics. Papilionaceos, resupinate flowers








which have a bearded, geniculate style identify the genus. Specimens

in fruit become more difficult to recognize due to the variation of

the legume. However, odd-pinnate compound leaves (usually 3-foliate),

persistent stipules, stipels, bracts, and bracteoles, plus a

persistent, five-toothed, infundibular calyx are reliable characters

that, taken collectively indicate the genus Clitoria from other genera.

The genus Centrosema (Glycineae) is often confused with Clitoria.

Linnaeus established Clitoria in his "Species Plantarum" in 1753, and

included four species. Two of those species, C. brasiliana and

C. virginiana, were later placed in Clitoria section Centrosema

by de Candolle in 1825. Bentham (1858) segregated Centrosema from

Clitoria and elevated the taxon to the level ot genus. In the keys of

most floristic and revisionary treatments, Clitoria is distinguished

from Centrosema by its "bearded style" or "pubescent style." This

character is often misinterpreted. Clitoria can be distinguished

morphologically from Centrosema by several characters. These

differences are summarized in Table 1.

The last revisionary work on Clitoria was a synopsis published by

Bentham in 1858. This treatment included a total of three sections

with twenty-six species, more than half of which were originally

described by Bentham. Bentham's treatment has been the major

reference relied upon by nearly all authors of floristic works that

included the genus. Bentham's treatment contains a number of

deficiencies, however, and now includes some conflicts with the

International Code of Botanical Nomenclature (Stafleu, 1972).

A number of problems presently exist within the genus that need

resolution, and these can be categorized into six areas. One problem











Table 1. A morphological comparison of the genera Clitoria and
Cent rosema


CHARACTER


Calyx


Vexillum

Alae vs Carina


Style


Ovary


Legume


Leaves


CLITORIA


Infundibular, lobes
broad, shorter than
to nearly equal tube
length (rarely longer)


Spurless


Alae longer than carina

Geniculate, bearded
lengthwise to base


Stipitate (or
subsessile in
Subgenus Clitoria)

Stipitate, ecostate or
with 1 lateral nerve
near midline of valve

Pinnate typically
3 (1) leaflets
(Subgenus Clitoria
commonly 5, 7, 9, or
11 leaflets


CENTROSEIMA


Short-campanulate,
lobes narrow, equal to
or longer than tube
length

Spurred or gibbous

Alae subequal carina

Broadly incurved
(U-shaped), pubescent
only beneath stigma

Subsessile



Subsessile, 2 prominent
nerves laterally, one
near each suture

Pinnate typically
3 leaflets, rarely
5, 7, or 11 leaflets


_~_








is the proliferation of names and their dulineition into specific and

subspecific taxa. Eleven genera must be included in synonymy under

Clitoria. Bentham cited only two, although by tracing back through

his cited references, the genera included in synonymy increased to

six. To date, nearly two-hundred and fifty published binominals (plus

ca. twenty nom. in sched.) are included within the total concept of

the genus Clitoria, senso lato. Of these, one hundred and sixty of

the binominals were published under the naie Clitoria. Bentham

included only sixty of these binominals in his treatment, and since

Bentham's revision, nearly fifty new species have been described,

almost double the total number of species that he recognized. Bentham

did not include subspecific taxa in his synopsis. Twenty-one of the

twenty-three subspecific taxa published were d scribed after Bentham's

treatment.

A second problem is the lack of complete and detailed descriptions

of species, one probable cause of the large abundance of names. Few

of the published specific names were accompanied by detailed

descriptions. Many were published with short diagnostic descriptions

that included a number of characters that are typical of the genus or

subgeneric groups. Many of the descriptions contain relative terms

(i.e. short vs. long, large vs. small) and lacd precise measurement

ranges. Often a character included in one description is lacking in

the description of another species.

Some floristic treatments have included descriptions of their

species, occasionally offering more detail than the original

description. These descriptions can be helpful, but a few have

compounded the problem. An example would be MRicbride's treatment





5


(1943) of the Peruvian species of Clitorid. Many of the specimens

examined by Macbride had been misidentified and did not match the

type collections. This has resulted in descriptions placed under

inaccurate names (e.g., C. aiiazonum and C. nervosa, neither species

occurring in Peru).

A third problem that occurs in conjunction with the first two is

the lack of keys to the species of Clitoria. Bentham did not include

a key in his treatment, although he used some diagnostic phrases that

separated the species into groups. The only keys that exist are those

found within certain floras. Keys which included five or more species

of Clitoria are found in floristic treatments for only a few countries

in South America, a few in Central America, plus one treatment for

Southeast Asia. They often include characters which work well

locally, but not when more species are added from neighboring areas.

A fourth problem is incomplete information, on types. A number of

species had incomplete citations, making type interpretation difficult.

Bentham was partially responsible in that he otten paraphrased data

from herbarium labels. Specimens labeled as "type" in the personal

herbarium of Bentham and in other herbaria which he examined, often

bear a collection number or date which Bentham did not cite.

Occasionally type information has been distorted. For example,

Howard (196/) proposed the name C. fairchildiana as a substitute for

the homonym C. racemosa Benth., correctly citing the Brazilian type

collection of Pohl. Then he distributed speci;,ens (Howard 17052)

from a cultivated tree at the Jennings Estate in Miami, Florida,

United States, as the type tree.









A fifth problem is the number of species that bear illegitimate

names as defined by the International Code of Botanical Nomenclature.

Although Bentham was not bound by a code developed nearly a century

after his revision, he did select names by a process that conflicted

with the Code standards in practice today. Later botanists who

assumed that those names were acceptable under the Code have often

been unaware of these differences and have used names that cannot

stand today.

A sixth problem is lack of information as to the geographic

distribution of the species. Distribution statements have often been

based upon one or a few collections and appear to sometimes have

been wholly hypothetical. The problem has been compounded in some

species which were described originally from cultivars outside the

normal distribution range (e.g., C. arborescens). Distributions in

some species have been distorted when floristic treatments have

mistakenly included a species within their geographical areas or

inaccurately reported the species from other areas.

This author has examined nearly eighty percent of the type

collections for the two hundred and fifty binominals associated with

Clitoria. Many of the types not examined belonged to species described

and placed within the genus Clitoria, but later transferred to other

genera, often Centrosema. In most cases, descriptions of these

species contained diagnostic phrases (i.e. campanulate calyx, spurred

vexillum) sufficient to indicate that the species does not belong to

the genus Clitoria.

Bentham's section Clitorianthes was most poorly understood.

Nearly fifteen percent of the specimens that were examined in this








section were unidentified to species. Foi those specimens with names,

slightly more than fifty percent were misi.lentified. A large portion

of the species in Bentham's section NeurocJrpuYi were also misnamed.

However, many of these specimens were correctly identified to species,

but bore illegitimate synonyms. Relatively few problems were found

within Bentham's section Ternatea.

The purpose of this investigation has been to thoroughly

re-examine available specimens and types, to clarify the nomenclature

and species descriptions, and to prepare Tunctional keys and

illustrations to legitimate taxa.















ECONOMIC IMPORTANCE


With the exception of C. Ternatea, the species within the genus

Clitoria have been of relatively little economic use, except where

they are locally important. Clitoria ternatea is important

economically as a medicinal herb, as an o inamental, as a food and

cover crop, as a dye, as green manure, and as an aphrodisiac.

Clitoria ternatea has been prized as a cultivar at least since the

late seventeenth century, and has been carried by man throughout the

tropics, where it often has escaped and bc,:ome established. As a

result, thi, species now has a pantropical distribution. Attention

has been focused on C. Ternatea as a typical representative of the

genus because it is widespread, well know, and easily accessible for

economic use. Some species within the genus are reported to have some

of the same economic properties as C. tern ,tea, although these species

have not received the same publicity of C. ternatea. Additional

properties have been noted for a few species, for example, the

possible use as powerful insect and vertebrate poisons.

The genus Clitoria has great potential to be an economically

important legume genus. The economic uses of a particular species

properly belong under the treatment of that species. The purpose of

this chapter is to discuss tlhe economic properties known for the genus

and to point the way toward research in the economic botany of

Clitoria as a whole. Clitoria ternatea will be the major








representative of the genus in this discussion. Information obtained

from herbaria is documented with the collector's name and his

collection number enclosed within parentheses, for example, (Hahn

234).


Medicinal Properties

Nearly all the medicinal properties of Clitoria are recorded for

the species C. ternatea as a result of investigations by workers

interested in medicinal Indian, Southeast Asian, and Indonesian plant

species. Remedies and treatments reported are grouped below by

medicinal use for each plant organ used.

Roots

Root or root bark is often reported to be an effective laxative,

but may not always be safe. The degree of cleansing action varies in

the reports. Dalgado (1896), Kirtikar and Basu (1918), and Dutt

(1928) report use of the roots as a laxative. Kirtikar and Basu ibidd)

cited Mr. Mooden Sheriff (1891) who noted that the root bark was a

laxative in some cases. Chopra, Chopra, Honda, and Kapur (1958) noted

that the root bark was used as a laxative. Chopra, Badhwar, and Ghosh

(1949) and Chopra, Chopra, Honda, and Kapur ibidd) reported the roots

as a powerful cathartic similar to Jalap (Exogonium purga,

Convolvulaceae), but that it was not a safe medicine. Dey (1896),

Burkill (1935). Quisumbing (1951) citing Dey, and Gardner and Bennett

(1956) reported the roots as carthartic. Dey reported preparing the

remedy as an alcoholic extract of the roots. Quisumbing ibidd),

citing Rajan (1926), reported the roots as a purgative and narcotic

toxin. Rajan ibidd) indicated symptoms of the narcotic poisoning to








include unconsciousness attended with extreme irritability and a

peculiar loss of memory. Chupra, Badhwar, and Gosh ibidd), citing

O'Shaughnessy (1841), indicated that an alcoholic extract of the root

acts as a brisk purgative in doses of five to ten grains, but produces

gripping anJ tenesmus, and the patient will feel feverish and uneasy.

Quisumbing ibidd), citing Rajan ibidd), Sanyal and Ghose (1934), and

Burkill ibidd), reported the roots as an aperient. Roots were noted

to be a purgative medicine or remedy used in Martinique (Hahn 234).

Roots dre also used in stimulating urination and are useful in

ascites and fevers. Rajan (1926), Dutt (1928), citing the Hindus,

Sanyal and Ghose (1934), Qui umbing (1951), citing the three earlier

works, and Chopra, Chopra, Honda, and Kaputr (1958) reported the roots

as diuretic. Chopra, Chopra, Honda, and l.ipur ibidd) also reported

the root bark used as a diuretic. Kirtikar and Basu (1918), citing

Sanskrit writers, described Lhe root as diuretic, useful in ascites

(an accumul ition of excess fluid in the al.Jomiiial cavity) and fever.

They also cite Mooden Sheriff who noted that an infusion of the root

bark acts as a diuretic.

Quisumiing (1951), citing Sanyal and Ghose (1934) who cited

Mooden Sheriff (1891), and Kirtikar and Basu (1918), who also cited

Mooden Sheriff, reported that Sheriff spole highly of the infusion of

the root bark as a demulcent in cases of irritation of the bladder and

of the urethra.

Roots ire also used as a medicine in promoting menstrual

discharge. Quisumbing (1951) cited Crevost and Petelot (1929) who

reported the roots used as an emmenagogue.








Roots or root juice has been reported to cause nausea and

vomiting and to be useful for lung ailments and removing the phlegm

of chronic bronchitis. Lindley (1938) and Gardner and Bennett (1956)

reported the root as an emetic. Quisumbing (1951), citing Dalgado

(1896) and Crevost and Petelot (1929), indicated that each reported

the roots used as a vomitive. Drury (1873) and Dymock (1885), each

cited by Quisumbing ibidd), quote Mr. Ainslie (1826) who mentions the

use of the root in soup to remove phlegm in chronic bronchitis and to

bring on nausea and vomiting. Kirtikar and Basu (1918), citing Dymock

ibidd), indicated that in the Concan region of India, two tolas1 of

the root juice are given in cold milk to remove the phlegm of chronic

bronchitis through nausea and vomiting. Chopra, Badhwar, and Ghosh

(1949) indicated that opinions differed with regard to the emetic

properties of the root. In the Sudan, the roots are chewed for lung

and throat complaints (Prichard 4).

Dymock (1885), cited by Kirtikar and Basu (1918), indicated that

the juice of the root of a white-flowered form of C. ternatea is blown

up the nostrils as a headache remedy. A similar practice in the Sudan

has other effects. Before a cow is sent to the bull, the root is

chewed and blown up her nostrils to ensure fertility (Pritchard 4).

Roots are reported to be used as an antidote against snake-bite

and scorpion-stings. Chopra, Badhwar, and Gosh (1949) reported the

roots as a possible antidote against snake bites. Chopra, Chopra,

Honda, and Kapur (1958) reported that the roots were used in Indian


1. A tola is an Indian unit of weight equal to one silver rupee,
or 180 grains.








indigenous medicine as a treatment for snuae-bites (Cobras and

Daboias) and for scorpion-stings (genera huthus and Palamnoeus).

However, they found that remedies administered to dogs and rabbits

injected with the snake or scorpion venom in strict conformity with

the directions given in standard Indian medicinal books, had no

preventive, antidotal, or therapeutic effi:ct.

Leaves

The leaves are reported to be used against eruptive conditions

by Chopra, lKadhwar, and Ghosh (1949). Kirtikar and Basu (1918, citing

Watt (1889-96), and each later cited by Quisumbing (1951), reported

that an infusion of the leaves is used for eruptive conditions.

Heyne (1927), later cited by Burkill (1935) and Quisumbing (1951),

reported that the leaves of a white-flowered form of C. ternatea are

used as poultices in Java. Kirtikar and l.su (1918) and Quisumbing

ibidd), each quoting Mr. Taylor,2 stated that the juice of the leaves

mixed with qreen ginger is administered in cases of colliquative

sweating in hectic fever. Quisumbing (ibid) quoted Kirtikar and Basu

ibidd) who quoted Mr. Mukerji (1889-1904) in "Watt's Dictionary" that

juice of the leaves is mixed with common salt and applied all around

the ear for earaches, especially when accompanied with the swelling of

neighboring glands. Quisunbing ibidd) quoted Tavera (1892) who

reported that the leaves are used in the Philippines for swollen

joints. Chopra, Chopra, Honda, and Kapur (1958) reported that the

juice of the leaves is mixed with water to form a jelly which is taken


2. Taylor was quoted by both authorities; however, neither one
documented the reference source of Taylor.








as a cooling medicine for gonorrhea, and used externally for eczema,

prurigo, and impetigo.

Seeds

Seeds are used also as a laxative and considered as an effective,

safer method than the use of roots or root bark. Waring (1868),

Drury (1873), Dey (1896), and Crevost and Petelot (1929), each cited

by Quisumbing (1951), reported that seeds are used as a mild purgative.

Quisumbing ibidd) also cited Kirtikar and Basu (1918) who reported the

seeds as a purgative and an aperient. Chopra, Badhwar, and Ghosh

(1949) indicated that seeds in powdered form have purgative and

aperient properties, and are considered more useful and a safer

medicine than the roots. Quisumbing ibidd) cited Burkill (1935) who

reported that the seeds are used as an apcrient and contained a toxic

alkaloid. Quisumbing ibidd) cited Nadkarni (1927) who reported that

the seeds contained a fixed oil, a bitter acid resin, tannic acid,

glucose, and six percent ash. The testa was brittle and contained

cotyledons full of granular starch. Chopra, Chopra, Honda, and Kapur

(1958) cited Dymock, Warden, and Hooper (1890-93) who reported that

the seeds contained a fixed oil, a bitter resinous principle, and

tannin. (Chemical investigations on the seeds are reported under the

chapter "Prior Research on Clitoria." (Reler to that chapter,

subheading "Chemical Research," for details.)

Burkill (1935) quoted Tavera (1901) lwho reported that the seeds

are used in the Philippines as poultices for swollen joints. In an

earlier edition of his book (1892), Tavera had attributed this

medicinal property to the leaves.








Seeds are supposed to be effective in destroying or expelling

tapeworms. Waring (1868), Drury (1873), Dey (1896), and Crevost and

Petelot (1929), each cited by Quisumbing (1951), reported that the

seeds were used as an antihelminthic.

Flowers

Burkill (1935) cited "Medicinal Book of Malayan Medicine (Gard.

Bull. S.S. 6, 1930, p. 381)" which reported that the floral juice of

a white-flowered form of C. ternatea was used for inflamed eyes.

Additional Notes

Chopra, Chopra, Honda, and Kapur (1958) reported that C. ternatea

is alleged to have anti-dysenteric and anti-tubercular properties.

No plant organ was noted.


Aphrodisiacs


A few species are reported to ensure fertility and are used as

sexual stimulants. This topic is usually not included in the

literature possibly through delicacy complicated by the fanciful

resemblance of the flower to the female sexual apparatus. Yet, in

widely dispersed geographical areas with different species, the native

people use Clitoria plants as aphrodisiacs for man and domestic

animals. It is presumed that the natives follow the ancient

principles of the Doctrine of Signatures, which is based upon belief

that the plant structures which resemble portions of the human body

have been so structured as to advertise their ability to provide

remedies for ailments of those body portions.

The use of the plant in conception was implied by Rumpf (1747)

who noted that the Portuguese name "Fula criqua" was derived from the








flowers, and that Breyne (1678) had appropriately named the plant Flos

clitoridis ternatensibus (this plant now known as Clitoria ternatea).

The Portuguese translation of "Fula criqua" means "to speed or

hurry-up creation." In the Sudan, before a cow is sent to the bull,

the root of C. ternatea is chewed and blown up her nostrils to ensure

fertility (Pritchard 4). Near Anaconda Island on the Rio Napo,

Ecuador, this author attended a wedding presided over by a tribal

witch-doctor. In a discussion with him the night before the wedding,

the flower of C. pozuzoensis (a species that this author was searching

for locally) was described by cupping the hands for the vexillum and the

arching of the thumbs for the keel and wings. The grinning witch-doctor

apparently recognized the plant in question immediately. He accurately

described its white flowers, climbing habit, and vegetative morphology.

He reported through the translator that he gave this plant to couples

that had difficulty bearing children.

In Minas Gerais, Brazil, catuaba (C. guianensis) is put in

cachaca "to make a man out of you" (Williams, Assis, and Moreira

5416),which the collectors noted that the effect was "Potencia"!


Ornamental Properties


Most of the species of the genus Clitoria have potential for

becoming economically important as ornamentals because of their large

and beautifully-colored, showy flowers and the various habits they

possess. Frequently prized are those species which display climbing

habits, used commonly for garden trellises, and less frequently on

other vegetation. Also much esteemed are those species which possess








the "azure blue" flowers and those of "delicate shades of pink."

Other species are noted for their foliage. A few examples will be

cited here.

The oldest and most widely cultivated ornamental is C. ternatea.

This species is probably indigenous to Eastern Africa with subsequent

expansion into central and western Africa and the Indian subcontinent.

Man has transported this species throughout the tropical belt where it

often has escaped and become established as a wild plant. Today it is

commonly cultivated and is found wild in many Pacific Islands, in

portions of Australia, Indonesia, Africa, and the Americas, and in the

Antilles. Breyne (1678) reported the species as a cultivar, taken

from the Mollucan island of Ternate and planted in European gardens,

where Curtis (1812) reported that it needed protection from the

climate, and that it would not set seed. Because of its brilliant

blue flowers of large size which bloom nearly year round, it is a

popular cultivar in tropical regions, growing rarely as a small shrub,

but most commonly as a perennial climber. The plant grows rapidly,

forming many intertwined stems, and when placed upon open ground,

develops tangled mats. It can be trained through cutting to form

low-growing shrubs. The white-flowered and pale blue-flowered strains

are not as popular as the azure blue-flowered strains.

Clitoria ternatea is also important because of a double-flowered

variety. These plants produce large, actinomorphic, free, banner-like

petals of deep azure to bright purple colors, and rarely white flowers.

This variety was first reported by Commelin (1701). Today it is

commonly cultivated and is naturalized in portions of the Caribbean,

Sri Lanka (=Ceylon), and Indonesia, with isolated populations

elsewhere in the neotropics.








Another early prized ornamental was C. heterophylla which was

reported by Curtis (1820). This species is endemic to the islands of

Madagascar and Mauritius. It is valued for its climbing habit, blue

flowers, and unusual pinnate leaves of two distinct shapes found upon

a single plant.

The most commonly cultivated, erect plant species has been

C. laurifolia (synonym: C. cajanifolia). It is native to the

neotropics and introduced into central Africa (naturalized in Zaire)

and Southeast Asia-Indonesia. In the latter areas, it is cultivated

as a shrub with several plants often planted closely together to

form a hedge.

Two woody lianas from the neotropics are popular in protected

gardens: C. arborescens of northern South America has large leaves,

silky beneath, many racemes of numerous purple flowers, and it is a

tall climber; C. javitensis of South America and Panama is valued for

its climbing habit and numerous lilac to delicate pink shaded flowers.

A less popular liana species, C. lasciva, a native of Madagascar, is

valued for its luxurient foliage and blue flowers.

The tall erect shrub of C. amazonum is desirable in protected

gardens because of its erect habit with upper branches climbing, and

its huge, showy, purplish flowers which are numerous in each of many

inflorescences.

Clitoria fairchildiana (synonym: C. racemosa Benth.) is a

popular tree species. It is endemic to the deltas of the Amazon River

and the other large rivers to the southeast. This tree is planted

extensively along beach roads in Rio de Janeiro, and valued for its

elongated racemes of numerous purple flowers, and for the luxurient








foliage borne on branches which arch and droop to the ground, forming

a canopy.


Poisons and Pesticides


Several species are known to have toxic properties. It has

already been pointed out (Rajan, 1926) that the roots of C. ternatea

taken as a purgative can produce narcotic poisoning in man. Chopra,

Badhwar, and Kapur (1949) included C. ternatea in a list of important

plants poisonous to man and livestock.

Gardner and Bennett (1956) reported that C. arborescens was used

as a fish poison, the seeds containing alkaloids with weak curariform

action. The seeds of a related species from Colombia, C. arborea,

were reported as alkaloid positive (Schultes 24120). In Amazonas,

Brazil, C. froesii was reported as being considered the most toxic

plant in the region of Rio Icana, and was planted by the Indians and

found in fair quantities (Froes 12441/185).

The genus is not noted for its pesticidal properties, although

some species contain the insecticide ingredient rotenone. In Surinam,

the stems and fleshy roots of C. falcata have a strong odor of

rotenone (Archer 2848). A cultivated population of C. laurifolia in

Sao Paulo, Brazil, was noted not to be attacked by fungus or insects

(Norris 278). In Thailand, tuber juices of C. macrophylla are sprayed

on vegetables to kill green flies and root juice is used to kill worms

in the backs of buffaloes (Collins 1441).








Dyes and Fibers


Clitoria ternatea is the only species of the genus that is

reported to be used as a dye plant. Rumpf (1747) and Uphof (1968)

reported that in the Moluccan island Ambon, the flowers were boiled

with rice to give the rice a blue tinge. Burkill (1935) reported

identical use in Malaya. Burkill also noted that the juice of the

leaves is used on some occasions in the Dutch Indies to color food

green. He also reported that the flowers were used as a dye to give a

temporary color to white cloth. The dye was used in the Rhio

Archipelago (=Riouw Archipelogo, a portion of the Malaya Archipelago)

and as a reagent for detecting acid and alkaline solutions, as does

litmus paper.

Liana species are used locally by the natives to make ropes.

Uphof (1968) reported that C. lasciva stems are the source of a fiber

used by natives of Madagascar for making ropes. Clitoria sagotii

was noted as a bush rope by Jenman 4930, in Guyana.


Food for Man


Uphof (1968) reported that the pods of C. ternatea are consumed

as boiled vegetables.


Fodder and Green Manure


A number of references report that C. ternatea is a good forage

crop and is edible by cattle, sheep, and goats. Burkill (1935), for

example, reported that sheep and goats will eat the foliage. In

Venezuela, the species is used as a forage crop (Zambrano 105).

Several experimental studies have reported on the establishment of


L








these species in pastures, and other studies indicate its nutritional

value. These studies will be detailed in the later chapter "Prior

Research on Clitoria," under the section "Agroromical Research."

Burkill (1935) reported that C. ternatea has been used as a green

manure, but that it climbs too much, although it was fairly successful

in some respects, on land that was to be fallow for a short time.

Burkill (1935) reported that C. laurifolia was only fair as a green

manure and that better plants existed for the purpose. Uphof (1968)

reported that this species was used as a jreen manure in warm

countries. Holland and Joachin (1933) found that C. laurifolia was

well adapted as a hedge plant to control terraces on steep slopes in

tea plantations in Ceylon (=Sri Lanka). tBrkill (1935) noted that

the erect stems made this species well suited ,:s a hedge plant on

contoured terraces.















HISTORY OF CLITORIA


The history of Clitoria begins nearly a century before its

officially recognized origin in 1753. Although nomenclaturely

irrelevant, the pre-Linnaean names published and the accompanied

data presented provides better understanding of the genus and its

close relationships with the genus Centrosema.


Origin of the Generic Name


The debatable origin of the generic name has been a very

distasteful subject to many botanists in the past. The most commonly

supposed origin of the name is from the Latin "clitoris," an

anatomical term by which it is indicated that the flower bears a

resemblance to the human female sexual apparatus. This unscientific

notion of comparing the flower with the female reproductive organs, a

practice noted by some as one of questionable taste or immoral

behavior, has been disturbing to a number of botanists. Solutions to

this problem have been various. First, ignore the problem by making

no reference to the subject. Second, circumvent the issue by

concluding that the origin of the name is lost in history, and today

has no significant meaning. Third, gently touch upon the problem by

indicating that the origin is from the flower's fantasized resemblance

to "a female organ," or to "a female structure," or to "an organ."

Fourth, circumvent and attack the problem by indicating that the


I








origin is from the Greek "cloisto: to enclose' which refers to the

manner in which the organs of fructification ate enclosed within the

keel. Or one can use as the origin, Clitor, a town in Arcadia

(Stearn's "Botanical Latin," 1966, has Arcddia = Arkadhia,

Peloponnisos, S. Greece, a European country side outside the

distribution of the genus). Fifth, a direct attack upon the problem

by indicating the immoral practice and unscientific method of

selection of the generic name, which should be selected to reflect

some botanical structure, to reflect a geographic origin, or to honor

a distinguished person by naming the genus after him. Hence the genera

established as substitutes, Ternatea Tourn. for Clitorius Petiver, and

Nauchea Desc. and Vexillaria Eaton substituted for Clitoria L.

To this author, it seems clear that tHe o0 igin of the name can be

traced to Breyne (1678) who published Flos clitoridis Ternatensibus

for an exotic cultivar from the Moluccan island of Ternate. He

indicated the vernacular name to be "Bokyni cotele de Principisse

Clitoris." Rumpf (1747) expanded upon this by indicating the Ternatice

names to be "Saja Cotele" and "Bokyma Cotele" Irom which Breyne erected

Flos clitoridis (flowers clitoris-like) based upon clitoris principisse

(or origin from clitoris), and from which the Portuguese name, Fula

criqua (to speed or hasten breeding or creation) were derived and

appropriately named. The Portuguese and Ternatice names were probably

based upon the ancient "Doctrine of Signatures," a belief that plant

structures which resembled human structures were created to furnish

remedies for the ailments of those body structures. A number of

Clitoria species are used as aphrodisiacs, both for man and for

animals, in widely separated geographical -reas and cultures. Breyne








evidently recognized a resemblance in the flower to the female anatomy

as others hkid, because ne iin:orporated the con rpt that the flowers

were clitoris-like in his polynominal name


G neral Histor,,


Breyne (1678), a German botanist, was the first to describe and

to illustrate d representative of the genu, Clitoria. In "Exoticarum

Plantarumn Cnturia Priiia" he described a plant, from the Indian

subcontinent and transferred to Europe for cultivation, as Flos

clitoridis ler!natensibus. SBieyne describe I th- showy papilionaceous

flowers as inverted, with the vexillum ventrali/ located, a unique

characteristic that segregates Clitoria fmin n mt other legumes.

However, in tho excellent illustration (pi te I, drawn by Stech1),

the flowers ar r not resipinale (possibly .:, arl ist's or engraver's

error). The v,,xillum appear ;d dorsally a, in typical papilionaceous

flowers. 'Breyne's description included hi,,it, calyx, bracteoles

(Breyne used the term "leafl ts"), and fruit (i.ot illustrated), major

characters sedl by later authors in segreg tini Clitoria species. The

illustration clearly shows tie odd-pinnat( y c poundd leaves and

persisting bra:ts and sLipuls (the stipul *s ai e drawn too large for

the species), additional characters used .irly in the determination of

species. Breyne's represent.ition is easily recognized as Clitoria

ternatea L.. the type species for the genu .reyne reported the plant

from the Moluccan island of Ternate, and ited that it bore the name


1. Hunt Iot. Cat. 1:37.1 attribute d ,winc, to Andreas Stech and
engraved by Issac Saal. Plate bears initi Is "AS" at left bottom,
"IS" at rigit iottoii.








of "Bokyni iotele de Principisse Clitoris.' the source for Breyne's

polynominal name. Breyne reported a coll, tioi by Sevenhuysen in

1667 from Domiingo (=Doiiinican Republic) a: related to the Ternate

specimen. If verified as a dlitoria, it would d represent an earlier

record of tie genus.

Breyne also described and illustrated (plate 32) a second

representative as Plantt Legiminosa Brasiliana, Phaseoli facie, flore

purpureo maAiro. This specimen provided Ire earliest record of the


species known as Clitoria braisiliana L., 1

becoming the type species of the genus Cen

Plukenat (1691) in his 'Phytographia'

Foenum1 grae. u!iin phaseoloides ;'irginiana fle

the earliest record of Lhe species known o

later corre tly transfer' red LO the genus (

Petiver (1704) in his Appendix to hi.

Siccis Petiverianus" published two polynoh

data. His number 54 was Cli :orius Marianu,

record of this polynominal has been found

number 55 w.s Clitorius Marinnus trifolia(

the earliest record of the species known n.

was the first to use Clitorii or a form oi

"generic" nine.

Tourneforc (1706) in hi, article "Sui

queques Nouveaux Genres de Plantes" descri


iter transferred to and

Lrost ,a.

described and illustrated

'e aiiplo caeruleo. This is

Cl toria virginiana L.,

ntr semla.

"Ca' logus Plantarum Hortis

nal both lacking additional

trifoliatus viridis. No

iteo after this date. His

I sibtus glaucus. This is

Cl toria Iiariana L. Petiver

the basic root word as the



ie di L'etablissement de

,ed the genus Ternatea. He


provided an illustration of the flower wil cal/x, individual petals,

fruit, and :,eels. Thre.t meniiers of the g,us 'are included. His

first species was Ternaea flore sim plici, caeijleo based upon Breyne's








specimen. Breyne's specimen became the type for the genus and the

source of the generic name, which Tournefort indicated was named after

the Moluccan island of Ternate. Tournefort's second species was

Ternatea flore pleno, caeruleo based upon a double-flowered specimen

described by Commelin in 1697. His third species was Ternatea flore

simplici albido, a white-flowered specimen recognized for the first

time. All three species are now Clitoria ternatea L.

Dillenius (1732) in his "Hortus Elthamensis" described and

illustrated Clitorius trifolius, flore minore caeruleo. He segregated

this specimen from those of Breyne, Petiver and Tournefort, and

concluded that it represented a new member of the genus Clitorius as

defined by Petiver. His specimen is Clitoria virginiana L., although

no reference was made to Plukenet (1691), the earliest describer of

this species.

Linnaeus (1737b)in his "Hortus Cliffortianus" became the first to

use the generic name Clitoria. He recognized two species. His first

species was Clitoria foliis pinnatis, based upon specimens described

by Breyne (1678), Rheede (1688), Plukenet (1691), Rivinius (1691)

Commelin (1697), Tournefort (1706), Dillenius (1719), and Burman

(1737). Breyne's Flos clitoridis ternatensibus was the source of the

generic name. Neither Petiver (1704) nor Dillenius (1737), both of

whom had used Clitorius, was cited. The diagnosis was based upon

Phaseolus foliis pinnatis Riv. The distribution was given as

Malabaria (=S.W. India), Zeylona (=Ceylon, now named Sri Lanka), and

Ternate. Linnaeus recognized three varieties. His typical variety had

simple blue flowers (=Tournefort's first member). His variety alpha

was white flowered (Tournefort's third member). His variety beta had








double blue flowers (Tournefort's second member). In addition,

Linnaeus gave a brief Latin description of variety beta. His second

species was Clitoria foliis ternatis based upon Breyne's Brazilian

specimen and Plumier (1693).

Linnaeus (1748) in his "Hortus Upsaliensis" recognized the same

two species, but used binominal names. His first species Clitoria

foliis pinnatis became Clitoria ternatea. His species Clitoria foliis

ternatis became Clitoria brasiliana.

Nomenclaturally, the official history of the genus began with

Linnaeus (1753) in his "Species Plantarum." Linnaeus established the

genus Clitoria (p. 753) with four species: C. ternatea and C.

brasiliana from his earlier works (1737band 1748), with the addition

of C. virginiana (based upon Dillenius, 1737, and Gronovius, 1739) and

C. mariana (based upon Petiver, 1704). His brief Latin diagnosis for

each species was based upon leaflet number and calyx shape. India,

Brazil, Virginia, and northern America were given as the ranges of

his four species.

The International Code of Botanical Nomenclature (Stafleu, 1972)

officially recognizes that those generic names which first appear in

first edition (1753) of Linnaeus's "Species Plantarum" (e.g. Clitoria)

are associated with the first subsequent description given under those

names in the fifth edition (1754) of Linnaeus's "Genera Plantarum."

Clitoria is described under genus number 796 (p. 334) of Linnaeus's

"Genera Plantarum" as:


CAL. Perianthium monophyllum, erectum, tubulatum,
quinquedentatum, persistens. COR. Vexillum maximum
rectum, emarginatum, margin undulatum, patens. Alae
oblongae, rectae, obtusae, vexillo breviores. Carina








alis breviores. STAM. Filamenta diadelpha (simplex
& novemsidum). Antherae simplices. PIST. Germen
oblongum. Stylus adsendens. Stigma obtusum. PER.
Legumen longissimum, lineare, compressum, uniloculare.
SEM. plura, reinformia.


This diagnosis represents the first validly published description

of the genus Clitoria L.

Miller (1754) in his "Gardner's Dictionary" provided the first

encyclopedic reference work on the genus. He adapted Tournefort's

treatment (1706) and recognized the genus Ternatea. Thus, three of the

four species he recognized followed Tournefort's treatment, not

Linnaeus's (1737) who treated these as three varieties of one species.

A fourth species, Ternatea Americana perennis flore caerulea from

Jamaica, was included. In a later edition of his work, Miller (1759)

adopted Linnaeus's generic name Clitoria and his four species. The

three species of Tournefort included in his earlier edition were

included as C. ternatea in the later edition. The treatment of the

Jamaican species is unclear, as no reference to this specimen was

made. It may have been included under C. brasiliana or C. virginiana,

or treated under another genus, but it could not have been included

in C. mariana, a species which does not occur in Jamaica.

Miller described the flower as papilionaceous, with the vexillum

longer .than and hiding the keel and wings, the fruit opening by the

sutures, and the seeds kidney-shaped. The first description of the

stamens and stigma were given. He referred to the flower as a

butterfly type.

Buchoz (1775) in his "Histoire Universelle du Regne V6egtal ou

Nouveau Dictionnaire Physique et Economique de toutes les Plantes qui

Croissent sur la Surface due Globe" became the first to describe the








genus Clitoria in greater detail. Clitoria was described as: calyx

tubular, 5-toothed, persistent; corolla papilionaceous; vexillum very

large, straight, notched, with undulating margins; wings oblong,

straight, obtuse, much shorter than standard; carina much shorter than

wings, round underneath, furrowed; stamens diadelphous, anthers simple;

ovary oblong, style ascending, stigma obtuse; legume very long, linear,

flat, bivalved, awl-tipped; seeds numerous, reniform. Buchoz followed

the Linnaeai treatment and included five species, C. ternatea L.,

C. brasiliana L., C. virqiniana L., C. mariana L., and C. galactia

Crantz. Each species ,was described with notes on origin, culture, and

economic importance.

Lamarcl. (1786) in his "incyclopedie tiethodique, Botanique"

described seven species, of which Clitorid heterophylla (East Indies)

and Clitori, falcata (Santo Domingo) were newly described.

Description, lacked measurements, but included habit, stem, leaf

shape and pubescence, leaflet number, peduicle location and relative

size as compared to petiole size, calyx, flower size and color, and

fruits. All sizes were relative and compared to similar structures of

other species. These characters, along with the addition of bracteoles

and stipules, were to become the major characters utilized by later

authors in describing and segregating Clitoria species. Lamarck's

style of describing Clitoria species established a trend that continued

through Bentham's revision of the genus in 185' and beyond. Poiret

(1811), in supplement to Lamarck's work, described three new species

and brought the total number to thirteen.

Desvaux (1813) in his journal article "Precis de Caracteres de

plusieures ienres, de la Famille des Legui. ineuses" described a new








genus Neurocarpum, with a tubular calyx of five, equal teeth and a

stalked fruit characterized by a longitudinal nerve on each side. He
2
indicated Aublet's Crotalaria gajanensis (published by Aublet as

Crotalaria guianensis) and a new species, N. ellipticum Desv., were

placed in the genus. In a succeeding article a year later, Desvaux

(1814) described the new species N. ellipticum and included

Crotalaria guyanensis Aubl. and Crotalaria longifolia Lam. as

synonyms of a species called Neurocarpum janense Desv.

Eaton (1817) in his "Manual of Botany for the Northern States"

established the genus Vexillaria as a substitute for Clitoria because

J. E. Smith severely censured the name Clitoria (reference of the name

reflecting the female anatomy) in Rees' "Cyclopaedia" (1807).3 Rickett

and Stafleu (1959), in an expanded footnote regarding Vexillaria Benth.

(=Centrosema), quoted Smith as:


CLITORIA, in Botany, (from XAElw, claudo, include),
expressing the manner in which the essential organs
of fructification are enclosed or shut up in the
keel and wings of the corolla. Whatever may have
been in the thoughts of Petiver, by whom the name
was first introduced into botany; or of the
illustrious naturalists, by whom it has since been
continued, reformed, or sanctioned, we cannot refrain
from entering our decided protest against every
attempt to associate it directly with an anatomical
term, to which, though derived from the same Greek
theme, it has in fact only a very remote,


2. Desvaux erred in his interpretation of Aublet's species.
Desvaux's type (P-59!) does not agree with Aublet's type (BM!).
Therefore, Aublet's name is not synonymized with Desvaux's name. A
complete discussion can be found under the type treatment of subgenus
Neurocarpum, p.
3. This author has not been able to obtain a copy of Eaton's
publication, nor Smith's article. However, a copy of the eight edition
of Eaton's Manual (1840) has been obtained. Clitoria was not mentioned.
Vexillaria Eaton contained three species, V. virginiana, V. mariana,
and V. plumieri, each treated as a Clitoria species by almost all other
botanists.








fictitious analogy. It is greatly to be lamented,
that a fondness for these gross allusions should
ever have been indulged by any, who, in all other
respects, have deserved highly of natural science,
and whose splendid talents should have rendered
them far superior to such grovelling ideas. By
this conduct they may have done all in their power
to pollute a study, which is, perhaps, more than
all others, suited to the loviest part of the
human race, and which, without concealing any
essential part of the sexual system, may easily be
so conducted, as not to excite an unpleasant
sensation in the most delicate female mind. We
do not mean to exempt from the full severity of
this censure our great master, Linnaeus, himself;
for when, not only the purity of moral feeling,
but also the common decorum of polished life, is
infringed, the nullis in verba of the poet will, we
trust, be uniformly our principle and our practice).


Vexillaria Eaton does not appear in Willis' Dictionary, Index Kewensis,

Gray Card Index, nor other standardized index references, nor in

articles associated with Clitoria. Vexillaria Eaton apparently was

not accepted by other botanists of the time, and was thus easily

overlooked by the compliers of the major index references. Vexillaria

Eaton is a probable synonym of the genus Clitoria.

Rafinesque (1818) in a two-line note in a journal article remarked

"Clitoria mariana must form a particular genus Vexillaria." Thus

Vexillaria Raf. was established and recorded by the major standardized

index references. Other botanists did not agree with Rafinesque and

neither. Vexillaria mariana Raf. was used nor were other species of

Clitoria transferred to the genus Vexillaria Raf.

Leandro do Sacramento (1821) in a journal article described a new

species under a new genus Martia, named in honor of the German botanist

Carl F. P. Martius. The description of Martia physodes was detailed

with an illustration provided. The generic characters which indicated

a distinct genus were the lack of a corolla, small stamens with








rudimentary filaments, a style in contact with its own anthers, and a

legume with a protruding lateral nerve. The description was based

upon a cleistogamous flowering collection of Clitoria falcata from

Brazil.

Schultes (1822) in his "Mantissa in Volumen Primum Systematis

Vegetabiliuii" published the species of Leandro do Sacramento under the

name Martiusia physalodes. Perhaps Schultes changed the generic name

because of the homonym Martia Sprengel (1819) which is synonymous with

a non-legume genus.

Humboldt, Bonpland, and Kunth (1824) in their "Nova Genera et

Species Plaiitarum" described four new species of Neurocarpum. A new

genus was cited in synonymy as "Rhombolobium Rich. mss.," probably

based upon a Paris specimen of Richard (P- 2!) whose label bears the

generic description of Rhombolobium, and less likely based upon a

Geneva specimen of Richard (G-545!) whose label bears the data

"Rhombolobiumj Richard, ijenre nouveau qui doit etre place a coti du

Clitoria. 11 y a un angle que [handwriting illegible for this

term] du milieu du Legume du chaque coti." Clitoria L. and Ternatea

Tourn. were treated as distinct genera by Humboldt, Bonpland, and

Kunth.

A. P. de Candolle (1825) published the first major revision of

Clitoria in his "Prodromus Systematis Naturalis Regni Vegetabilis."

Almost published simultaneously was his 'Ti iijires sur la Famille des

Legumineuses" which gave explanations of his reasons underlying the

classification adopted in his "Prodromus." A. P. de Candolle

characterized the genus Clitoria as: caly-, of five teeth with a pair

of bracteoles at its base; the corolla and stamens inserted at the








calyx base at a point visible to the exterior; a large round standard;

a style more or less dilated at the summit; a resupinate flower; and

a fruit of two valves with seeds separated by a cellular partition.

The genus Ternatea Tourn. was placed in synonymy. A. P. de Candolle

recognized four sections within the genus, segregated by the calyx

shape, bracteoles, presence or lack of a vexillum spur, and a leaflet

number.

A. P. de Candolle called section I "Ternatea," based upon Kunth's

treatment of Ternatea in 1824. This section was characterized by a

tubular calyx, spurless vexillum, and five or seven leaflets. Two

species were included, Clitoria heterophylla Lam. and Clitoria

ternatea L. A. P. de Candolle named section II "Euclitoria," based

upon his statement that this section included the true clitorias.

Section II was characterized by a tubular calyx, spurless vexillum,

and three leaflets. Six species were included: Clitoria mariana L.,

Clitoria mexicana Link, Clitoria angustifolia H.B.K., Clitoria

formosa H.B.K., Clitoria glycinoides DC., and Clitoria poitaei DC.,

the latter two newly described species. It should be noted that the

type species for the genus, C. ternatea L., is not in this section,

and thus the section cannot contain the "true" clitorias. Section III

was named "Centrosema" by de Candolle, based upon the spurred vexillum,

a unique characteristic in the legumes. Section III was characterized

by a campanulate calyx of five divisions, a spurred vexillum,

longitudinally striated bracteoles, and three leaflets. Three species

were included: Clitorid virginiana L., Clitoria brasiliana L., and

Clitoria plumieri Turp. Section IV was named "Glycinopsis" by

de Candolle, based upon the resemblance to the genus Glycine.








Section IV was characterized by a campanul.te calyx of five teeth,

corolla and stamens inserted at the calyx uase, style dilated at

summit, longitudinally striated bracteoles, and three leaflets. One

new species, Clitoria berteriana DC., was placed here. A. P.

de Candolle also noted five additional species that may possibly

belong in the genus Clitoria, bringing the total number of species

included to seventeen. The five species that were briefly noted, and

not placed in any of the four distinct sections, were: Clitoria

coccinea Schrad., Clitoria speciosa Cav., Clitoria laurifolia Poir.,

Clitoria vicioides Nees. & Mart., and Clitoria arborescens Ait.

For each species placed in a section, de Candolle gave a brief

Latin description in his "Prodromus." He included the general

distribution of the species dnd cited collections he had seen for

the new species.

A. P. de Candolle treated the genus Nieurocarpum Desv. as distinct

from Clitoria L. in his "Prodromus" following the tradition of other

botanists of his time. He characterized the genus Neurocarpum as:

calyx tubular with five, subequal, acuminate teeth bearing two

bracteoles at its base; vexillum large, subrotund; stamens diadelphous,

legume stipitate, compressed, subtetragonous, valves bearing a

prominent mnidial longitudinal nerve; and with seeds in cellular

partitions. Seven species were included: Neurocarpum simplicifolium

Kunth, NeurLcarpum angustifolium Kunth, Neurocarpum guianensis Desv.,

Neurocarpum ellipticum iesv., Neurocarpum javitense H.B.K., Neurocarpum

macrophyllum H.B.K., and Neurocarpum falcatum (Lam.) DC. A. P.

de Candolle included a new genus, Rhombifolium Rich., in synonymy

based upon a specimen iii Richard's herbari im.








A. P. de Candolle treated the genus Martiusia Schult. as also

distinct from Clitoria and Neurocarpum. He characterized the genus

Martiusia as calyx slightly bilabiate with five teeth, lowermost

longer; corolla lacking; stamens four with two sterile and two fertile,

subciliate anthers; filaments much shorter than the ovary; legume

stipitate, compressed, subtetragonous, bearing a medial longitudinal

nerve. A. P. de Candolle noted that the calyx and legume were those

of Neurocarpum, and that the genus may not be distinct from Neurocarpum.

He included only one species, Martiusia physalodes Schult. The genus

Martia Leandr.-Sacr. was treated in synonymy.

Descourtilz (1826) in his journal article "Sur les Nauchees,

Genre Nouveau dans la Famille des Legumineuses" criticized the use of

the name Clitoria and its reference to an organ from which the name

was derived. He felt that the botanical language should consecrate

the memory of those illustrious persons of importance by adopting

names which recognized their contributions. He proposed the

substitution of the name Nauchea for Clitoria. The genus Nauchea was

named in honor of M. fe docteur Nauche.4 Twelve species of Clitoria

were transferred to Nauchea and a new species was described. Each

species included a description and a distribution. Descourtilz gave

each species a common name, based upon the specific epithet (i.e.

Nauchea multiflora was "Nauchee a fleurs nombreuses"). Other botanists

did not adopt the generic name of Nauchea.

Desvaux (1826) in his article "Observations sur la Famille des

Legumineuses" proposed a revisionary change in the genus Neurocarpum.


4. A prominent French physician, Jacques-Louis Nauches (1776-
1843), who pioneered work in electrical shock therapy.








Two species were placed near those described in 1813. They were

Neurocarpum barbatum Nees, newly describe(, and Neurocarpum laurifolium

Desv., a transfer of Poiret's species (thus should have been N.

laurifolium (Poir.) Desv.). Three species were recognized as different.

They were Neurocarpum rubiginosum Desv., a transfer of Jussieu's

species published by Persoon (thus should nave been N. rubiginosum

(Juss. ex Pers.) Desv.), Neurocarpum glycinoides N.,5 a transfer of

de Candolle's species (thus should have been N. glycinoides (DC.)

Nees), and Nleurocarpum villosum N., newly Jescribed. Desvaux proposed

the name Pilanthum for this group based upon Puiret's name used for

the type of C. gycinoides DC. (=N. glyciilides Nees). The genus

Pilanthus Puir. ex End]. (Genera Plantarum p. 1289, 1841), presently

synonymized with the genus Centrosema, may be based upon Poiteau's

name Pilantlhos, although Endlicher's article has not been seen. The

specimen referred to by Desvaux and also b/ de Candolle (Pilanthum

cited under C. glycinoides in his Prodromun 2:234, 1825), bears the

name Pilanthos tetragonus Poiret. Two new species of Clitoria were

also described, C. laurifolia Nees, a homonym for C. laurifolia

Poiret (1811), and C. sinuata Nees.

Don (1832) in his "A General History of the Dichlamydeous Plants"

followed de Candolle's treatment of Clito ria, Neurocarpum, and Martiusia

with minor revisions. Of the five dubious species listed by

de Candolle under the uenus Clitoria, Don laced two in the section

Centrosema (C. speciosa Cav. and C. arborescens Ait.), one in the

section Glycinopsis (C. coccinea Schrad.), and left one in a dubious


5. Desvaux's use of N. was an abbreviation for Nees.








list (C. vi(ioides Nees & Mart.). The filth species was transferred

to the genus Neurocarpuii (C. laurifolia Poir.) along with C.

glycinoides DC. from section Euclitoria. lie also described two new

species, Clitoria raceimosa and Clitoria alba, and placed them both in


the section Euclitoria.

and Martiusia with the

Bentham (1837) in

de Candolle's treatment

There were three major

segregated the section

to the level of genus,

botanists have adopted

century later, botanist

and Centrosema (DC.) Be

Glycinopsis by transfer


No changes were iimde in the genera Neurocarpum

exception of the transfers already noted.

a journal article on legume genera followed

of these genera with some revisionary changes.

changes in the genus Clitoria. First, Bentham

Centrosema from Clitoria and elevated the taxon

using de Candolle's name Centrosema. Succeeding

this revision, and presently, more than a

s continue to reco!;gize both genera, Clitoria L.

rith. Second, Benthdm eliminated the section

ring Clitoriana berteriana DC. to a


non-clitoritn genus, Periandra, and placed Clitoria coccinea Schrad.

in a dubious status. Succeeding botanists adopted this revision as

the section Glycinopsis disappeared from literature after this date.

Third, Bentham recognized de Candolle's sections Ternatea and

Neurocarpum, and added to them a new section, Bractearia Mart.ex

Benth. Bracteria was described as fruticose species with trifoliate

leaves and bracteoles equal to or longer than the calyx. To this new

section, Bentham transferred Clitoria racemiosa Don and Clitoria

poitaei DC. He also described three new species and placed them in

this section, Clitoria amazonum, Clitoria acuminata, and Clitoria

arborea.








Benthain's treatment of section Ternatea followed de Candolle and

Don with the addition of a new species, Clitoria lasciva Boj. ex Benth.

Section Euclitoria underwent minor revision. Clitoria formosa H.B.K.

and Clitoria angustifolia H.U.K. were transferred to the genus

Centrosema. Bentham added a newly described species, Clitoria

pedunculata Boj. ex Benth., to the two remaining species, Clitoria

mariana L. and Clitoria mexicana Link. Bentham also added notes on a

number of miscellaneous species published as Clitoria, but which had

been transferred to other genera.

Benthain's treatment of the genus Neurocarpum followed de Candolle

and Don with two important changes. First, Bentham placed the genera

Martia Leandr.-Sacr. and Martiusia Schult. in synonymy. Second,

Bentham remarked on the close relationship between Neurocarpum and

Clitoria by stating that Neurocarpum is a Clitoria except that it has

a different legume (i.e. Neurocarpum has a costate legume; Clitoria

has an ecostate legume). Eleven species ware recognized and placed in

two unnamed, artificial groups. Group 1 iid suberect stems whereas

group 2 had procumbent or climbing stems. Five species were newly

described and several species were placed in dubious status.

Bentham (1839), in a publication on British Guianan plants

collected by Schomburgk, substituted the nime Dendrocyamus for his

section Bractearia in the genus Clitoria. He reasoned that the name

Bracteria was inappropriate since it had been given to a genus of the

Rubiaceae and had been used as a section of the genus Chaetogastra

(Melastomaceae).

Bentham (1858) published a second revision of the genus Clitoria

in a Journal of the Linnean Society article. This was the last revision








of the genus until this work over a century later, and was adopted by

all succeeding botanists in their treatment of the genus. The major

change from his earlier revision (1937) was the uniting of the genus

Neurocarpum with Clitoria. Bentham reasoned on p. 35:


In a systematic point of view I had formerly
endeavoured to render Clitoria more natural, by the
elimination of DeCandolle's section Centrosema; and
I now find it necessary for the same purpose to unite
it with the Neurocarpum of Desvaux, hitherto universally
adopted by other botanists, myself included. This
entails the giving up, as a generic character, one which,
in Leguminosae, is usually considered as absolute, the
raised longitudinal nerve or wing along the center of
each valve of the pod. It is the same peculiarity which
has induced the separation of Tetragonolobus from Lotus
among European plants. But in the division of Lotus, as
well as in that of Clitoria, this purely technical
character is unaccompanied by any other differences, and
I now have instances in Clitoria where it is inconsistent
in one and the same species, and even on the same
specimen.


Bentham continued to recognize three sections, although he changed

their names. Section Ternatea remained unchanged, whereas Neurocarpum

was substituted for Euclitoria, and Clitorianthese was substituted for

Bractearia (=Dendrocyamus in 1839) used in Bentham's earlier revision

(1837). No comment was made on the change in section names, nor was

there any reference to the earlier sectional names. However, a

comparison of the two revisions by Bentham indicated that the sectional

diagnosis agreed (although the 1837 description was expanded upon in

1858), the species placed in each section agreed, the order in

which each section was treated agreed, and the number given by Bentham

for each section agreed.

Section Ternatea was characterized as: stems herbaceous or

rarely fruticose, prostrate, twining or climbing; leaflets 5-9, rarely








3; legume valve flat or slightly convex, non-costate, seeds subreniform,

compressed, smooth. Six species were recognized, all known previously.

They included Clitoria lasciva Boj. ex Benth., Clitoria ternatea L.,

Clitoria pilosula Wall., Clitoria heterophylla Lam., Clitoria biflora

Dalz., and Clitoria pedunculata Boj. ex. Benth.

Section Neurocarpum was characterized as herb, twining or

prostrate or short erect; leaflets 1 or 3; legume valve convex, medial

longitudinal costa or rarely ecostate; seeds globose, ovoid, or thick

subreniform, with a viscid coat. This section contained eleven species

placed into three groups based upon the stem habit. The groups were

unnamed and distinguished by one to three "*" symbols. The first group

with twining stems included: Clitoria macrophylla Wall., Clitoria

mariana L., and Clitoria glycinoides DC. The second group with

prostrate stems included two newly described species: Clitoria

flagellaris and Clitoria rufescens. Group three with stems arising

from a lignose rhizome, erect to ascending, included: Clitoria

cajanifolia Presl, Clitoria simplicifolia (Kunth) Benth., Clitoria

guianensis (Aubl.) Benth., Clitoria densiflora (Benth.) Benth., and

Clitoria stipularis (Mart.) Benth., plus a newly described species,

Clitoria nana.

Section Clitorianthes was characterized as erect shrub or tall

climbing; leaflets 3; legume valve flat or slightly convex, coriaceous;

seeds in matured state unknown. This section contained nine species

divided into two unnamed groups based upon the bracteoles. The first

group with narrow bracteoles, or bracteoles much shorter than the calyx,

included Clitoria polystachya Benth., Clitoria brachystegia Benth.,

Clitoria arborescens Ait., Clitoria javitensis (H.B.K.) Benth., plus








two newly described species, Clitoria sellui and Clitoria leptostachya.

The second group with ovate, coriaceous bracteoles subequal to the

calyx, included: Clitoria aiiazonum Mart. ex Bunth., Clitoria racemosa

Benth., and Clitoria hoffmanseggii Benth.

The treatment of each species included a Latin diagnosis, the

distribution as then known, the citation ol one to a few collections,

and synonymy/. Characters used in the diagnosis of the species

included: stem habit; leaf shape and pubescence, and leaflet number;

peduncle size, location, and number of flowers borne, calyx;

bracteoles; legume curvature and costa; ard sometimes stipules or

vexillum pubescence. -enthaii occasionally gave additional descriptive

data with a minimum quantity of measurements.

After lentham's revision, the next century of publications on

Clitoria was minly of a floristic nature, not revisionary. Isolated

descriptions of newly described species placed in the genus also

occurred. Two exceptions to the lack of revision are worth noting.

First, Baker (1879), in a floristic treatment of legumes in "Hooker's

Flora of British India," elevated Bentham's sections to the level of

subgenus. Subgenus Ternatea was characterized as having flat legumes

without a costa. Clitoria mariana L. and Clitoria macrophylla Wall.

were transferred here from Bentham's section Neurocarpum. Subgenus

Neurocarpum was characterized as having a turgid, costate legume.

Bentham's section Clitorianthes was not treated, probably because the

species within this section do not occur in the Paleotropics. Second,

Kuntze (1891) transferred all Clitoria species to the genus Ternatea,

a revision not adopted by later botanists. Occasionally, floras would

include Clitoria and Martiusia as separate genera, the latter genus








containing the species of Neurocarpur (e.jg., Small, 1933; Britton,

1924).

In the earlier parr. of Lhe twentieth century, research in various

areas began on Clitoria (e.g., agronomy, anatomy, chemistry, cytology,

development). The results of these studies are enumerated in the

chapter "Prior Research on Clitoria."

These selected articles represented a partial, highlighted history

of the genus Clitoria. The important portions of the nomenclatural

history-of the genus, above Ihe level of species, are summarized in

Figures 1 ard 2. For both figures, genera were expressed in capital

letters and enclosed in a "box" for better visualization. Genera

included in synonymy were enclosed in parentheses. For the few

polynominals included, ihich are important in tracing the origin of

some generit- names, the key terms are used and underlined, followed

by "..." which represented the missing, nonessential terms. Sections

are numbered and subgenera are preceded by the abbreviation "Subg."


Historical Chronol gy


A more complete historical chronology is presented below. Each

article is presented in a chronological order, by year, with a brief

synopsis of the publication, and nomenclatiral notes. Those articles

marked (*) are expanded upon in the "General History" section

previously discussed, and the reader is referred to its coverage.

Those articles marked (R) are expanded upon in the next chapter, "Prior

Research on Clitoria," and the reader is referred to that chapter for

details.

















SE 0
_0 4-) S






o ro
-- -r -
o ro o o o





















00 < .
J LU __jC






t \-



4-_
Mo *r- o 00 C
-C \ < : :-.
2c)






O 0.-. Q CAD
01















LL
CD \ Q.4- c

I- C CD- o 021) -
) of) C- =U 4- E >
C- u -u co

ff H- C C+ Q\

LLJ CO i\





















w -. ro ,r
H- ro < g .
-- U LU ar )6

L Co I- to -








0 C
t ct o to o


0f C- 0ot -1 c










4 a) L \ L
.1S-o
r-_ \ \ in













0 4 I r.-


r- L- 0 r <_ oII t- nc4



o rU to C t
4-' C ~ 4- a
C* U u-







43





E
S- q
I 00
0- I
-1 -r
-0 0 IQ ---

SO 0 2 0
N LVU L I 0-3
) I- |< T ) C) -I
S 'dC _d^* 0 H- -z.
(0TJ I


-- 0) |
4-- CO F- 2-.-


LLU








", r---r
I 0
















:7 +* ;


n.. -" .-











4 --' u 4i-
CD 2T r C 0D


CD U
S- ..- -











C cc I









{. -2 s- t'- C C
--l D I )
U) < ^ O












O4-
L U L U J CL







C C-.















ro-


H- Ln (
z c -- 4 u
LU (0 I CZ

H- E 0 0r(








(0 I LU

























O-
Q 0 ; r0 '0 0L

' CD (I / i
0 '- H- 1- U
Lrl LU LU LU 0 r o




I I ^ -











lt L







4'L r 1








Pre-Linnean (1650-1753: _Po/lnominal Peri;l)

1667: Sevenhuysen was reported by Bieyne (1678) to have collected

a specimen Irom Santo Domingo that was related to the Moluccan specimen

Flos clitoridis ternatensibus. If this specimen is verified as a

Clitoria, this would be the earliest known record of the genus.

1678: Breyne (*) described and illustrated two specimens that are

the framework upon which Linnaeus established two of the original four

species of Clitoria. His Table 31 described Flos clitoridis

ternatensibus (=Clitoria ternatea L.) froii the island of Ternate,

Moluccan Islands. His Table 32 described Planta leguminosae

Brasiliana, phaseoli fcie, flore purpureo maxirno (= Clitoria

brasiliana L.).

1688: Rheede tot Draakestein described Schonga cuspi from Malabar

(=S.W. Indid) which Linrnaeus later cited in synonymy for Clitoria

ternatea. From its description and a photograph of Schongi cuspi as

illustrated in his Table 38, one can easily conclude that this plant

is not a Clitoria specimen.

1691: Plukenet published Foenum graecum phaseoloides Virginianus

flore amnplo caeruleo which is the origin of the third original Linnaean

species, Cl itoria viriniana L.

1693: Plumier described Phaseolus am;ple flore peltato, siliquis

nigris et anqulosis (=Clitoria plumieri Turp.) based upon a specimen

from Santo Domingo.

1696: In his catalogue of Jamaican plants, Sloane included

Phaseolus minor lactescens flore purpureo (=Clitoria galactia Crantz).

1698: Dr. David Kreig brought a specimen to Great Britain that

was collected in Maryland. [his specimen is now deposited in the Royal








Botanic Garden Herbarium at Edinburgh (E-127, Herb. Dubois!). This

specimen represents the oldest known representative of the genus

Clitoria (=C. mariana L.).

1700: Plukenet published Phaseolus Indicus, coeruleus,

Glycyrrhizae foliis alatis, flore amplo Cliterio (=Clitoria ternatea

L.) and Phaseolus flore vexillo, silique rectis teretibus (= Clitoria

brasiliana L.).

1701: Commelin described and illustrated in his Table 24,

Phaseolus Indicus, glycyrrhizae foliis, flore amplo, caeruleo, pleno.

This was the first known example of the unique double flower of

Clitoria ternatea L.

1703: Dr. Edward Bulkley sent a specimen to Great Britain that

was collected from Fort St. George in the East Indies. This specimen

is now deposited in the Royal Botanic Garden Herbarium at Edinburgh

(E-69, Herb. Dubois!). This specimen represents the oldest known

specimen of the type species of the genus Clitoria (=C. ternatea L.).

1704: Petiver (*) published two polynominals in an appendix, the

first to use "Clitorius" as the leading key term of the polynominal.

His number 54 was Clitorius Marianus trifoliatus viridis which

disappeared from Clitoria literature. His number 55 was Clitorius

Marianus trifoliatus subtus ajaucus which was the origin of the fourth

original Linnaean species, Clitoria mariana L.

1706: Tournefort (*) established the genus Ternatea based upon

a specimen from Ternate, a Moluccan Island. Three elements were

published which later served as the framework for the three varieties

of Clitoria ternatea L. These were Ternatea flore simplici, caeruleo








(agreed with Breyne, 1678); Ternatea flore pleno, caeruleo (agreed

with Commelin, 1701); and Ternatea flore simnplici albido.

1709: Petiver published Plukenet's rime (1691) and illustrated

a specimen in his Table 104, Figure 19.

1732: Dillenius (*) described and illustrated Clitorius trifolius

flore minor caeruleo (=Clitoria virginiana L.) which he concluded was

segregated from, yet related to the specimens of Breyne (1678) and

Petiver (1704). No reference was made to Plukenet (1691).

1737: Burmann reported Flos clitorius flore coeruleo (=Clitoria

ternatea L.) from Ceylon (=Sri Lanka presently) citing Breyne (1678)

and Plukenet (1700).

1737: Linnaeus (*) established the genus Clitoria and recognized

two species. The first species was Clitoria foliis pinnatis (=Clitoria

ternatea L.) based upon Breyne (1678), Rheede tot Draakestein (1688),

Plukenet (1700), Commelin (1701), Tournefort (1706), and Burmann (1737).

Three "varieties" were recognized using Tournefort's segregation. The

second species was Clitoria foliis ternatis (=Clitoria brasiliana L.)

based upon Breyne (1678) and Phaseolus flore vexillo amplitismo,

filiquis rectis teretibus published by Plumier in "Spec. 8."6

1739: Gronovius published Clitoria foliis ternatis, calycibus

oblongus, from Virginia (=U.S.A.), citing Petiver (1704) number 55 in

synonymy. He provided the first record of a specimen citation for

Clitoria, Clayton 108 (=Clitoria mariana L.). He also reported

Clitoria foliis ternatis, calycibus campanulatis, based upon Dillenius

(1732), and Clayton 112 (-Clitoria virginiana L.).


6. This author was unable to establish the publication title and
therefore was unable to obtain a copy of Plumier's publication.








1740: Royen reported three species from Batavia (=Java). The

first species was Clitoria foliis pinnatis of Linnaeus (1737). The

second species was Clitoria foliis pinnatis, caule decumbente (=non-

Clitoria). The third species was Clitoria foliis ternatis of Linnaeus

(1737).

1747: Rumpf (Rumphius) described "Bongu Biru" from the island of

Amboin, a Moluccan Island, as Flos coeruleus (=Clitoria ternatea L.).

Vernacular names in several languages were noted.

1748: Linnaeus (*) established binominals for his two species

published in 1737. The two species became Clitoria ternatea and

Clitoria brasiliana, respectively.

Post-Linnaeain (1753-1824: Descriptive Period

1753: Linnaeus (*) officially established the genus Clitoria,

and recognized four species, Clitoria ternatea and Clitoria brasiliana

based upon his prior works (1737; 1748), Clitoria virginiana based upon

Dillenius and Gronovius, and Clitoria mariana based upon specimen

number 55, Petiver (1704). India, Brazil, Virginia, and northern

America were given as the ranges of the four species.

1754: Linnaeus (*) published a description of the genus

Clitoria. This description is the first description published that is

officially recognized by present botanists through the rules established

in the International Code of Botanical Nomenclature (Stafleu, 1972).

1754: Miller (*) in his fourth edition of "Gardner's Dictionary"

followed Tournefort (1706) in the treatment of the genus Ternatea. Thus

the genus Ternatea Tourn. ex Miller became officially established. In

the seventh edition (1759) Miller adopted the Linnaean treatment. This








was the first encyclopedia type of reference work on the genus.

Descriptions and horticultural notes were provided.

1756: Browne reported two species from Jamaica with Linnaean

polynominals cited in synonymy. They were Clitoria major scandens,

foliis subrotundo-ovatis, floribus qeminatis (=Clitoria virginiana L.)

and Clitoria minor scandens, foliis subvillosus oblongo-ovatis,

floribus ge!inatis (=Clitoria ternatea). His Table 32 bore the title

"Clitoria galactia"; however, the plant was described under the genus

Galactia in the text, with Sloane's Phaseolus minor lactescens cited

in synonymy. From the illustration and description, this plant is not

a Clitoria.

1758: Linnaeus in his tenth edition of "Systema Naturae" cited

two additional species in addition to those established in 1753.

Neither were numbered as the other species were numbered. Clitoria

lactescens disappeared from Clitoria literature after this date, except

for Richter (1840), who cited the species in synonymy for Clitoria

galactia L. The second species, Clitoria zoophthalmum, also disappeared

from the literature until cited by Richter as synonymous with the

non-Clitorian species, Dolichos urens L.

1762: Gronovius published the same two polynominals in the second

edition of his Virginia Flora as he did in 1739. A third non-Clitorian

species published in 1739 (Clitoria foliis pinnatis, caule decumbente)

was now transferred to the genus Cracca.

1763: Linnaeus described a new species, Clitoria galactia, based

upon Sloane's specimen cited by Browne (1756). No reference was made

to Clitoria lactescens.








1766: Crantz followed Linnaeus in his treatment of C1itoria.

He listed tne five species in Linnaeus "Species Plantarum" (1763), but

altered the spelling on two species, published as C. ternatensium and

C. galactea.

1775: Aublet described a new species, Crotalaria guianensis, from

French Guiana. He was the first to use bracteole and stipule

characteristics, and the first to note the prominent costa on the fruit,

a characteristic that was important in the segregation of genera and

section's by later botanists. Aublet had cJllected several plants from

French Guiana, which included two species similar in gross appearance

that seemingly matched the illustration provided in his Plate 305.

Thus, from the misidentification, the namL Crotalaria guianensis Aubl.

would become associated with two distinct species (i.e. Clitoria

guianensis iAubl.) Benth. and Clitoria laurifolia Poir.) and provide

confusion ii later literature and specimen identification.

1775: Buchoz (*) followed the Linnaedn treatment and provided

the first detailed generic description of Clitoria, which included

perianth, stamens, pistil, legume, and seeds. Each species was

described aid notes on origin, culture, and economic properties were

included.

1775: Forskal (or Forsskal) described a new species, Lathyrus

spectabilis (=Clitoria ternatea L.) from Egypt. He was the first to

indicate the villous style typical of the genus.

1786: Lamarck (*) described two new species of Clitoria, C.

heterophylla and C. falcata, along with the basic five Linnaean species.

He also described Crotalaria longifolia citing Aublet's Crotalaria

guianensis in synonymy. Thus, Lamarck's riime is superfluous for

Aublet's name.








1786: Scopoli described Clitoria micrantha and segregated Royen's

C. brasiliana from the species and called it C!itoria quadralupensis.

Both species plus C. galactic L. were described with detailed

descriptions and illustrated.

1788: Swartz described ?Clitoria multiflora with a two-line

Latin diagnosis based upon a plant from Santo Domingo.

1788: Walter's Carolina flora recognized two species, C. mariana

L. and C. virginiana L., segregated on leaflet number and calyx type.

1789: Browne republished his treatment of Clitoria originally

published in 1756, still not adopting the Linnaean treatment. His

Table 32, Figure 2 continued to bear the totle "Clitoria galactia"

while described under the genus Galactia.

1791: Gaertner, in his publication on fruits and seeds of various

plants, described and illustrated those of C. ternatea L.

1792: Richard described Clitoria capitata with characteristics

that are too vague for placement of the species without the specimen

in hand. No type was given.

1796: Salisbury published Clitoria spectabilis which was based

upon Lathyrus spectabilis Forssk. and C. ternatea L. Salisbury's name

was superfluous for the Linnaean species.

1797: Raeuschel published Clitoria aquiliupensis, a nomen nudum,

in his list of Clitoria species.

1798: Roth described Clitoria amoena with a very detailed Latin

description (clearly marking it as a Centrosema member). It was

published in Roemer's "Archiv fur die Botanisk" and republished by

Roth in his "Catalecta Botanica" in 1800.

1802: Cavanielles described a new species, Clitoria speciosa.








1804: Desfontaine in his "Tableau dt 1 'ecole de Botanique" listed

Clitoria glbella and Clitoria galactia glbella as Clitoria glabre.

Both C. glalella Desf. and C. glabre Desf. were without any description

or discussion and thus, nomen nudum.

1806: Salisbury described and illustrated Clitoria calcarigera.

Three varieties were described based upon leaflet shape. Variety alpha

was the typical representative. Variety beta was based upon Dillenius

(1732). Variety gamma was based upon Clitoria virginiana L. Thus

Salisbury's name was superfluous for the Linnaean name, C. virginiana.

1807: Persoon included ten species of Clitoria in his treatment

of the genus. Each species had a Latin diagnosis with habitat. Two

species fro., Domingo were newly described. Clitoria plumieri Turp. was

based upon a specimen sent to him by Turpin bearing a Latin diagnosis

under the nime. The species was named for Plumier who described the

species in 1693. The second new species w,.s Clitoria rubiginosa Juss.

ex Pers. often cited as C. rubiginosa Pers.

1811: Poiret in a supplement to Lamarck's "Encyclopedie

Methodique Botanique" added seven species to those listed by Lamarck,

the last one, C. capitata Rich., lacking a number. Newly described

were Clitoria polyphylla and Clitoria laurifolia from Porto Rico and

Clitoria bracteata of unknown origin. The other three species included

the new pair described in Persoon plus Swartz' species. Clitoria

amoena Roth was synonymized with C. brasiliana. Clitoria galactia was

reported as the type for a new genus Galactia Brown which was adopted

by Michaux.

1812: Curtis illustrated C. ternatea in color in his "Botanical

Magazine" in plate 1540.








1813: Balbis listed three species ot Clitoria in his "Catalogus

Stirpium." Clitoria broussonetii, named for the French botanist who

sent the specimen to Balbis under a name loti coerulei, was newly

described in a footnote. The footnote contained a Latin diagnosis and

placement of the species near C. galactia.

1813: Desvaux (*) described the genus Neurocarpum and indicated

that it included Aublet's Crotalaria gajanensis ( published as

C. guianensis by Aublet) and a new species, N. ellipticum, which lacked

a description and is a nomen nudum.

1814: Desvaux (*) described Neurocarpum ellipticum and included

Crotalaria guyanensis Aubl. (Aublet published as: guianensis) and

Crotalaria longifolia Lam. as synonyms of Neurocarpum janensis Desv.

1814: Roxburgh listed five species of Clitoria cultivated in

Calcutta. Clitoria erecta, a new species, lacked a description and

thus was a nomen nudum.

1814: Brown described a new species, Clitoria arborescens, in

Aiton's "Hortus Kewensis," based upon a cultivar from Trinidad. This

species is usually cited as C. arborescens Ait.

1817: Eaton (*) substituted the name Vexillaria for Clitoria

because of the criticism regarding the alleged resemblance of the

Clitoria flower to the female reproductive anatomy. Eaton's

superfluous name was not adopted by botanists.

1818: Edwards illustrated and described in detail Clitoria

plumieri Turp., with notes presented on the history of the species.

1818: Rafinesque (*) established the genus Vexillaria Raf. by

his published note "Clitoria mariana must form a particular genus








Vexillaria." Neither the generic name nor Vy. mariana (L.) Raf. were

adopted by botanists.

1820: Curtis published a color plate and notes on Clitoria

heterophylla Lam. (Plate 2111).

1821: Schrader described Clitoria coccinea, and noted it near,

yet distinct from Clitoria falcata Lam.

1821: Nees included a note that Clitoria coccinea Schrad. was

near his species, but uncertain where Clitoria falcata Lam. belonged.

This was the source of Clitoria falcata Nees, a name associated with

C. coccinea Schrad. in synonymy in later literature, but the name was

not validly published. Hence, C. falcata Nees is an illegitimate name.

1821: Trdttinnick cited fourteen species in his discussion of

the genus Clitoria, and described and illustrated two cultivars,

C. ternatea and C. brasiliana.

1821: Leandro do Sacramento (*) described a new genus Martia

based upon a cleistogamous specimen from Brazil. The genus was named

in honor of the German botanist Carl Martius. One species, Martia

physodes, was placed in the genus. Martia is a homonym of Martia

Spreng. (1818).

1822: Schultes (*) published Leandro do Sacramento's species

under the name Martiusia physalodes. The change in generic name is

presumed to be due to the recognition of Martia as a homonym. The

change in the spelling of "physodes" is unclear.

1824: Kunth described in detail and illustrated two new species

of Neurocarpum, N. simplicifolium (plate LIX) and N. angustifolium

(plate LX). Both species were from Brazil.








1824: Humboldt, Bonpland and Kunth (k) described four species,

two of which were published three months errlicr by Kunth. The other

two were newly described in detail. They were N. javitense from

Javitam (=Yavita, Amazonas, Venezuela) and N. iacrophyllum from

Novo-GranatJ (=Columbia). A new genus Rhoinbolobium Rich. based upon

Richard's manuscript was cited in synonymy. The genus Ternatea Tourn.

was recognized as separate and included T. vul(aris, a new name for the

Linnaean C. ternatea. Under a third genus, Clitoria, two new species,

C. angustifolia and C. forlmosa, were described from Brazil.

1824: Nees and Martius described two new species from Brazil,

Clitoria vi:ioides and Clitoria angustifol ia. A Latin diagnosis was

given for Clitoria tomentosa which was observed as differing from

C. rubiginosa Pers.

1824: Schrank described Clitoria gladiata, a new species from

Brazil.

Post-Linnaeon (1825-1853: Revisionary Period)

1825: Hamilton miscited the genus Neurocdrpum Desv. as

"Neurocarpon." He split Desvaux's species (1814) such that N.

ellipticum was based upon a specimen in Dcsvaux Herb. and N. guianensis

included Crotalaria guianensis Aubl. and (rotalaria longifolia Lam.

He noted the relationship to Clitoria laurifolia Poir. and C.

rubiginosa Pers., both bearing the prominent lateral nerve on the

legume.

1825: Velloso described and illustrated four new species,

Clitoria flumensis (plate 123), Clitoria brasiliana (plate 129),

Cl-itoria gewna (plate 130) and Clitoria insulana (plate 131).








1825: A. P. de Candolle (*) published in mid November the first

major revision of the genus Clitoria in his "Prodromous." The genus

was divided into four sections. Section I (Ternatea Kunth) included

two species. Section II (Euclitoria) included six species, C.

glycinoides and C. poitaei newly described. Section III (Centrosema)

included three species. Section IV (Glycinopsis) included only the

newly described C. berteriana. In discussion of C. glycinoides

de Candolle referred to Poiteau's Pilanthus. This was an earlier

record of Pilanthus Poit. ex Endl. (Genera Plantarum 1289, 1841). The

genera Martiusia (one species) and Neurocarpum (seven species) were

treated as separate genera. This publication is the source for

N. falcatum DC. which was an error in citation by later botanists.

Lamarck's species was transferred by de Candolle from Clitoria and

should have been cited as N. falcatum (Lam.) DC.

1826: A. P. de Candolle (*) published his "Memoires sur la

Famille des Legumineuses" nearly simultaneously (early February) with

his "Prodromous," in which he explained his reasons for the classifi-

cation used in his "Prodromous." This was the source cited for the

origin of C. berteriana; however, the "Prodromous" predated this work

by almost three months.

1826: Desvaux (*) proposed a new group within Neurocarpum that

would bear Poiteau's name Pilanthum. This section included two

transfers cited as N. glycinoides N. (C. glycinoides DC.) and

N. rubiginosum Desv. (C. rubiginosa Pers.) plus newly described

N. villosum N. Two species were placed near Desvaux' species described

in 1813, N. laurifolium Desv. (C. laurifolia Poir.) and N. barbatum N.,

newly described. Two species of Clitoria were newly described,

C. laurifolia Nees, a homonym of Poiret (1811), and C. sinuata Nees.








1826: Descourtilz (*) substituted the superfluous name Nauchea

for Clitoria and listed thirteen species, one newly described as

N. pudica.

1827: Tussax described and illustrated Turpin's C. plumieri from

the West Indies.

1828: Wallich listed four species in his catalogue. Three were

new species which lacked descriptions, but had specimens cited for

each. They included Clitoria macrophylla Wall. (no. 5345), Clitoria

acumina-ta Grah. (no. 5346) and Clitoria pilosula Wall. (no. 5347), each

a nomen nudum.

1829: Descourtilz described the genus Nauchea, and illustrated

and described three species, N. pudica, N. virginiana and N. rubiginosa,

each with medicinal notes.

1830: Sweet listed twelve species in his catalogue with Clitoria

occidentalis listed as a substitute for the non-Linnaean virginianaa

B.R. 1047." The species is not described and the citation is unclear.

1830: Guillenim and Perrottel in their "flora Senegambiae

Tentamen" describe a new species, Clitoria micrantha, a homonym of

Scopoli (1786).

1832: Curtis illustrated C. arborescens. This illustration was

a composite of two species, C. arborescens and C. javitensis which

resulted from additions made to an unpublished drawing of 1822.

1832: Presl described and illustrated Neurocarpum cajanifolium.

1832: Geel described and illustrated the "Plumier Clitoria."

1832: Zuccarini described Martia mexicana in detail.

1832: Don (*) adopted de Candolle's trealinent of Clitoria,

Martiusia and Neurocarpum. He described Clitoiia racemosa and Clitoria








alba and placed them in Sect. Euclitoria. Four of the five dubious

species listed by de Candolle were placed. Clitoria broussonetii was

transferred to the genus Cologania. Clitoria phyrne Juss., C. mariana

ioc. and Sesse ined., C. anustifolia Nees and Mart. and C. rubiginosa

Nees and Mart. were synonymized with Galactia species.

1835: Pritzel listed Nauchea clypeata Desc. based upon "Desc.

Ant. 8,591," a publication not obtained by this author.

1837: Hooker described and illustrated Clitoria virdiflora Bouton

ex Hook'.

1837: Bojer described Clitoria lasciva and reported Clitoria

virdiflora Bouton mss. Both Hooker and Bojer cite "Bouton mss. in hb.

nostr." Stafleu's "Taxonomic Literature" (no. 537) indicated Hooker

published sometime within October 7 to November 8. The date of Bojer's

publication was unknown except for the year, although if it preceded

Hooker, then C. virdiflora Boj. was a nomen nudum. Bojer made no

reference to Hooker.

1837: Schlechtendal reported Clitoria mariana L. from Mexico.

This was the source of Clitoria mariana Schlecht. cited by other

botanists. Schlechtendal had misidentified his specimens, which are

Clitoria mexicana Link.

1837: Bentham (*) published a revision of de Candolle's treatment

of Clitoria, Neurocarpum and Martiusia. Two sections of Clitoria were

eliminated. Section Centrosema was elevated to a genus level. Section

Glycinopsis was eliminated through the transfer of its only species to

another genus. To the remaining two de Candolle sections, Bentham

added the section Bractearia. Martiusia and Martia were synonymized

with Neurocarpum. Bentham noted the close relationship of Clitoria and








Neurocarpum segregated only by the prominent costa on the legume. New

species included Clitoria amazonum Mart. ex Beiith., Clitoria acuminata

Benth., Clitoria arborea Benth., Clitoria raceinosa Benth., Clitoria

pedunculata Boj. ex Benth., Neurocarpum longifolium Mart. ex Benth.,

Neurocarpum frigidulum Mart. ex Benth., Neurocarpum rufescens Benth.,

Neurocarpum densiflorum Benth., and Neurocarpum bracteatum Mart. ex

Benth.

1838: Schlechtendal described Clitoria schiedeana.

1838: Lindley reported that C. ternatea roots were emetic.

1839: Bentham's 1837 article was republished with the title

revised.

1839: Bentham reported on Schomburgk's Biitish Guiana collections

and substituted the name Dendrocyamus for his section Bractearia. He

reasoned that the name Bractearia was preoccupied by a genus in the

Rubiaceae and in a section of Chaetogastra, and thus inappropriate for

Clitoria.

1839: Bentham described Clitoria polystachya in his enumeration

of Hartweg collections.

1840: Bentham described Neurocarpum flagellare and synonymized

N. frigidulum with N. longifolium in Hooker's "Journal of Botany."

1840: Steudel published a list of Clitoria species and species

transferred. The following binominals, each a nomen nudum, appeared for

the first time: C. brasiliana Arrab., C. !Irahdmi Steud. (C. acuminata

Grah.), C. insulana Arrab., C. micrantha Smith, C. phryne Commers. and

C. pudica Steud. (Nauchea pudica Desc.). The latter name was a

transferred name and legitimate. Clitoria grahami was an apparent








substitute name for the homonym C. acuminata Grah. which to this date

was still undescribed.

1841: Steudel's second volume published two new binominals for

Neurocarpum, each a nomen nudum. They included Martia brasiliensis

Zuccar. (N. ellipticum Desv.) and N. mexicanum Steud. (Martia mexicana

Zuccar.), both transferred to Neurocarpum, the latter described

previously, the former undescribed but placed in synonymy.

1841: Hooker and Arnott described Neurocarpum multiflorum in their

enumeration of the plants collected on the Beechey's voyage.

1841: Brunbury described Neurocarpum restipinatum from Brazil.

1842: Hasskarl described two new species from Java, Clitoria

oblonga (C. vir~jniana L. var elliptica DC.) arid Neurocarpus retusus.

1842: Bertoloni described and illustrated Clitoria alabamensis.

1843: Martius and Galeotti described Cliloria multiflora and

Clitoria grandiflora from Mexico.

1844: Bentham described Clitoria brachystegia from Ecuador in his

enumeration of plants collected on the voyage of the Sulphur.

1844: Hasskarl published Neurocarpus, an orthographic variant of

Neurocarpum, and Neurocar us retusus.

1844: Paxton illustrated in color and described Clitoria fulgens

(Centrosema coccinea).

1845: Voigt transferred Clitoria erecta Roxb. to Neurocarpum. He

listed six species of Clitoria cultivated in Calcutta.

1845: Blanco reported one species, C. ternatea, from the

Philippines.

1845: Tenore described Clitoria tristis.








1847: Burnett illustrated and described a new variety, C.

ternatea var. major, based upon a cultivar raised from seeds sent from

New South Wales. He also illustrated and discussed Paxton's species.

1848: Hasskarl provided a detailed description of his species,

Neurocarpus retusus, published in 1844.

1348: Schomburgk published three new species and a new genus

based upon a manuscript of Klotzsch. Each was a nomen nudum. The

genus was Macrotrullion K. with two species, M. spendens K1. and

M. elegans K1. The third species was Neurocarpum speciosum Kl.

1850: Dalzell described Clitoria biflora, a new species, from

India.

1851: Miquel described Clitoria amoena, a new species from the

lower Marrowyne River.

1852: Bentham described two species in a footnote that were

published previously as a nomen nudum. They were Clitoria macrophylla

Wall. (published 1828) and Clitoria grahani Steud. (published in 1852;

synonym C. acuminata Wall. published in 1828.)

1853: Duchass and Walpers describe Neurocarpum argentum, a new

species from Panama.

1854: Beurling published Clitoria portobcllensis, a new species

from Panama.

1856: Miquel described a new species front Java, Clitoria javanica.

1858: Bentham (*) published his second revision of the genus

Clitoria which included Ternatea, Neurocarpum, Martia and Martiusia in

synonymy. The major change was the synonyjiizing of the latter three

genera with Clitoria. Bentham noted that the character of the legume

costa broke down. Another change was the riame Clitoriathus used for








section three instead of Bentham's earlier names Bractearia (1837) or

Dendrocyamus (1839). New species described were C. flagellaris,

C. rufescens, C. nana, C. selloi, C. leptostachya; C. hoffmanseggii

(superfluous name for C. arborea) and C. stipularis. The species

transferred from Neurocarpum included C. simplicifolia, C. guianensis

and C. densiflora.

Post-Linnaean (1859-1932): Floristic Period)

1862: Bentham in Martius and Eichler's "Flora Brasiliensis"

listed *fifteen species with descriptions, synoinymy, and a key to the

species. This was the first record of a key to Clitoria species and

used characters of: habit; leaflet shape, pubescence and number;

legume costa; peduncle length and number of flowers borne; and

bracteole shape and length relative to calyx or pedicel. New

binominals included C. yuyanensis (orthographic variant of C. guianensis)

and C. cajanlifolia (transfer of species from N(urocarpum). A new

variety described was C. glycinoides var. juraritiaca.

1864: Griesbach reported four species of Clitoria in his flora

of the British West Indies.

1864: Bentham described a new species in his "Flora Australiensis,"

Clitoria australis.

1866: Griesbach described a new species in his "Catalogus

Plantarum Cubensium," Clitoria glomerata.

1871: Oliver reported one species from tropical West Africal in

his flora.

1875: Micheli described a new species from Brazil, Clitoria

pedunculata. This name was a homonym for l[ojet ex Benth. (1837).

Micheli placed his species in Section Clil riarthes.








1878: Lowis provided some observations on the Indian C. ternatea.

1878: Hance reported on rare Chinese plants and listed Clitoria

macrophylla Wall. for a specimen in his herbarium collected by Sampson.

This was the source for C. macrophylla Hance cited by later botanists.

1878: In a second Philippine flora, Blanco continued to report

one species of Clitoria as he had in 1845.

1878: Vatke described Clitoria zanziharensis, a new species from

East Africa.

1879: Baker (*) reported five species of Clitoria in his flora of

British India. He made a revisionary change that elevated Ternatea

and Neurocarpum to the subgenus level, the first botanist to do so.

Section Clitorianthes was not reported, presumably because the species

within this group did not occur in India. A new variety was reported,

C. ternatea var. Oilosula (C. pilosula Wall.).

1882: Sagot reported five species of Clitoria in his French

Guiana flora. A new variety, C. javitensis var. gjabra, was described.

1887: Watson described two new species from Mexico. Clitoria

triflora was placed near C. quianensis in the section Neurocarpum.

Clitoria sericea was described from a fruiting specimen. Watson did not

place this species in any section.

1887: Hemsley described Clitoria Hanceana based upon a specimen

in Hance's herbarium collected by Sampson. He noted that this was

equivalent to C. macrophylla Hance (cf 1878), but not C. macrophylla

Wall.

1887: Sesse and Mocino described two new species from Mexico.

Clitoria racemosa (homonym of Don, 1832, and Bentham, 1837) was first

reported in synonymy by Don (1832) and probably by de Candolle (1825)








as a synonym of C. plumieri based upon a iinusL.ript of Mocino and Sesse

cited as "fl. Mex. ic. ined." The second species Clitoria uirginiana,

was probably an orthographic error (cf. 1 93) because Gronovius'

"Flora Virginiana" was cited (also typographic error with "U" for "V"

and corrected by Sesse and Mocino in the second edition of their

publication.

1889: Deflers reported one species from Yemen.

1893: Sesse and Mocino published their second edition of "Plantae

Novae Hispaniae" and cited C. virginiana. This corrected the

orthographic error of their first edition (cf. 1887).

1891: Kuntze (*) transferred twenty-nine species to the genus

Ternatea because Clitoria was a homonym of Clitorius Dillenius

(1732).

1894: Taubert in Engler and Prantl's "Die Naturlichen Pflanzen

Familien" adopted Bentham's treatment. They indicated about thirty

species in the genus were found in the tropical zone, but their species

were reported by sections and totaled to only twenty-four. They placed

ca. six species in Section Ternatea (Asia and Africa), ca. fifteen

species in Section Neurocarpum (Asia, Australia, and America), and ca.

three species in Section Clitorianthes (tropical America).

1894: Nairne reported one species of Clitoria in his Western

India flora.

1894: Robinson described the flowers of C. sericea (cf. Watson,

1887) and noted that the style was similar to Centrosema, yet had other

flower characters of Clitoria.

1894: Sesse and Mocino's "Flora Mexicana' followed their 1893

treatment of Mexican plants.








1894: Siebert and Voss reported Clitoria coelestis as a horti-

cultural synonym of C. ternatea. This new binomial is a nomen nudum.

1897: Britton and Brown reported one species of Clitoria in their

flora of the northern United States.

1897: Duss reported two species from Martinique that occur

natively plus one cultivated species. A new variety, C. glycinoides

var. ecostata, was a noiien nudum.

1897: Micheli described a new species from East Africa, Clitoria

tangani-censis.

1899: Rose reported nine species of Clitoria in Mexico with six

additional species as doubtful and excluded. Two new species were

Clitoria humilis and Clitoria subsessilis. Rose indicated that Clitoria

sericea may be a Cracca and that Clitoria grandiflora Mart. & Gal.

and Clitoria schiedeana suggested Centrosema.

1900: Bailey reported two Australian species of Clitoria in his

Queensland flora.

1901: Huber described a new species from Brazil, Clitoria cearensis

and noted its affinities with C. glycinoides, C. guyanensis and

C. cajanifolia.

1903: Merill reported vernacular names of C. ternatea in his

Philippine Dictionary.

1904: Chodat and Hassler described a new variety from Brazil,

Clitoria cajanifolia var. latifolia.

1907: Harms reported the first occurrence of cleistogamy in

Clitoria in a detailed article. Cleistogaious flowers in three species

were described and illustrated. They were C. glycinoides, C.

guianensis crnd C. cajanifolia. Harms discussed the history of Martia








physodes, a species described based upon a cleistogamous flowered

specimen from Brazil (cf. 1821).

1908: Mattei described a new species from northern Africa,

Clitoria albiflora. He noted its affinities to C. ternatea and C.

tanganicensis.

1909: Bailey listed the same two species of Clitoria in his

Queensland catalogue that he reported in his flora (1900).

1909: Gibbs reported one species of Clitoria in his Fiji flora.

1909: Herzog described a new species from Bolivia, Clitoria

nervosa. He noted its affinities to C. hoffmansegii and C. racemosa

Benth.

1909: Huber described a new species from Brazil, Clitoria

obidensis.

1910: Hassler described a new variety of Clitoria from Paraguay,

C. densiflora Benth. var. mucronata.

1911: Dunn added C. mariana to the list of species found in

China.

1911: Hayata reported one species of Clitoria from Formosa.

1912: Merrill reported one species of Clitoria from Manila in

his flora.

1912: Compton (R) described the seedling morphology and vascular

structure of C. ternatea. This was the first non-floristic, non-

nomenclatural, non-horticultural research undertaken on the genus

Clitoria.

1913: Harms described a new species of Clitoria from the Congo,

Clitoria kaessneri.








1914: Baevecke reported one species of Clitoria from the Eastern

United States in his flora.

1915: Gagnepain described a new species from Laos, Clitoria

linearis, and placed it near C. hanceana.

1916: LeCompte reported six species of Clitoria from Indo-China.

A key, descriptions, distributions, and citations were included.

Clitoria linearis was illustrated. A new variety was described,

C. hanceana Hemsl. var. laureola.

1917: Ewart reported one species of Clitoria in his Australian

flora on the Northern Territory.

1918: Britton reported one species of Clitoria from Bermuda in

his flora.

1918: Kirtikar and Basu reported the Indian vernacular names and

medicinal properties of C. ternatea.

1918: Pittier described and illustrated the keel and wings of a

new Venezuelan species, Clitoria dendrina.

1919: Brandegee described a new species from Baja California,

Clitoria monticola, and segregated it from C. mariana.

1920: Britton reported one species of Clitoria from the Bahamas

in his flora.

1920: Fawcett reported two species of Clitoria from Jamaica in

his flora. They were segregated by leaflet number and flower color.

1921: Harms described a new species from Brazil, Clitoria

brachycalyx and reported its affinities with C. brachystegia.

1922: Ducke described a new species from Brazil, Clitoria

snethlageae, and noted its affinities with C. leptostachya.








1922: Ridley reported two species of Clitoria from the Malayan

Peninsula in his flora and noted that C. cajanifolia was introduced

into Indonesia on Brazilian cattle which had the viscid seeds in their

hair.

1922: Standley reported six species of Clitorian trees and shrubs

from Mexico in his flora. A key and brief diagnosis were included.

1923: Merrill reported one species of Clitoria from the

Philippines in his flora.

1924: Britton reported three species from Puerto Rico and the

Virgin Islands in the survey conducted by the New York Academy of

Science. One species was reported under the genus Clitoria. Two

species from the Neurocarpum section were reported under the genus

Martiusia. Even through the original species N. ellipticum Desv. was

listed in synonymy, Britton did not use the older generic name

Neurocarpum. The two transferred species were Martiusia rubiginosa

(Juss.) Britton and Martiusia laurifolia (Poir.) Britton. This was the

first example of a botanist diverting from Bentham's treatment of 1858.

This trend of recognizing two genera, Clitoria and Martiusia, was

followed by a few American botanists.

1925: Fragoso and Ciferri (R) described several ascomycete

species from branches and legumes of "siccis Clitoriae triandrae"

collected in the Dominican Republic. This was the source for two

names, Clitoria triandra and Clitoria riandra.

1925: De Wild described a new species from British East Africa,

Clitoria merrnsi.

1926: Small described a new species from Florida, Clitoria

fragrans andl noted its affinities with C. mIaridna.








1928: Standley reported four species of Clitoria from the Panama

Canal Zone in his flora. A key and habitats were included.

1929: Baker reported four species of Clitoria in his treatment

of Tropical Africa legumes. A key and diagnosis were included. He

noted that C. albiflora and C. mearnsi were equivalent to C. ternatea.

A new variety that had been nom. in sched. was described as Clitoria

ternatea var. angustifolia Hochst. ex Bak. Baker published Clitoria

tanganyicensis Micheli as an orthographic correction for C.

tanganicensis licheli (1897).

1929: Hutchinson & Dalziel transferred C. racemosa Don to the

genus Vigna.

1929: Standley transferred C. sericea Wats. to the genus

Tephrosia.

1930: Barker and Darbeau reported four species of Clitoria from

Haiti which included C. plumieri Tuss., a nomen nudum.

1930: Ducke described a new species from Brazil, Clitoria

grandifolia, and noted its affinities to C. arborescens.

1930: Macbride described a new species from Peru, Clitoria

pozuzoensis. He noted that some of the characters that Bentham used to

distinguish the species of Section Clitorianthes broke down since

Macbride's new species keyed to C. javitensis from which it was totally

different.

1931: Gleason reported a species of Clitoria collected on the

Tyler-Duida expedition to Colombia.

1931: Malme described a new species from Brazil, Clitoria

chapadensis and noted its affinities with C. guyanensis. He also

described a new variety, Clitoria qlycinoides var. megapotamica.








1931: Sandwith reported his observations on the type of

nomenclatural problems of C. javitensis.

1931: Standley reported two species of Clitoria from the

Lancetilla Valley of Honduras.

1931: Williams reported four species of C1itoria from Trinidad

and Tobago. A key is included.

Post-Linnaeanr (1932-present: Research Period)

1932: Tschechow and Kartaschowa (R) reported the first chromosome

count for the qenus with 2n=16 for C. terrntea. The chromosomes were

illustrated in a camera lucida drawing. Relationships of C. ternatea

with other species in tne Phdseoleae tribe were discussed.

1932: Rylberg reported one species of Clitoria in his flora of

the Central North Ameri.an plains and prairies.

1933: Small reported oie species of :litoria and two species of

Martiusia it the Southestern United States. His floristic study

adopted Britton's treatment (1924). Small's species (1926) was

transferred as Martiusii frayrans Small.

1933: Holland and Joachin (R) reported their results on soil

erosion experiments with C. (ajanifolia in Ceylon (=Sri Lanka).

1935: Burkill reported the economic uses of two species of

Clitoria of thte Malayan Peninsula.

1936: Brown reported one species of :litoria in his flora of

Southeastern Polynesia.

1936: Standley reported two species of Clitoria from British

Honduras. Either a key nor descriptions were included.

1936: Stahl included three species or Clitoria from Puerto Rico.

1937: Standley reported four species of Clitoria from Costa Rica.








1937: Stehle reported on the colonizationi of some plants in

Martinique ond included Clitoria rubiginos.i Juss. var. ecostata (Urb.)

Stehle, a numen nudum.

1939: Amshoff reported eight species of Clitoria in Pulle's

Suriname flora. A key was included. Amshoff adopted Bentham's

classification (1858).

1940: Standley described a new species from Panama, Clitoria

velutina.

1941: Burkart described and illustrated j new species from

Argentina, Clitoria cordobensis. He discussed cleistogamous flowers

in the species and its affinities with C. Iariana, C. guyanensis,

C. multiflora ind C. rubiginosa.

1943: Kol, Machado and Meirelles (R) reported their results of

utilizing C ternatea ihy for sheep digestion.

1943: Ma,.bride reported eight species of Clitoria from Peru. A

key, descriptions, and citations were included. However, this treatment

was of poor quality because of the numerous misidentified specimens that

Macbride examined and the difficulty within the group Clitorianthes.

Macbride reported variability in several species, and along with his

additional notes, indicated his awareness of the problem. Some species

reported do not occur in Peru. Other species not reported do occur in

Peru. Peru has several endemics not recogiiized by Macbride.

1943: Mocoso reported four species of Cl itoria from the Dominican

Republic in his flora. Clitoria polyphylla was included in synonymy

with a Barbieria species.

1943: Pittier described a new specie, from Venezuela, Clitoria

glaberrima. He noted its affinities with C. arborescens and C. dendrina.








The description included "Pedicelli 4-5 cm' which was a typographical

error for "mm."

1944: Pittier published a second description of C. dendrina

(cf. 1918) ind illustrated the perianth.

1945: Hassell (R) reported his results of experiments to

establish C. ternatea in pastures of Queensland, Australia.

1946: Standley reported five species of Clitoria in his flora of

Guatemala. A key was included.

1946: Strang reported ihe ornamental properties of C. racemosa

Benth. in Brazil.

1948: Robyns reported one species ot Clitoria in his flora of

Albert National Park in Africa.

1948: Stehle and Quentin reported t,,j species of Clitoria from

Guadeloupe in their flora. They included rnbiinosa var. genuina

(illegitimate name for the typical variety) and C. rubiginosa var.

ecostata (Ub.) Stehle, a noiren nudum (cf. 1937).

1949: Burkart described and illustrated a new species from

Argentina, Clitoria epetiolata. He noted its affinities and compared

the species to C. guiarrensis. The variety C. epetiolata var.

latiuscula w as described.

1949: Chopra, Nath, Badhwar, and Ghosh reported the economic uses

and problems of C. ternatea in their "Indijenous Drugs of India."

1949: Henderson reported C. laurifolia as an established Malayan

wildflower introduced from Brazil early in the nineteenth century.

1949: Johnston reported C. portobellonsis from San Jose Island,

Gulf of Panama, in his flora and discussed its confusion with Clitoria

arborescens, the name the Panama plants were given by botanists.








1949: Merrill in his "Index Rafines(eue" reported that C. parviflora

Raf. was C. ternatea.

1950: Metcalfe and Chalke (R) included the genus Clitoria in their

descriptive anatomy of the Papilionaceae. Leaf, axis and wood

characteristics were included. The species studied were not reported.

1950: Velez and Overbech included C. terratea as a desirable

cultivar in their book.

1951: Krapovickas and Frapovickas (R) reported a chromosome count

of 2n=24 for C. cordoboensis and discussed its relationship to C.

ternatea.

1951: Leon reported three species of- Clitoria from Cuba in his

flora. Clitonria glomerita was included in synonymy with a Galactia

species.

1951: Quisumbing reported the medicinal properties of C. ternatea

in the Philippines.

1951: Rau (R) described endosperm development in C. ternatea.

1952: Andrews reported one species of Clitoria from the Sudan in

his flora.

1952: Burkart reported five species of Clitoria from Argentina.

A key was included.

1952: Lemee reported six species of Clitoria in his French Guiana

flora.

1952: Schery described a new species, Clitoria coriacea, from

Brazil.

1953: Frahm-Leliveld (R) reported a chroniosome count of 2n=16 for

C. ternatea.








1954: Berhaut reported two species or Clitoria from Senegal in

his flora. Within his key he published C. ternatea var. alba.

1954: Wilcez reported three species of Clitoria in his Belgium

Congo flora. A key and descriptions were included. He recognized

Baker's variety (1929) of C. ternatea and included C. tanganicensis

and C. mearnsi as synonyms.

1955: Sirdeshmukh (R) reported a double flower occurring within

C. ternatea.

1956: Bunting and Lea (R) reported their observations on the

introduction of three strains of C. ternatea in Sudan pastures.

1956: Fidalgo (R) reported his anatomical study of the Brazilian

tree ornamental, C. rac.emosa Benth. (=C. rairchildiana). This was the

first and only anatomictAl study of a Clitotia species, except for the

survey of MNtcilfe and Chalk (1950). TwerLy plates of illustrated

anatomical structures were presented.

1956: Gardner and Bennett included C. ternatea as a poisonous

plant in their book "Toxic Plants of Western Australia" and reported

the curariform properties of the seeds of U. atborescens.

1957: Cowan described a new species from Venezuela, Clitoria

cerifera, and noted its affinities to C. a:borescens. The species was

illustrated in his Figure 65.

1958: Chopra, Chopra, Honda, and Kapur included C. ternatea in

their book "Indigenous Drugs of India" and reported the species' use and

effects as: a cathartic, ; remedy for snake bites and scorpion sting,

an anti-dysentaric drug, a diuretic drug, j laxative, a drug for gonor,

and a skin jelly for eczema, prurigo and iipetigo.








1958: Santapau reported one species of Clitoria in his flora of

Purandhar, India.

1959: Rizzini described a new speciL, from Paraguay, Clitoria

burkartii.

1959: Tiwari and Gupta (R) reported the first chemical study of a

Clitoria species and reported a new chemical "dparajitin" extracted

from the leaves of C. ternatea.

1960: Sinha (R) identified a sterol from the yellow oil of the

seeds of C. ternatea. In a second article, Sinha (R) reported a

chemical and its properties extracted froii the leaves of C. mariana.

Although unnamed by Sinha, the author of the present study found the

properties LO agree with those of "aparajitin.

1961: Hundley reported two species uf Cl itoria from Burma and

their vernacular names.

1961: Sen (R) reported and illustrated binucleate pollen mother

cells in C. ternatea.

1961: Sen and Krishnan (R) reported their genetic experimental

results with the double flowers of C. territea and concluded that the

double flowers originated through a dominant gene mutation.

1961: Saroja (R) reported a chromosome count of n=8 for the

double flowered C. ternitea. He also reported his results on the

fertility of its pollen grains with those grains of the papilionaceous

flower type. Saroja concluded that the mutant gene was pleiotropic.

1962: Lind and Tallantire reported one Clitoria species from

Uganda.

1962: White reported one species of Clitoria from Northern

Rhodesia (-'ambia).








1963: Backer and Bakhuizen von Brink reported four species of

Clitoria from Java within a detailed descriptive key. They synonymized

C. javanica with C. rubiginoa.

1963: Maheshwari reported one species of Clitoria from Delhi.

1963. Rizzini reported twenty-three species of Clitoria from

Brazil. His treatment included keys, descriptions, morphological notes,

extensive citations, and illustrations of leaflets and some germination

diagrams, lie reported that for the few species studied, seed

germination type was an excellent sectional character. He found that

Sect. Ternatea and Sect. Clitorianthes was of the epigean type and that

Sect. Neurocarpumi was of the hypogean type. Newly described were

Clitoria rubiginosa f. longifolia, C. rubiginosa f. latifolia, C.

guyanensis i. chapadensis (Milme) Rizz., L. amizonum f. rotundifolia

and C. raceimosa f. obovatifolia. Despite 1 feu; nomenclatural errors,

Rizzini's floristic treatment was the best of any seen by the author of

the present study.

1965: Angley reported three species of Clitoria from Parana,

Brazil without naming the species.

1965: Gooding, Loveless,and Proctor reported one species of

Clitoria from Barbados.

1965: Liogier transferred two species from Martiusia in Britton

and Wilson's flora (1924) to Clitoria, which corresponded to Bentham's

treatment.

1966: Chavan reported two species of Clitoria from Gujarat, India.

1966: Hutchinson and Dalziel reported two species of Clitoria

from West Tropical Africa.








1966: Lourtig discussed the specimens within the herbarium of

Johan Burman and reported thit Flos Clito) ius flore coeruleo Burm was

Clitoria ternratea (cf. 1737).

1966: Padmanabhai (R) reported the tirst tissue culture studies

on a Clitorian species. His results on the excised embryonal axes of

C. ternatea were discussed.

1967: Boulos reported one species of Clitoria in his weed flora

of Aswan, Egypt.

1967: D'Arcy reported one species ot Clitoria from the Virgin

Islands.

1967: Howard reported his observatiios on the woody Clitoria

species and substituted the iame Clitoria iairchildiana for the homonym

C. racemosa Benth.

1967: Mullick, Pra:kash and Chatterji (R) reported their results

on seed germination of C. ternatea under v.irious ecological and

physiological conditions.

1967: Mullick and Chatterji reported their results of the effects

of sodium cyanide on seed germination of C. ternatea.

1967: Mitra and Datta reported a chromosome count of n=8 for

C. ternatea.

1968: Baum discussed binominals used by Linnaeus in his "Systema

Naturae, ed. 10" and reported C. lactescens was C. galactia and that

C. zoophthalmum was Dolicho> urens.

1968: Gupta and Lil (R) reported their isolation of a hexacosanol,

a sitosterol and an anthoxanthin glucoside froii the seeds of C. ternatea.

1968: Lakshmann and Padmanabhan (R) reported their results of the

antibiotic .scuchitine on th. growth of C. ternatea in vitro.








1968: Radford reported one species of Clitoria from the Carolinas.

1968: Uphof reported the economic iI~portance of three Clitoria

species in his dictionary.

1970: Datta and Sjha (R) reported tl, floral vasculature of

C. ternatea and concluded that it represented an advance stage in the

Phaseolae.

1970: Katiyar, Ranjhan and Shukla (R) reported their experimental

results on the yield and nutritive of C. tcrnatea for sheep.

1970: Oakes (R) reported his results on C. ternatea grown for

livestock consumption.

1970: Correll and Johnston reported one native species and one

cultivated species of Clitoria in Texas.

1970: Verdcourt reported one species of Clitoria from Tropical

East Africa and described a new variety C. rubiginosa var. glabrescens.

He proposed that the following species of the "rubiginosa-mariana"

complex be reduced to subspecific levels under the two species

C. mariana and C. rubiqinosa based upon stipule shape. His complex

included C. mariana L., C. grahami Steud. ex Benth., C. javanica Miq.,

C. australis Benth., C. frag'ans Small, C. mexicana Link and

C. rubiyinol d Pers.

1971: Cameron reported C. ternatea reacted favorably to a small

extent in becoming established where trifluralin was used to control the

black pigweed Trianthemna portulascastrum.

1971: D'Orey and iberato reported one species of Clitoria from

Portugese Guinea.

1971: Gillett, Pohill and Verdcourt reported three species of

Clitoria from Tropical fast Africa. Clitoria ternatea was concluded









to be a polymorphic species that included C. zanzibarensis, C. mearnsii,

C. tanganicensis and C. ternatea var. angustifolia, and no variety was

recognized. Refer to the African members Jf C. ternatea in this study.

1972: Adams reported two species of Clitoria from Jamaica.

1972: Liberato reported one species of Clitoria from S. Tome and

Principe.

1973: Aiyar, Narayanan, Seshadri and Vydeeswaran (R) reported

three glycosides of kaempferol extracted from the leaves of C. ternatea.

1974: Croat reported six species of Clitoria in Panama, with the

new combination C. guianensis var. subsessilis. Croat correctly

provided the first explanation of the plants known by the name C.

arborescens as represented by two species, a liana C. javitensis and

the tree, C. glaberrima.

1975: Fantz sent a manuscript on Clitoria for inclusion within

the Leguminosae in the Flora of Panama series, which was accepted and

approved by the editor. Six species were reported and described. They

represented the same six species identified by Croat, but with some

changes in the species names to correct nomenciatural errors to conform

with the Code. Croat's variety was not recognized as his segregating

characters broke down. The publication will appear when the entire

family has been treated.

1976: Fantz presented a preliminary paper comparing diagnostic

characters that segregate Clitoria from Centrosema. This was in

response to some individual suggestions and herbaria proposals that the

two genera should be recombined, a proposal he does not agree with as

the genera are distinct entities.















PRIOR RESEARCH ON CLITORIA


Most of the prior research on Clitoria that is non-floristic and

non-nomenclatural occurred after 1950. The amount of research on

Clitoria has been relatively small given the size of the genus. Most

of the work has involved the species Clitoria ternatea. This species

is commonly cultivated in gardens and often escapes, becoming

naturalized, giving it the largest distribution (pantropical and

subtropical) in the genus. Because of its easy accessibility,

C. ternatea was usually selected as the representative of the genus

Clitoria when research problems included several legume genera.


Anatomical Research


The few anatomical studies on Clitoria have resulted from surveys

including a number of genera. Only Fidalgo (1956) provided a detailed

anatomical study of a species within the genus.

Metcalfe and Chalk (1950) included Clitoria as one of the genera

in their descriptive anatomy of the Papilionaceae. No reference was

made as to which Clitoria species were used in their study. They often

described an anatomical characteristic followed by a list of genera

that included species with this characteristic. The following

anatomical characteristics were noted for Clitoria: Leaf: Non-glandular,

hooked hairs, with short basal cells, and a larger bent, terminal cell

(p. 505); lower surface papillose or subpappilose (p. 506); epidermis









often including a proportion of mucilaginous cells (p. 506); rod-shaped

crystals (styloids) in the palisade tissue (p. 511). Axis: Young stem

with cork arising in the middle of the cortex (p. 513); tanniniferous

cells, colored brown in dried material, and sometimes containing

proteins, mucilage, and other materials as well, often occurring in

groups situated in the pith and phloem or in the primary cortex

(p. 515). Wood: Parenchyma very abundant, in broad, moderately

regular bands 4 cells or more wide in some species (p. 521); rays

exclusively uniseriate or with only occasional biseriate rays in some

species (p. ;24); rays up to 1 mm in height in some species (p. 524);

sheath cells present in some species (p. 525); fibres forming small

islands on the cross-section owing to the abundance of wood and ray

parenchyma (p. 525).

Datta and Saha (1971) included Clitori a ternatea as one

representative member in their study of the floral vasculature of the

tribe Phaseoieae. Their findings include tiat the common bundles for

perianth, stamen and disc remained undividd.l for a short distance; that

the disc traces are feebler, branching immediately after origin in the

swellings at the bases of the stamens for a short distance; that the

common bundles for the perianth and stamens are elongated; that the

style is hollow having one dorsal and two marginal traces, with the

upper part solid having a single trace; that the vascular trace

continues almost up to the stigma tip. A camera lucida drawing of a

transaction of the floral axis is illustrated in Figure 1 and a graphic

diagram of the floral vasculature is illustrated in Figure 17.

Fidalgo (1956) did ,in anatomical study of Clitoria fairchildiana

(published under the synonym (litoria racemiisa), a tree species









cultivated in parts of Brazil as an ornamental, and infrequently

cultivated elsewhere. He provided 43 page, of his techniques and

descriptive anatomical results on the roots, steis, leaves, flowers,

fruits, and seeds. He included 20 plates (52 figures) of drawings that

illustrate the anatomical structures. He includedd that there is a

great occurrence of calcium oxalate, monoclinal, principally in the

cortical parenchyma; small particles of starch were found in woody

parenchymous cells, in great abundance in the pith; tannins occur

frequently, not from internal differentiation of cell, but from

secretary pockets or canals; mucilaginous canals commonly not in

parenchymatous pith rays; external epidermis provided with uniseriate

protective hairs (not peduncle) with 1 or 2 basal cells, partially or

totally caducous with developing organ; secretary hairs occur in hidden

parts of peduncle, with a single basal cell and a terminal swelling;

development of 20 structures of stem emerges between layers of apical

collenchyma, strata of suberized cells occupy various positions;

petioles invaginated from bark by activity of cambium and isolated

woody elements forming a secondary internal woody ring of bark;

pulvinus and rachis express an interesting evolution of cortical weave

which permits movement of the leaves; leaves present numerous

adaptations to xerophytic conditions, with great cavities in the lower

epidermis, obliterated by long protective hairs over large, scanty

stomata; fruit with two patterns of sclerenchymous layers orientated

in different directions; seed coat formed of numerous weave patterns

in the external epidermis, not containing toxic proteins, presenting

great advantages for their use through cattle alimentary tracts.









Chemical Research


Chemical research has been conducted by Indian investigators on

the seeds and leaves of Clitoria species that occur natively in India

and are reported by the Indians to be of medicinal value.

Tiwari and Gupta (1959) reported a new chemical they named

"aparajitin" that was obtained from an alcoholic extract of the dried

leaves of C. ternatea. The chemical name was derived from the plant's

Indian vernacular name of "aparajita." Based on experimental results,

aparajitin (C26H5002) was determined to be a o-lactone of 2-methyl-4-

hydroxy-n-pentacosanoic acid with the chemical structure:

CH

CH3(CH2) 18-CH2CCHCH2 CH2HCO

0----

The properties of aparajitin were described as insoluble in water,

cold alcohol, chloroform, ether, and petroleum ether; sparingly soluble

in acetone; soluble in hot alcohol, benzene, ethyl acetate and

pyridine; did not decolorize bromide water or alkaline potassium

permangate solution and gave no acetyl derivative; did not reduce

Fehling's solution nor respond to any keto group test; laevo-rotatory

with M.W. of 391 and m.p. of 92-93'; treatment with hydroiodic acid

and hydrobromic acid produced 2-methyl-4-iodo-n-pentacosanoic acid

(C26H51021) and methyl-4-bromo-pentacosanoic acid (C26H5102Br)

respectively; esterification produced methyl acid, 2-methyl-4-hydroxy-

n-pentacosanoic acid, which formed the acetyl derivative of 2-methyl-

4-hydroxymethyl-n-pentacosanoate; oxidation produced arachidic acid

and B-methylglutanic acid.









Sinha (1960) identified the yellow fixed oil (yield 18.78%)

obtained from the seeds of C. ternatea as a y-sitosterol based upon

the preparation of acetate and benzoic derivatives and comparing their

comparative compositions, melting points and rotations with those of

known sitosterols. The sterol's properties were described as:

obtained as colorless shiny plates; molecular formula of C20H500;

negative test for N,P,S, and halogens; positive Salkowski reaction

and digitonin test; coloration assumed purple to blue, then to green

in Liebermann Burchard reaction; purple coloration in Steinkle

Kehlenberg reaction, which turned cobalt blue on exposure to light.

Sinha (1960) described a o-lactone compound from alcoholic extract

of the dried leaves of Clitoria mariana with a molecular formula of

C26H5002. Front the nearly identical properties described and

treatments of the new extract, the chemical is "aparajitin," although

Sinha has no reference to the work of Tiwari and Gupta.

Gupta and Lal (1968) reported the isolation of hexacosanol,

B-sitosterol and an anthoxanthin glucoside from the seeds of

C. ternatea. On acid hydrolysis, the anthoxanthin glucoside yielded

quercetin and glucose. Chromatography of the amino acid composition

enabled the identification of the essential amino acids lysine, valine,

methionine, phenylalanine and isoleucine in the protein hydrolysate,

along with aspartic acid, serine, glycine, alanine, glutamic acid,

tryosine, proline, argisine and histidine.

Kulshrestha, Kumar and Khare (1968) detected cersulfate positive

compounds by thin layer chromatography in the butanol-soluble fraction

of the seeds of C. ternatea. Four of the six crystalline compounds

were identified as adenosine, kaempferol-3-rhamnoglucoside,









p-hydroxy-cinnamic acid and ethyl-alpha-D-galactopyranoside. The

remaining two compounds were a polypeptide and a phenolglycoside.

Aiyar, Narayanan, Seshadri, and Vydeeswaran (1973) reported three

glycosides of kaempferol from the leaves of C. ternatea. The ethyl

acetate soluble part of the ethanol extract gave three compounds on

chromatography identified as follows: Compound A (m.p. 198-200) as

kaempferol-3-monoglucoside; Compound B (m.p. 215-70) as kaempferol-3-

0-rhamnosyl (1-6) glucoside; Compound C (m.p. 188-90) as kaempferol-

3-rhamnosyl (1-6) galactoside. Light petroleum ether and ether

extracts on chromatography yielded waxy matter, chlorophyll and

B-sitosterol.


Cytological Research


Very little research has been done on the chromosomes of Clitoria.

From the few counts reported, the species appear to fall into two

distinct groups which correlate with the morphological characters which

indicate two subgeneric lines for the species involved. The subgenus

Clitoria has x=8, and is based on one species. The subgenus

Neurocarpum has x=12, and is based on four species. The chromosome

counts reported for Clitoria species are tabulated in Table 2.

Tschechow and Kartaschowa (1932) included C. ternatea as one of

the species in their cytological survey of the tribe Phaseoleae. They

provided the first chromosome count for the genus (2n=16). A camera

lucida drawing of the chromosomes was provided. They remarked that

C. ternatea differs uniquely from the other species in the tribe both

by the number of x=8 (versus x=ll, 12, 19, 20 in other species, the

latter pair of counts pertaining to other species in the same subtribe)














Table 2. Reported chromosome numbers in Clitoria species.



Subgenus CLITORIA


Species

ternatea

ternatea

ternatea

ternatea

ternatea

ternatea

ternatea (double)


n 2n date

16 1932

8 1938

16 1953

16 1957

16 1962

8 1967

8 1961


reporter

Tschechow & Kartaschowa

Senn

Frahm-Leliveld

Frahm-Leliveld

Shibata

Mitra & Datta

Saroja


Subgenus NEUROCARPUM


Species n ?n date

cordobensis 24 1951

falcata* 24 1960

laurifolia 24 1957

macrophylla 12 1971


*Chromosome number reported under the


reporter

Krapovickas & Krapovickas

Frahm-Leliveld

Frahm-Leliveld

Larsen


ynonym C. rubiginosa








and by the two distinct chromosome sizes, four long chromosomes and

four short ones. Even among the two groups, these eight chromosomes

are not equal in size, although there appears to be a 3:1 size

relationship between the long and short chromosomes.

Krapovickas and Krapovickas (1951) provided a count (2n=24) and

an illustration of the chromosomes of Clitoria cordobensis. They

concluded, after comparison to C. ternatea, that there are either two

distinct groups of species with x=8 and x=12 respectively, or that

C. cordobensis was a triploid.

Sen and Krishnan (1961b) reported that binucleate pollen mother

cells occur at a very low frequency in C. ternatea. One of the two

nuclei was always smaller. Meiosis was not observed in these cells

beyond early diakinesis. An illustration of a binucleated cell with

both nuclei at pachytene was presented.


Developmental Research


Developmental research has been undertaken almost entirely upon

C. ternatea, primarily by Indian researchers. Studies have been

conducted on endosperm, embryo growth, seed germination, and seedling

growth.

Rau (1951) described the endosperm development in C. ternatea

along with four other selected papilionaceous species. The endosperm

is of the nuclear type, with the free endosperm nuclei arranged along

the sides of the inner layer of the inner integument. As the seed

cavity at the chalazal end widens in later stages, the massive nucellus

projects into it. The antipodal end of the embryo sac is wedged into

this nucellar mass. Free endosperm nuclei lying in dense cytoplasm








gather here. Cytokinesis of endosperm nuclei occurs when the embryo is

in an advanced stage of development. The cell formation is restricted

to just one or two layers around the embryo; the remaining part retains

free nuclei.

Padmanabhan (1966) reported growing excised embryonal axes in

vitro on Nitsch's basal medium supplemented with vitamins. He

concluded that the course of axis growth was remarkably similar to

natural germination events, namely; (1) growth of root (day 2-3),

(2) elongation of hypocotyl (day 4), and (3) activation of shoot bud

(after day 20). Elongation was not affected by removal of the apical

bud. From these results the observation that upper hypocotyl curves,

he concluded that the embryonal organs are capable of independent

development when isolated, but follow the same developmental pattern

as in the entire embryo.

Lakshmann and Padmanabhan (1968) reported the effect of the

antibiotic ascochitine on in vitro growth of Clitoria ternatea. They

concluded that, in general, the toxin exerted its effects on the

embryos grown in vitro at much lower concentrations than is required

to cause perceptible effects on the germinating seeds. Ascochitine

produced total inhibition of root growth, presumably because root

proteins were easily denatured. A low percentage of embryos did

produce callusing of the root apex and the hypocotyl. The shoot

meristem was not affected by the toxin, presumably because the shoot

meristem converts a large quantity of the toxin to dihydroascochitine,

rendering this toxin benign.

Mullick and Chatterji (1967a) reported on seed germination

of C. ternatea. Imbibition and germination behavior was








studied under various eco-physiological conditions in the laboratory,

conditions similar to those found in the wild. They concluded: light

had no effect on either behavior; seeds germinated mostly at a

temperature of 35' 20C; better imbibition and germination behavior

were obtained when seeds were subjected to physical and chemical

treatments of impaction with sand, impaction without sand, hot water,

concentrated sulphuric acid, absolute alcohol and normal potassium

hydroxide. Seed imbibition and germination behavior were about fifteen

per cent under laboratory conditions of light and temperature.

Mullick and Chatterji (1967b) reported the effects of sodium

cyanide on seed germination of C. ternatea. They concluded that seeds

soaked for 24 hours in 100 ppm of aqueous sodium cyanide solution

accelerated imbibition, germination, and promoted early growth of the

seedlings. Increased concentrations produced inhibitory effects

leading to the suppression of normal seed germination and growth.

Light and temperature did not influence the chemical effect on the

seeds.

A. P. de Candolle (1825) described seedling development based upon

three species of Clitoria and illustrated various stages in Plate IX

(Figures 33-35). Two of these species are now placed in the genus

Centrosema. The illustrations show plants with 3-foliate leaves, but

the leaflet shape tends to indicate de Candolle's Figure 33 as

C. ternatea by elimination. Clitoria ternatea typically is 5- and

7-foliate; however, Rizzini's study (1963) clarifies the apparent

discrepancy in leaflet number. With the data representing a mixture

of two genera, de Candolle's treatment becomes a non-reliable source

of seedling development information in Clitoria.




Full Text
GíiOGFAPH (jF THE GENUS L1 i noria (LEGUMINOSA.
r'i YP1 'M
v^í_ ! U ilf
By
PAUL R.
A DISSERTATION PRESENTED TO THE ORA '.'AT
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULE
DEGREE CF DOC
REOUIREÍ1EN
COUNCIL OF
FOR THE
'OR CT PHILOSOPHY

COPYRIGHT
By
PAUL R. FANTZ
1977

ACKNOWLEDGEMENTS
I would like to express my appreciation for the interest, patience,
guidance, and helpful criticisms of Dr. Daniel B. Ward, under whose
supervision this monograph was accomplished. I also wish to thank the
other members of my committee, Dr. Dana G. Griffin III, Dr. Willard
Payne, Dr. Terry Lucansky, Dr. Jack Ewel, and Dr. John Kaufmann, for
their assistance during this study and for reading and providing
helpful editorial evaluations of the manuscript.
I am very grateful to the curators and staff of the herbaria
listed in the chapter on "Systematic Criteria and Taxonomic Notes" for
lending their material, including type-specimens, for use in this study.
Special thanks are extended to Gus Kovalick and the Staff of Hume
Library and Library West for their diligence and extensive aid in
searching for titles and institution of deposit, and obtaining xerox
copies of many older botanical publications; to Dr. Armstrong for her
assistance in providing appropriate geographical indexes, maps, and
atlases during my search for the obscure localities noted by the plant
collectors; and to Dr. Armstrong, Dr. Zimmermann, and Dr. Jones for
providing me with special library privileges to borrow needed references
for extensive periods during this study.
Thanks are extended to Dr. Daniel B. Ward and his staff at the
Herbarium of the Agricultural Experiment Station, for their assistance

in the necessary correspondence and the securing and returning of
borrowed herbaria material used in this study, and for providing the
needed facilities and space for the storage and examination of this
material.
I am indebted to Dr. Robert H. Mohlenbrock of Southern Illinois
University, who made this monograph possible by introducing me to the
world of plants, and whose enthusiasm stimulated me to change careers
and pursue a lifetime study in Vascular Plant Systematics.
I am grateful to Barbara Smerage and her assistant, Linda
Nordstedt, for their time, patience, and cooperation in the typing of
this manuscript.
A special thanks is extended to my wife, Janet, for her patience,
understanding, and the sacrifices she has made during the five years
it took to complete this study and for her assistance in portions of
this study.
i v

TABLE OF CONTENTS
Page
CHAPTER
ACKNOWLEDGEMENTS iii
ABSTRACT xi
INTRODUCTION 1
ECONOMIC IMPORTANCE 8
Medicinal Properties 9
Leaves 12
Seeds 13
Flowers 14
Additional Notes 14
Aphrodisiacs 14
Ornamental Properties 15
Poisons and Pesticides 18
Dyes and Fibers 19
Food for Man 19
Fodder and Green Manure 19
HISTORY OF CL I TOR IA 21
Origin of the Generic Name 21
General History 23
Historical Chronology 41
Pre-Linnaean (1 650-1753: Polynominal Period) 44
Post-Linnaean (1753-1824: Descriptive Period) 47
Post-Linnaean (1825-1858: Revisionary Period) 54
Post-Linnaean (1859-1932: Floristic Period) 61
Post-Linnaean (1 932- present: Research Period) 69
PRIOR RESEARCH ON CLITORIA 79
Anatomical Research 79
Chemical Research 82
Cytological Research 84
Developmental Research 86
Morphological Research 90
Agronomic Research 91
Palynological Research 94
v

Page
MORPHOLOGY 95
Habit 95
Stems and Branches 97
Pubescence 98
Leaves 100
Leaflets 101
Stipules and Stipels 101
Petiole and Rachis 102
Petiolule 102
Observations 102
Inflorescence 103
Chasmogamous Inflorescences 105
Cleistogamous Inflorescences 109
Chasmocleistogamous Inflorescences 109
Bracts 110
Pedicels 113
Bracteoles 113
Observations 114
Chasmogamous Flowers 114
Calyx 118
Vexillum 121
Alae 123
Carina 124
Androecium 124
Observations 127
Cleistogamous Flowers 133
Calyx 133
Corolla. . 138
Androecium 138
Gynoecium 139
Observations 139
Legume 140
Stipe 141
Valves 145
Observations 148
Seeds 148
Seed Germination 151
Observations 151
Summary 152
Species Descriptions 152
Characters Occurring in Relatively Few Species 153
DISTRIBUTION OF CLIT0RIA 159
SYSTEMATIC CRITERIA AND TAXONOMIC NOTES 173
SYSTEMATIC TREATMENT 180
VI

Page
Subgenus Bractearia 193
Section Bractearia 198
1. Cl i toria arbórea 201
a. var. arbórea 211
b. var. longiramosa 215
c. var. pseudoamazonica 216
2. Cl i tori a fa irchil diana 218
3. Cl i tori a andrei 230
4. Cl itoria juinensis 234
5. Cl i toria moyo bam ben si s 242
6. Clitoria amazonum 247
a. f. amazonum 255
b. f. vulgaris 256
7. Cl itoria nervosa 258
Section FIexuosa 265
8. Cl itoria woytkowski i 267
9. Clitoria flexuosa 275
a. var. fl exuosa 282
b. var. brevibracteola 283
10. Cl itoria pozuzoensis 284
a. var. pozuzoensis 292
aa. f. pozuzoensis 292
ab. f. subpalmata 292
b. var. schunkei 293
Section Brachycalyx 295
11. Clitoria brachystegia 299
12. Cl itoria hermannii 307
13. Clitoria glaberrima 311
14. Cl itoria canescens 321
15. Cl itoria brachycalyx 326
16. Clitoria dendrina 332
17. Cl itoria f roe si i 344
Section Caul iflorae 349
18. Cl itoria arborescens 357
19. Clitoria javitensis 372
a. var. javitensis 398
aa. f. javi tensi s 399
ab. f. bracteosubtenda 401
b. var. portobel 1 ensis 403
ba. f. portobel 1 ensis 404
bb. f. pilosa 407
be. f. truncata 409
c. var. 1 ongi loba 409
d. var. grandifol ia 410
e. var. klugi i 412
20. Cl itoria cavalcantei 413
21. Clitoria coriácea 418
22. Cl itoria tunuhiensis 425
23. Cl itoria sagoti i 429
a. var. sagoti i 439
b. var. caniculata 442
c. var. sprucei 444
vi i

Page
24. Clitoria kaieteurensis 446
25. Cl i tori a pendens 451
26. Clitori a 1eptostachya 457
a. var. leptostachya 465
b. var. fruticosa 466
27. Clitoria sel 1 oi 466
28. Cl i tori a obidensi s 474
29. Cl i toria pi umosa 480
Subgenus Clitoria 486
30. Cl itoria lasciva 494
31. Clitoria ternatea 504
a. var. ternatea 536
aa. f. ternatea 536
ab. f. pauciflora 571
ac. f. fasciculata 571
ad. f. a 1 biflora 574
b. var. angustifol ia 584
c. var. pleniflora 591
ca. f. pleniflora 596
cb. f. 1 eucopetal a 600
cc. f. subpolyadel pha 601
32. Clitoria heterophylla 602
a. var. heterophyl 1 a 609
b. var. pedunculata 610
33. Clitoria biflora 612
34. Clitoria kaessneri 619
Subgenus Neurocarpum 628
Section Mexicana 640
35. Cl i toria polystachya 645
a. var. polystachya 654
b. var. PXJH9^£Í 659
c. var. congesta 659
36. Cl itoria monticola 660
37. Clitoria triflora 665
38. Clitoria mexicana 671
39. Cl itoria humi 1 us 682
40. Cl itoria cordobensis 686
41. Clitoria fragrans 696
42. Cl itoria mariana 705
a. var. mariana 718
aa. f. mariana 722
ab. f. pedunculata 743
ac. f. pubescentia 746
b. var. oriental is 749
Section Tanystyloba 758
43. Clitoria macrophylla 762
a. var. macrophyl la 774
b. var. sericea 777
44. Clitoria javanica 779
45. Cl itoria cordiformis 784
v i i i

Page
46. Clitoria 1inearis 791
47. CÍItoria hanceana 795
a. var. hanceana 802
b. var. thailanensis 805
c. var. laureola 805
d. var. latifol ia 809
e. var. petiolata 810
48. Clitoria australis 811
Section Neurocarpum 820
49. Clitoria stipularis 826
a. var. stipularis 833
b. var. latifol ia 836
50. Cl itoria densi flora 837
51 • Cl itoria irwini i 846
52. Clitoria laurifolia 852
a. f. laurifol ia 863
b. f. glabrior 873
c. f. fasciculata 879
d. f. parvifol ia 879
e. f. petiolata 880
53. Clitoria guianensis 881
a. var. guianensis 891
aa. f. guianensis 892
ab. f. macrofructa 903
ac. f. imperfecta 903
ad. f. unifol iata 905
b. var. macrocleistogama 905
c. var. chapadensis 906
54. Clitoria epetiolata 910
a. var. epetiolata 916
b. var. angustissima 919
c. var. latiuscula 920
55. Cl itoria simp! icifol ia 921
56. Clitoria flagellaris 927
57. Cl itoria falcata 932
a. var. falcata 943
aa. f. falcata 943
ab. f. heteromorpha 960
ac. f. stipulacea 964
ad. f. longirachis 965
b. var. 1 atifol ia 966
c. var. aurantiaca 966
d. var. glabrescens 971
58. Cl itoria nana 972
a. var. nana 978
b. var. caaguazuensis 979
NUMERICAL LIST OF TAXA 980
NOMINA ET SYN0NYMA CLITORIA C0MPLECTENS 986
ix

Page
PHYLOGENETIC CONSIDERATIONS 1007
Evolutionary Trends in Clitoria 1008
The General Plant Structure 1009
Vegetative Structures 1013
The Inflorescence Structures 1014
The Flowers 1016
The Fruits and Gynoecium 1018
Pubescence 1020
Nonmorphological Data 1021
Phylogenetic Pathways in Clitoria 1022
BIBLIOGRAPHY 1026
BIOGRAPHICAL SKETCH 1051
x

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
A MONOGRAPH OF THE GENUS Clitori a (LEGUMINOSAE: GLYCINEAE)
by
Paul R. Fantz
August 1977
Chairman: Daniel B. Ward
Major Department: Botany
This monographic treatment of the genus Clitori a L. (Leguminosae:
Glycineae) is the first since that of George Bentham in 1858. The
genus is treated on a world-wide basis. The study is based upon
examination of approximately 7000 specimens from thirty-three lending
institutions, and from field studies of United States species. Treated
in synonymy are Clitori us Petiv. ex Dill., Clytoria Presl, Macrotrul1 ion
Klotsch ex Schomb., Martia Leandr. Scar., Martiusia Schult., Nauchea
Desc., Neurocarpon Hamilton, Neurocarpum Desv., Neurocarpus Hassk.,
Rhombifolium Rich, ex DC., Rhombolobi urn Rich, ex H.B.K., Ternatea Tourn.,
Ternatea Tourn. ex Mill., Vexillaria Eaton, and Vexillaria Raf.
Clitoria includes fifty-eight species and fifty-one nontypical
subspecific taxa. Forty-six species are native to the neotropics and
subtropics. One paleotropical species has been introduced and
naturalized in many regions of the neotropics. One species is native
to temperate North America and to tropical Southeast Asia. Eleven
species are native to the paleotropics in Africa and Madagascar, the
Indian subcontinent, Indonesia, and Arnhem Territory of northern
Australia. Two neotropical species have been introduced and naturalized
xi

in portions of Africa and Indonesia. One species is introduced and
naturalized in many Pacific islands.
A history of Clitori a is provided in which the origin of the
generic name is traced to Breyne in 1678. A detailed morphology of the
genus is presented. Cleistogamous flowers, as well as chasmogamous
flowers, are noted to occur in thirteen species. Species of Clitori a
are recorded to have economic uses as medicinal plants, ornamentals,
aphrodisiacs, poisons, pesticides, dyes, fibers, food, fodder, and
green manure. Prior research on species of Clitoria is summarized in
the areas of agronomy, anatomy, chemistry, cytology, developmental
botany, morphology, and palynology.
A systematic treatment is given which recognizes three subgenera
and eight sections. These include sections Bractearia (7 spp.),
FIexuosa (3 spp.), Brachycalyx (7 spp.), and Caul if1orae (12 spp.) of
subgenus Bractearia; subgenus Cl i tor ia (5 spp.); sections Mexicana
(8 spp.), Tanystyloba (6 spp.), and Neurocarpum (10 spp.) of subgenus
Neurocarpum. Each species is described and illustrated. A complete
synonymy is given. A nomenclatura! discussion of type collections,
phenology, distribution, subspecific taxa, and citation of examined
specimens is provided under each species. Economic importance and
vernacular names are included for a species when known. Each taxon has
a mapped distribution. Keys are provided to each taxonomic level.
Phylogenetic considerations are given based upon the interpretation
of morphological data presented in this study. Subgenus Bractearia is
regarded as a primitive subgenus with subgenus Neurocarpum as advanced.
XT 1

INTRODUCTION
The genus Clitoria contains fifty-eight species located mostly
within the tropical belt with a few species distributed within the
temperate zone. The bulk of the genus (ca. eighty percent of the
genera).is neotropical. The genus is absent from Europe, most of
Asia (except the Indian subcontinent), and most of Australia.
Although there is no univeral name by which the genus is commonly
known, the most often used vernacular name for the genus is "butterfly
pea." Centrosema, a close morphologically and Historically related
genus, bears the same vernacular name or is sometimes distinguished
as the "spurred-butterf1y pea." When horticulturalists attempted to
standardize plant names (Kelsey and Dayton, 1942), they adopted
"butterfly pea" for the genus Centrosema and adopted "pigeon wings"
for the genus Clitoria, the latter name rarely used by botanists.
The Papilionaceous genus Clitoria is Included by most botanists
within the tribe Phaseoleae, subtribe Glyc ineae. In the few
traditional tribes, a number of genera are usually included in a
list of exceptions to the typical tribal descriptions. In order to
reduce the number of exceptions, Hutchinson (1964) elevated a number
of subtribes to the tribal level. Thus, he placed Clitoria within the
tribe Glycineae.
Flowering plants of Clitoria are easily determined in the field
by a few unusual characteristics. Papilionaceous, resupinate flowers
1

2
which have a bearded, geniculate style identify the genus. Specimens
in fruit become more difficult to recognize due to the variation of
the legume. However, odd-pinnate compound leases (usually 3-foliate),
persistent stipules, stipels, bracts, and bracteoles, plus a
persistent, five-toothed, infundibular calyx are reliable characters
that, taken collectively indicate the genus Clitori a from other genera.
The genus Centro sema (Glycineae) is often confused with Clitoria.
Linnaeus established Clitoria in his "Species Plantarum" in 1753, and
included four species. Two of those species, C. brasiliana and
C. virginiana, were later placed in Clitoria section Centrosema
by de Candolle in 1825. Bentham (1858) segregated Centrosema from
Clitori a and elevated the taxon to the level of genus. In the keys of
most floristic and revisionary treatments, Clitoria is distinguished
from Centrosema by its "bearded style" or "pub-ascent style." This
character is often misinterpreted. Clitoria can be distinguished
morphologically from Centrosema by several characters. These
differences are summarized in Table 1.
The last revisionary work on Clitoria was a synopsis published by
Bentham in 1858. This treatment included a total of three sections
with twenty-six species, more than half of which were originally
described by Bentham. Bentham's treatment has been the major
reference relied upon by nearly all authors of floristic works that
included the genus. Bentham's treatment contains a number of
deficiencies, however, and now includes some conflicts with the
International Code of Botanical Nomenclature (Stafleu, 1972).
A number of problems presently exist within the genus that need
resolution, and these can be categorized into six areas. One problem

3
Table 1. A morphological comparison of the genera Clitoria^and
Centrosema
CHARACTER
CL ITOR IA
CENTROSE M_A
Calyx
Infundibular, lobes
broad, shorter than
to nearly equal tube
length (rarely longer)
Short-campanulate,
lobes narrow, equal to
or longer than tube
1ength
Vexi11 urn
Spurless
Spurred or gibbous
Alae vs Carina
Alae longer than carina
Alae subequal carina
Styl e
Geniculate, bearded
lengthwise to base
Broadly incurved
(U-shaped), pubescent
only beneath stigma
Ovary
Stipitate (or
subsessile in
Subgenus Clitoria)
Subsessile
Legume
Stipitate, ecostate or
with 1 lateral nerve
near midiine of valve
Subsessile, 2 prominent
nerves laterally, one
near each suture
Leaves
Pinnate - typically
3 (1) 1 eaflets
(Subgenus Clitoria
commonly 5, 7, 9, or
11 leaflets
Pinnate - typically
3 leaflets, rarely
5, 7, or 11 1eaflets

4
is the proliferation of names and their delineation into specific and
subspecific taxa. Eleven genera must be included in synonymy under
Clitoria. Benthain cited only two, although by tracing back through
his cited references, the genera included in synonymy increased to
six. To date, nearly two-hundred and fifty published binominals (plus
ca. twenty noni. in sched.) are included within the total concept of
the genus Clitoria, senso lato. Of these, one hundred and sixty of
the binominals were published under the name Clitoria. Bentham
included only sixty of these binominals in his treatment, and since
Bentham's revision, nearly fifty new species have been described,
almost double the total number of species that he recognized. Bentham
did not include subspecific taxa in his synopsis. Twenty-one of the
twenty-three subspecific taxa published were dt scribed after Bentham's
treatment.
A second problem is the lack of complete end detailed descriptions
of species, one probable cause of the large abundance of names. Few
of the published specific names were accompanied by detailed
descriptions. Many were published with short diagnostic descriptions
that included a number of characters that are typical of the genus or
subgeneric groups. Many of the descriptions contain relative terms
(i.e. short vs. long, large vs. small) and lacl. precise measurement
ranges. Often a character included in one description is lacking in
the description of another species.
Some floristic treatments have included descriptions of their
species, occasionally offering more detail than the original
description. These descriptions can be helpful, but a few have
compounded the problem. An example would be Mocbride's treatment

5
(1 943 ) of ttie Peruvian species of Cl i tori a . Many of the specimens
examined by Macbride had been misidentified and did not match the
type collections. This has resulted in descriptions placed under
inaccurate names (e.g., C. arnazonum and C. nervosa, neither species
occurring in Peru).
A third problem that occurs in conjunction with the first two is
the lack of keys to the species of Clitoria. Bentham did not include
a key in his treatment, although he used some diagnostic phrases that
separated the species into groups. The only keys that exist are those
found within certain floras. Keys which included five or more species
of Clitoria are found in floristic treatments for only a few countries
in South America, a few in Central America, plus one treatment for
Southeast Asia. They often include characters which work well
locally, but not when more species are added from neighboring areas.
A fourth problem is incomplete information on types. A number of
species had incomplete citations, making type interpretation difficult.
Bentham was partially responsible in that he often paraphrased data
from herbarium labels. Specimens labeled as "type" in the personal
herbarium of Bentham arid in other herbaria which tie examined, often
bear a collection number or date which Bentham did not cite.
Occasionally type information has been distorted. For example,
Howard (196/) proposed the name C^ fa irehi 1 diana as a substitute for
the homonym C. racemosa Benth., correctly citing the Brazilian type
collection of Pohl. Then he distributed specimens (Howard 17052)
from a cultivated tree at the Jennings Estate in Miami, Florida,
United States, as the type tree.

6
A fifth problem is the number of species that bear illegitimate
names as defined by the International Code of Botanical Nomenclature.
Although Bentham was not bound by a code developed nearly a century
after his revision, he did select names by a process that conflicted
with the Code standards in practice today. Later botanists who
assumed that those names were acceptable under the Code have often
been unaware of these differences and have used names that cannot
stand today.
A sixth problem is lack of information as to the geographic
distribution of the species. Distribution statements have often been
based upon one or a few collections and appear to sometimes have
been wholly hypothetical. The problem has been compounded in some
species which were described originally from cultivars outside the
normal distribution range (e.g., C. arborescens). Distributions in
some species have been distorted when floristic treatments have
mistakenly included a species within their geographical areas or
inaccurately reported the species from other areas.
This author has examined nearly eighty percent of the type
collections for the two hundred and fifty binominals associated with
Clitoria. Many of the types not examined belonged to species described
and placed within the genus Clitoria, but later transferred to other
genera, often Centrosema. In most cases, descriptions of these
species contained diagnostic phrases (i.e. campanulate calyx, spurred
vexillum) sufficient to indicate that the species does not belong to
the genus Clitoria.
Bentham's section Cl itorianthes was most poorly understood.
Nearly fifteen percent of the specimens that were examined in this

7
section were unidentified to species. Foi those specimens with names,
slightly more than fifty percent were mis identified. A large portion
of the species in Bentham's section Neurocarpum were also misnamed.
However, many of these specimens were correctly identified to species,
but bore illegitimate synonyms. Relatively few problems were found
within Bentham's section Terñatea.
The purpose of this investigation has been to thoroughly
re-examine available specimens and types, to clarify the nomenclature
and species descriptions, and to prepare functional keys and
illustrations to legitimate taxa.

ECONOMIC IMPORTANCE
With ttie exception of C. Ternatea, the species within the genus
Clitoria have been of relatively little economic use, except where
they are locally important. Cl itoria ternatea is important
economically as a medicinal herb, as an ornamental, as a food and
cover crop, as a dye, as green manure, and as an aphrodisiac.
Clitoria ternatea has been prized as a cultivar at least since the
late seventeenth century, and has been carried by man throughout the
tropics, where it often has escaped and be :ome established. As a
result, this species now has a pantropical distribution. Attention
has been focused on C. Ternatea as a typical representative of the
genus because it is widespread, well known, and easily accessible for
economic use. Some species within the genus are reported to have some
of the same economic properties as C. ternatea, although these species
have not received the same publicity of £. ternatea. Additional
properties have been noted for a few species, tor example, the
possible use as powerful insect and vertebrate poisons.
The genus Clito_ria has great potential to be an economically
important legume genus. The economic uses of a particular species
properly belong under the treatment of that species. The purpose of
this chapter is to discuss the economic properties known for the genus
and to point the way toward research in the economic botany of
Clitoria as a whole. Clitoria ternatea will be the major
8

9
representative of the genus in this discussion. Information obtained
from herbaria is documented with the collector's name and his
collection number enclosed within parentheses, for example, (Hahn
234).
Medicina 1 Properties
Nearly all the medicinal properties of Clitori a are recorded for
the species C. ternatea as a result of investigations by workers
interested in medicinal Indian, Southeast Asian, and Indonesian plant
species. Remedies and treatments reported are grouped below by
medicinal use for each plant organ used.
Roots
Root or root bark is often reported to be an effective laxative,
but may not always be safe. The degree of cleansing action varies in
the reports. Dalgado (1896), Kirtikar and Basu (1918), and Dutt
(1928) report use of the roots as a laxative. Kirtikar and Basu (ibid)
cited Mr. Mooden Sheriff (1891) who noted that the root bark was a
laxative in some cases. Chopra, Chopra, Honda, and Kapur (1958) noted
that the root bark was used as a laxative. Chopra, Badhwar, and Ghosh
(1949) and Chopra, Chopra, Honda, and Kapur (ibid) reported the roots
as a powerful cathartic similar to Jalap (Exogonium purga,
Convolvulaceae), but that it was not a safe medicine. Dey (1896),
Burkill (1935). Quisumbing (1951) citing Dey, and Gardner and Bennett
(1956) reported the roots as carthartic. Dey reported preparing the
remedy as an alcoholic extract of the roots. Quisumbing (ibid),
citing Rajan (1926), reported the roots as a purgative and narcotic
toxin. Rajan (ibid) indicated symptoms of the narcotic poisoning to

10
include unconsciousness attended with exti erne irritability and a
peculiar loss of memory. Chopra, Badhwar, and Gosh (ibid), citing
O'Shaughnessy (1841), indicated that an alcoholic extract of the root
acts as a brisk purgative in doses of five to ten grains, but produces
gripping and tenesmus, and the patient will feel feverish and uneasy.
Quisumbing (ibid), citing Rajan (ibid), Sanyal and Ghose (1934), and
Burkill (ibid), reported the roots as an aperient. Roots were noted
to be a purgative medicine or remedy used in Martinique (Ha hn 2_34).
Roots are also used in stimulating urination and are useful in
ascites and fevers. Rajan (1926), Dutt (1928), citing the Hindus,
Sanyal and Ghose (1934), Quisumbing (1951), citing the three earlier
works, and Chopra, Chopra, Honda, and Kapur (1958) reported the roots
as diuretic. Chopra, Chopra, Honda, and Kapur (ibid) also reported
the root bark used as a diuretic. Kirtikar and Basu (1918), citing
Sanskrit writers, described che root as diuretic, useful in ascites
(an accumulation of excess fluid in the abdominal cavity) and fever.
They also cite Mooden Sheriff who noted that an infusion of the root
bark acts as a diuretic.
Quisumbing (1951), citing Sanyal and Ghose (1934) who cited
Mooden Sheriff (1891), and Kirtikar and Basu (1918), who also cited
Mooden Sheriff, reported that Sheriff spoke highly of the infusion of
the root bark as a demulcent in cases of irritation of the bladder and
of the urethra.
Roots are also used as a medicine in promoting menstrual
discharge. Quisumbing (1951) cited Crevost and Petelot (1929) who
reported the roots used as an emmenagogue.

11
Roots or root juice has been reported to cause nausea and
vomiting and to be useful for lung ailments and removing the phlegm
of chronic bronchitis. Lindley (1938) and Gardner and Bennett (1956)
reported the root as an emetic. Quisumbing (1951), citing Dalgado
(1896) and Crevost and Petelot (1929), indicated that each reported
the roots used as a vomitive. Drury (1873) and Dymock (1885), each
cited by Quisumbing (ibid), quote Mr. Ainslie (1826) who mentions the
use of the root in soup to remove phlegm in chronic bronchitis and to
bring on nausea and vomiting. Kirtikar and Basu (1918), citing Dymock
(ibid), indicated that in the Concan region of India, two tolas^ of
the root juice are given in cold milk to remove the phlegm of chronic
bronchitis through nausea and vomiting. Chopra, Badhwar, and Ghosh
(1949) indicated that opinions differed with regard to the emetic
properties of the root. In the Sudan, the roots are chewed for lung
and throat complaints (Prichard 4-).
Dymock (1885), cited by Kirtikar and Basu (1918), indicated that
the juice of the root of a white-flowered form of C^ ternatea is blown
up the nostrils as a headache remedy. A similar practice in the Sudan
has other effects. Before a cow is sent to the bull, the root is
chewed and blown up her nostrils to ensure fertility (Pritchard 4).
Roots are reported to be used as an antidote against snake-bite
and scorpion-stings. Chopra, Badhwar, and Gosh (1949) reported the
roots as a possible antidote against snake bites. Chopra, Chopra,
Honda, and Kapur (1958) reported that the roots were used in Indian
1. A tola is an Indian unit of weight equal to one silver rupee,
or 180 grains.

12
indigenous medicine as a treatment for snare-bites (Cobras and
Daboias) and for scorpion-stings (genera Buthus and Palamnoeus).
However, they found that remedies administered to dogs and rabbits
injected with the snake or scorpion venom in strict conformity with
the directions given in standard Indian medicinal books, had no
preventive, antidotal, or therapeutic effect.
Leaves
The leaves are reported to be used against eruptive conditions
by Chopra, Badhwar, and Ghosh (1949). Kirtikar and Basu (1918, citing
Watt (1889-96), and each later cited by Quisumbing (1951), reported
that an infusion of the leaves is used for eruptive conditions.
Heyne (1927), later cited by Burkill (1935) and Quisumbing (1951),
reported th.it the leaves of a white-flowered form of C. ternatea are
used as poultices in Jaya. Kirtikar and Basu (1918) and Quisumbing
o
(ibid), each quoting Mr. Taylor, stated that the juice of the leaves
mixed with green ginger is administered in cases of colliquative
sweating in hectic fever'. Quisumbing (ibid) quoted Kirtikar and Basu
(ibid) who quoted Mr. Mukerji (1889-1904) in "Watt's Dictionary" that
juice of the leaves is mixed with common salt and applied all around
the ear for earaches, especially when accompanied with the swelling of
neighboring glands. Quisumbing (ibid) quoted Tavera (1892) who
reported that the leaves are used in the Philippines for swollen
joints. Chopra, Chopra, Honda, and Kapur (1958) reported that the
juice of the leaves is mixed with water to form a jelly which is taken
2. Taylor was quoted by both authorities; however, neither one
documented the reference source of Taylor.

13
as a cool ini] medicine for gonorrhea, and used externally for eczema,
prurigo, and impetigo.
Seeds
Seeds are used also as a laxative and considered as an effective,
safer method than the use of roots or root bark. Waring (1868),
Drury (1873), Dey (1896), and Crevost and Petelot (1 929), each cited
by Quisumbing (1951), reported that seeds are used as a mild purgative.
Quisumbing (ibid) also cited Kirtikar and Basu (1918) who reported the
seeds as a purgative and an aperient. Chopra, Badhwar, and Ghosh
(1949) indicated that seeds in powdered form have purgative and
aperient properties, and are considered more useful and a safer
medicine than the roots. Quisumbing (ibid) cited Burki 1 1 ( 1 935) who
reported that the seeds are used as an aperient and contained a toxic
alkaloid. Quisumbing (ibid) cited Nadkarni (1927) who reported that
the seeds contained a fixed oil, a bitter acid resin, tannic acid,
glucose, and six percent ash. The testa was brittle and contained
cotyledons full of granular starch. Chopra, Chopra, Honda, and Kapur
(1958) cited Dymock, Warden, and Hooper (1890-93) who reported that
the seeds contained a fixed oil, a bitter resinous principle, and
tannin. (Chemical investigations on the seeds are reported under the
chapter "Prior Research on Clitoria." (Reier to that chapter,
subheading "Chemical Research," for details.)
Burki 11 (1935) quoted Tavera (1901) who reported that the seeds
are used in the Philippines as poultices for swollen joints. In an
earlier edition of his book (1892), Tavera had attributed this
medicinal property to the leaves.

14
Seeds are supposed to be effective in destroying or expelling
tapeworms. Waring (1868), Drury (1873), Dey (1896), and Crevost and
Petelot (1929), each cited by Quisumbing (1951), reported that the
seeds were used as an anti helminthic.
FIowers
Burkill (1935) cited "Medicinal Book of Malayan Medicine (Card.
Bull. S.S. 6, 1930, p. 381)" which reported that the floral juice of
a white-flowered form of C. ternatea was used for inflamed eyes.
Additional Notes
Chopra, Chopra, Honda, and Kapur (1958) reported that C^ ternatea
is alleged to have anti-dysenteric and anti-tubercular properties.
No plant organ was noted.
Aphrodisiacs
A few species are reported to ensure fertility and are used as
sexual stimulants. This topic is usually not included in the
literature possibly through delicacy complicated by the fanciful
resemblance of the flower to the female sexual apparatus. Yet, in
widely dispersed geographical areas with different species, the native
people use Clitori a plants as aphrodisiacs for man and domestic
animals. It is presumed that the natives follow the ancient
principles of the Doctrine of Signatures, which is based upon belief
that the plant structures which resemble portions of the human body
have been so structured as to advertise their ability to provide
remedies for ailments of those body portions.
The use of the plant in conception was implied by Rumpf (1747)
who noted that the Portuguese name "Fula criqua" was derived from the

15
flowers, and that Breyne (1678) had appropriately named the plant Flos
clitoridis ternatensibus (this plant now known as Cl itoria ternatea).
The Portuguese translation of "Fula criqua" means "to speed or
hurry-up creation." In the Sudan, before a cow is sent to the bull,
the root of C. ternatea is chewed and blown up her nostrils to ensure
fertility (Pritchard 4). Near Anaconda Island on the Rio Napo,
Ecuador, this author attended a wedding presided over by a tribal
witch-doctor. In a discussion with him the night before the wedding,
the flower of C_. pozuzoensis (a species that this author was searching
for locally) was described by cupping the hands for the vexillum and the
arching of the thumbs for the keel and wings. The grinning witch-doctor
apparently recognized the plant in question immediately. He accurately
described its white flowers, climbing habit, and vegetative morphology.
He reported through the translator that he gave this plant to couples
that had difficulty bearing children.
In Minas Gerais, Brazil, catuaba (C. guianensis) is put in
cachaca "to make a man out of you" (Will jams, Assis, and More ira
5416), which the collectors noted that the effect was "Potencia"!
Ornamenta!_ Properties
Most of the species of the genus Clitoria have potential for
becoming economically important as ornamentals because of their large
and beautifully-colored, showy flowers and the various habits they
possess. Frequently prized are those species which display climbing
habits, used commonly for garden trellises, and less frequently on
other vegetation. Also much esteemed are those species which possess

Ib
the "azure blue" flowers and those of "delicate shades of pink."
Other species are noted for their foliage. A few examples will be
cited here.
The oldest and most widely cultivated ornamental is C^ ternatea.
This species is probably indigenous to Eastern Africa with subsequent
expansion into central and western Africa and the Indian subcontinent.
Man has transported this species throughout the tropical belt where it
often has escaped and become established as a wild plant. Today it is
commonly cultivated and is found wild in many Pacific Islands, in
portions of Australia, Indonesia, Africa, and the Americas, and in the
Antilles. Breyne (1678) reported the species as a cultivar, taken
from the Mollucan island of Ternate and planted in European gardens,
where Curtis (1812) reported that it needed protection from the
climate, and that it would not set seed. Because of its brilliant
blue flowers of large size which bloom nearly year round, it is a
popular cultivar in tropical regions, growing rarely as a small shrub,
but most commonly as a perennial climber. The plant grows rapidly,
forming many intertwined stems, and when placed upon open ground,
develops tangled mats. It can be trained through cutting to form
low-growing shrubs. The white-flowered and pale blue-flowered strains
are not as popular as the azure blue-flowered strains.
Clitoria ternatea is also important because of a double-flowered
variety. These plants produce large, actinomorphic, free, banner-like
petals of deep azure to bright purple colors, and rarely white flowers.
This variety was first reported by Commelin (1701). Today it is
commonly cultivated and is naturalized in portions of the Caribbean,
Sri Lanka (=Ceylon), and Indonesia, with isolated populations
elsewhere in the neotropics.

17
Another early prized ornamental was C. heterophylla which was
reported by Curtis (1820). This species is endemic to the islands of
Madagascar and Mauritius. It is valued for its climbing habit, blue
flowers, and unusual pinnate leaves of two distinct shapes found upon
a single plant.
The most commonly cultivated, erect plant species has been
C. laurifolia (synonym: C. cajanifolia). It is native to the
neotropics and introduced into central Africa (naturalized in Zaire)
and Southeast Asia-Indonesia. In the latter areas, it is cultivated
as a shrub with several plants often planted closely together to
form a hedge.
Two woody lianas from the neotropics are popular in protected
gardens: C. arborescens of northern South America has large leaves,
silky beneath, many racemes of numerous purple flowers, and it is a
tall climber; javitensis of South America and Panama is valued for
its climbing habit and numerous lilac to delicate pink shaded flowers.
A less popular liana species, C. lasciva, a native of Madagascar, is
valued for its luxurient foliage and blue flowers.
The tall erect shrub of C_. amazonum is desirable in protected
gardens because of its erect habit with upper branches climbing, and
its huge, showy, purplish flowers which are numerous in each of many
inflorescences.
Clitoria fairchildiana (synonym: C_. racemosa Benth.) is a
popular tree species. It is endemic to the deltas of the Amazon River
and the other large rivers to the southeast. This tree is planted
extensively along beach roads in Rio de Janeiro, and valued for its
elongated racemes of numerous purple flowers, and for the luxurient

18
foliage borne on branches which arch and droop to the ground, forming
a canopy.
Po i sons an_d Pest i c i dej>
Several species are known to have toxic properties. It has
already been pointed out (Rajan, 1926) that the roots of C. ternatea
taken as a purgative can produce narcotic poisoning in man. Chopra,
Badhwar, and Kapur (1949) included CL ternatea in a list of important
plants poisonous to man and livestock.
Gardner and Bennett (1956) reported that C. arborescens was used
as a fish poison, the seeds containing alkaloids with weak curariform
action. The seeds of a related species from Colombia, C^ arbórea,
were reported as alkaloid positive (Schultes 24120). In Amazonas,
Brazil, froesii was reported as being considered the most toxic
plant in the region of Rio Icana, and was planted by the Indians and
found in fair quantities (Froes 12441/185).
The genus is not noted for its pesticidal properties, although
some species contain the insecticide ingredient rotenone. In Surinam,
the stems and fleshy roots of C^ falcaba have a strong odor of
rotenone (Archer 2848). A cultivated population of C^ laurifolia in
Sao Paulo, Brazil, was noted not to be attacked by fungus or insects
(Norris 278). In Thailand, tuber juices of C^ macrophylla are sprayed
on vegetables to kill green flies and root juice is used to kill worms
in the backs of buffaloes (Col 1ins 1441).

19
D_yes_and Fibers
Clitori a ternatea is the only species of the genus that is
reported to be used as a dye plant. Rumpf (1747) and Uphof (1968)
reported that in the Moluccan island Ambon, the flowers were boiled
with rice to give the rice a blue tinge. Burkill (1935) reported
identical use in Malaya. Burkill also noted that the juice of the
leaves is used on some occasions in the Dutch Indies to color food
green. He also reported that the flowers were used as a dye to give a
temporary color to white cloth. The dye was used in the Rhio
Archipelago (=Riouw Archipelogo, a portion of the Malaya Archipelago)
and as a reagent for detecting acid and alkaline solutions, as does
litmus paper.
Liana species are used locally by the natives to make ropes.
Uphof (1968) reported that C. lasciva stems are the source of a fiber
used by natives of Madagascar for making ropes. Clitoria sagotii
was noted as a bush rope by Jenman 4930, in Guyana.
Food for Man
Uphof (1968) reported that the pods of C. ternatea are consumed
as boiled vegetables.
Fodder and Green Manure
A number of references report that C. ternatea is a good forage
crop and is edible by cattle, sheep, and goats. Burkill (1935), for
example, reported that sheep and goats will eat the foliage. In
Venezuela, the species is used as a forage crop (Zambrano 105).
Several experimental studies have reported on the establishment of

20
these species in pastures, and other studies indicate its nutritional
value. These studies will be detailed in the later chapter "Prior
Research on Clitoriaunder the section "Agronomical Research."
Burkill (1935) reported that ternatea has been used as a green
manure, but that it climbs too much, although it was fairly successful
in some respects, on land that was to be fallow for a short time.
Burkill (1935) reported that laurifolia was only fair as a green
manure and that better plants existed for the purpose. Uphof (1968)
reported that this species was used as a green manure in warm
countries. Holland and Joachin (1933) found that laurifolia was
well adapted as a hedge plant to control terraces on steep slopes in
tea plantations in Ceylon (=Sri Lanka). Burkill (1935) noted that
the erect stems made this species well suited as a hedge plant on
contoured terraces.

HISTORY OF CLITORIA
The history of Cl i tori a begins nearly a century before its
officinal ly recognized origin in 1753. Although nomenclaturely
irrelevant, the pre-Linnaean names published and the accompanied
data presented provides better understanding of the genus and its
close relationships with the genus Centrosema.
Origin of_the Generic Name
The debatable origin of the generic name has been a very
distasteful subject to many botanists in the past. The most commonly
supposed origin of the name is from the Latin "clitoris," an
anatomical term by which it is indicated that the flower bears a
resemblance to the human female sexual apparatus. This unscientific
notion of comparing the flower with the female reproductive organs, a
practice noted by some as one of questionable taste or immoral
behavior, has been disturbing to a number of botanists. Solutions to
this problem have been various. First, ignore the problem by making
no reference to the subject. Second, circumvent the issue by
concluding that the origin of the name is lost in history, and today
has no significant meaning. Third, gently touch upon the problem by
indicating that the origin is from the flower's fantasized resemblance
to "a female organ," or to "a female structure," or to "an organ."
Fourth, circumvent and attack the problem by indicating that the
21

22
origin is from the Greek "clejsto: to enclose1 which refers to the
manner in which the organs of fructification a»e enclosed within the
keel. Or one can use as the origin, Clitor, a town in Arcadia
(Steam's "Botanical Latin," 1 966, has Arcadia = Arkadhia,
Peloponnisos, S. Greece, a European country side outside the
distribution of the genus). Fifth, a direct attack upon the problem
by indicating the immoral practice and unscientific method of
selection of the generic name, which should be selected to reflect
some botanical structure, to reflect a geographic origin, or to honor
a distinguished person by naming the genus after him. Hence the genera
established as substitutes, Ternatea Tourn. for Clitori us Petiver, and
Nauchea Desc. and Vexi11 aria Eaton substituted for Cl itoria L.
To this author, it seems clear that tne oi igin of the name can be
traced to Breyne (1678) who published Flos cl itoridis Ternatensibus
for an exotic cultivar from the Moluccan island of Ternate. He
indicated the vernacular name to be "Bokyni cotele de Principisse
Clitoris." Rumpf (1747) expanded upon this by indicating the Ternatice
names to be "Saja Cotele" and "Bokyma Cotele" from which Breyne erected
LlP§. cl itoridis (flowers clitoris-like) based upon cl i tori s principi sse
(or origin from clitoris), and from which the Portuguese name, Fula
criqua (to speed or hasten breeding or creation) were derived and
appropriately named. Tfie Portuguese and Ternatice names were probably
based upon the ancient "Doctrine of Signatures," a belief that plant
structures which resembled human structures were created to furnish
remedies for the ailments of those body structures. A number of
Clitoria species are used as aphrodisiacs, both for man and for
animals, in widely separated geographical ureas and cultures. Breyne

23
evidently recognized a resemblance in the flower to the female anatomy
as others hid, because ne in orporated tin com opt that the flowers
were clitoris-like in his polynominal name
G neral Histor.
Breyne (1678), a German botanist, was the first to describe and
to illustrate a representative of the genu. Clitoria. In "Exoticarum
Plantarum Centuria Prima" he described a plant, from the Indian
subcontinent and transferred to Europe for cultivation, as Flos
clitoridis Ternatensibus. Breyne describí I tht showy papilionaceous
flowers as inverted, with the vexillum ventral 1/ located, a unique
characteristic that segregates Clitoria firm must other legumes.
However, in the excellent illustration (pi tte I, drawn by Stech^),
the flowers are not resupina re (possibly <. . art ist's or engraver's
error). The vexillum appear'd dorsally ac in typical papilionaceous
flowers. â– ' Breyne's description included ha..it, calyx, bracteoles
(Breyne used, the term "leafl ts"), and fruit (not illustrated), major
characters used by later autnors in segregating Clitoria species. The
illustration clearly shows the odd-pinnati iy compound leaves and
persisting bracts and stipules (the stipul s are drawn too large for
the species), additional characters used t.-urly in the determination of
species. Breyne's representation is easil recognized as Clitoria
terna tea L.. the type species for the genu .. treyne reported the plant
from the Moluccan island of Ternate, and ited that it bore the name
1. Hunt [lot. Cat. 1:37-1 attribute dr .wing to Andreas Stech and
engraved by Issac Saal. Piale bears initi Is "AS" at left bottom,
"IS" at right bottom.

24
of "Bokyni i,otelé ele Principisse Clitoris, the source for Breyne's
polynominal name. Breyne reported a col If tion by Sevenhuysen in
1667 from Domingo (-Dominican Republic) as related to the Ternate
specimen. If verified as a 1itoria, it would represent an earlier
record of the genus.
Breyne also described aid illustrated (plate 32) a second
representative as Planta Leguminosa Brasil lana, Phaseoli facie, flore
purpureo ma/.imo. This specimen provided 1 ne earliest record of the
species' known as Clitoria brasiliana L., 1 iter transferred to and
becoming the type species of the genus Centrost.aia.
Plukenet (1691) in his ‘Phytographia described and illustrated
Foenuni graei.um phaseolo ides Virginiana fji.oe ampio caeruleo. This is
tiie earliest record of the species known , Cl itoria virginiana L.,
later correctly transferred to the genus i ntrv.sema.
Petiver (1704 ) in liis Appendix to hi- "Ca:ilogus Plantarum Hortis
Siccis Petiverianus" published two polynon ¡nal■ , both lacking additional
data. His number 54 was Cl i r.orius Mariano ; trt fol iatus viridis. No
record of this polynominal has been found iteci after this date. His
number 55 was Clitori us Marianus trifolia t is snbtus gl aucus. This is
the earliest record of the species known a . Cl 1 toria mariana L. Petiver
was the first to use Cl itoria or a form 01 the (jasic root word as the
"generic" name.
Tournefort (1706) in his article "Suite d< L1établissement de
queques Nouveaux Genres de Plantes" described the genus Ternatea. He
provided an illustration of the flower with cai/x, individual petals,
fruit, and seeds. Three members of the gt..us \ are included. His
first specit.'S was Tern a tea flore simpl ici, caei íleo based upon Breyne's

25
specimen. Breyne's specimen became the type for the genus and the
source of the generic name, which Tournefort indicated was named after
the Moluccan island of Ternate. Tournefort's second species was
Ternatea flore pleno, caeruleo based upon a double-flowered specimen
described by Commelin in 1697. His third species was Ternatea flore
simp!ici albido, a white-flowered specimen recognized for the first
time. All three species are now Cl itoria terjiatea L.
Dillenius (1732) in his "Hortus Elthamensis" described and
illustrated Clitorius trifolius, flore minore caeruleo. He segregated
this specimen from those of Breyne, Petiver and Tournefort, and
concluded that it represented a new member of the genus Clitori us as
defined by Petiver. His specimen is Clitori a virginiana L., although
no reference was made to Plukenet (1691), the earliest describer of
this species.
Linnaeus (1737b)in his "Hortus Cliffortianus" became the first to
use the generic name Clitoria. He recognized two species. His first
species was Cl itoria fo 1 iis pinnatis, based upon specimens described
by Breyne (1678), Rheede (1688), Plukenet (1691), Rivinius (1691)
Commelin (1697), Tournefort (1706), Dillenius (1719), and Burman
(1737). Breyne's Flos cl itoridis ternatensibus was the source of the
generic name. Neither Petiver (1704) nor Dillenius (1737), both of
whom had used Clitorius, was cited. The diagnosis was based upon
Phaseolus folij_s pinnatis Riv. The distribution was given as
Malabaria ( = S.IJ. India), Zeylona (=Ceylon, now named Sri Lanka), and
Ternate. Linnaeus recognized three varieties. His typical variety had
simple blue flowers (=Tournefort's first member). His variety alpha
was white flowered (Tournefort's third member). His variety beta had

26
double blue flowers (Tournefort1s second member). In addition,
Linnaeus gave a brief Latin description of variety beta. His second
species was Cl itoria foliis ternatis based upon Breyne's Brazilian
specimen and Plumier (1693).
Linnaeus (1748) in his "Hortus Upsaliensis" recognized the same
two species, but used binominal names. His first species Clitoria
foliis pinnatis became Clitoria ternatea. His species Clitoria fol i i s
ternati s became Cl itoria bra si 1 iajTa.
Nomenclatura!ly, the official history of the genus began with
Linnaeus (1753) in his "Species Plantarum." Linnaeus established the
genus Clitoria (p. 753) with four species: C. ternatea and C.
bra si 1iana from his earlier works ( 1 737b and 1748), with the addition
of C. virgin iana (based upon Dillenius, 1 737 , and Gronovius, 1 739) and
C. mariana (based upon Petiver, 1704). His brief Latin diagnosis for
each species was based upon leaflet number and calyx shape. India,
Brazil, Virginia, and northern America were given as the ranges of
his four species.
The International Code of Botanical Nomenclature (Stafleu, 1972)
officially recognizes that those generic names which first appear in
first edition (1 753) of Linnaeus's "Species Plantarum" (e.g. Cl itoria)
are associated with the first subsequent description given under those
names in the fifth edition (1754) of Linnaeus's "Genera Plantarum."
Clitoria is described under genus number 796 (p. 334) of Linnaeus's
"Genera Plantarum" as:
CAL. Peri anthi urn monophyllum, erectum, tubulatum,
quinquedentatum, persistens. COR. Vexil 1 urn maximum
rectum, emarginatum, margine undulatum, patens. A1ae
oblongae, rectae, obtusae, vexillo breviores. Carina

27
alis breviores. STAM. Filamenta diadelpha (simplex
& novemsidum). Antherae simplices. PIST. Germen
oblongum. Stylus adsendens. Stigma obtusum. PER.
Legumen longissimum, lineare, compressum, uniloculare.
SEM. plura, reinformia.
This diagnosis represents the first validly published description
of the genus Clitoria L.
Miller (1754) in his "Gardner's Dictionary" provided the first
encyclopedic reference work on the genus. He adapted Tournefort's
treatment (1706) and recognized the genus Ternatea. Thus, three of the
four species he recognized followed Tournefort's treatment, not
Linnaeus’s (1737) who treated these as three varieties of one species.
A fourth species, Ternatea Americana perennis flore caerulea from
Jamaica, was included. In a later edition of his work, Miller (1759)
adopted Linnaeus's generic name Cljtoria_ and his four species. The
three species of Tournefort included in his earlier edition were
included as ternatea in the later edition. The treatment of the
Jamaican species is unclear, as no reference to this specimen was
made. It may have been included under brasil iana or virginiana,
or treated under another genus, but it could not have been included
in C. mariana, a species which does not occur in Jamaica.
Miller described the flower as papilionaceous, with the vexillum
longer .than and hiding the keel and wings, the fruit opening by the
sutures, and the seeds kidney-shaped. The first description of the
stamens and stigma were given. He referred to the flower as a
butterfly type.
Buchoz (1775) in his "Histoire Universelle du Regne Vegetal ou
Nouveau Dictionnaire Physique et Économique de toutes les Plantes qui
Croissent sur la Surface due Globe" became the first to describe the

28
genus Cl i tor i a in greater detail. Cl i tona was described as: calyx
tubular, 5-toothed, persistent; corolla papilionaceous; vexillum very
large, straight, notched, with undulating margins; wings oblong,
straight, obtuse, much shorter than standard; carina much shorter than
wings, round underneath, furrowed; stamens diadelphous, anthers simple;
ovary oblong, style ascending, stigma obtuse; legume very long, linear,
flat, bivalved, awl-tipped; seeds numerous, reniform. Buchoz followed
the Linnaean treatment and included five species, C^ ternatea L.,
C. bras'i 1 iana L. , C. virqiniana L. , C. mar i ana L., and C^ gal act ia
Crantz. Each species was described with notes on origin, culture, and
economic importance.
Lamarck (1786) in his "Encyclopedic Methodique, Botanique"
described seven species, of which Clitori a heterophyl1 a (East Indies)
and Clitoris falcata (Santo Domingo) were newly described.
Descriptions lacked measurements, but included habit, stem, leaf
shape and pubescence, leaflet number, peduncle location and relative
size as compared to petiole size, calyx, flower size and color, and
fruits. All sizes were relative and compared to similar structures of
other species. These characters, along with the addition of bracteoles
and stipules, were to become the major characters utilized by later
authors in describing and segregating Clitoria species. Lamarck's
style of describing Cl i Loria species established a trend that continued
through Bentham's revision of the genus iri 1858 and beyond. Poiret
(1811), in a supplement to Lamarck's work, described three new species
and brought the total number to thirteen.
Desvaux (1813) in his journal article "Precis de Caracteres de
plusieures Den res, de la Famille des Legumineuses" described a new

29
genus Neurocarpum, with a tubular calyx of five, equal teeth and a
stalked fruit characterized by a longitudinal nerve on each side. He
indicated Aublet's Crotalaria gajanensis (published by Aublet as
Crotalaria guianensis) and a new species, N. ellipticum Desv., were
placed in the genus. In a succeeding article a year later, Desvaux
(1814) described the new species N. el 1ipticum and included
Crotalaria guyanensis Aubl. and Crotalaria 1 ongifol ia^ Lam. as
synonyms of a species called Neurocarpum janense Desv.
Eaton (1817) in his "Manual of Botany for the Northern States"
established the genus Vexi11 aria as a substitute for Clitori a because
J. E. Smith severely censured the name Clitoria (reference of the name
reflecting the female anatomy) in Rees' "Cyclopaedia" (1807).^ Rickett
and Stafleu (1959), in an expanded footnote regarding Vexillaria Benth.
( = Centrosema ), quoted Smith as:
CL I TOR IA, in Botany, (from xAetu>, claudo, includo),
expressing the manner in which the essential organs
of fructification are enclosed or shut up in the
keel and wings of the corolla. Whatever may have
been in the thoughts of Pet i ver, by whom the name
was first introduced into botany; or of the
illustrious naturalists, by whom it has since been
continued, reformed, or sanctioned, we cannot refrain
from entering our decided protest against every
attempt to associate it directly with an anatomical
term, to which, though derived from the same Greek
theme, it has in fact only a very remote,
2. Desvaux erred in his interpretation of Aublet's species.
Desvaux's type (P-59!) does not agree with Aublet's type (BM!).
Therefore, Aublet's name is not synonymized with Desvaux's name. A
complete discussion can be found under the type treatment of subgenus
Neurocarpum, p.
3. This author has not been able to obtain a copy of Eaton's
publication, nor Smith's article. However, a copy of the eight edition
of Eaton's Manual (1840) has been obtained. Clitoria was not mentioned.
Vexillaria Eaton contained three species, virginiana , V. mariana,
and V. plumieri, each treated as a Cl itoria species by almost all other
botanists.

30
fictitious analogy. It is greatly to be lamented,
that a fondness for these gross allusions should
ever have been indulged by any, who, in all other
respects, have deserved highly of natural science,
and whose splendid talents should have rendered
them far superior to such grovelling ideas. By
this conduct they may have done all in their power
to pollute a study, which is, perhaps, more than
all others, suited to the loviest part of the
human race, and which, without concealing any
essential part of the sexual system, may easily be
so conducted, as not to excite an unpleasant
sensation in the most delicate female mind. We
do not mean to exempt from the full severity of
this censure our great master, Linnaeus, himself;
for when, not only the purity of moral feeling,
but also the common decorum of polished life, is
infringed, the nul1 is in verba of the poet will, we
trust, be uniformly our principle and our practice).
Vexi11 aria Eaton does not appear in Willis' Dictionary, Index Kewensis,
Gray Card Index, nor other standardized index references, nor in
articles associated with â–¡ itorj_a_. Vexil 1 aria Eaton apparently was
not accepted by other botanists of the time, and was thus easily
overlooked by the compilers of the major index references. Vexi 11 aria
Eaton is a probable synonym of the genus Clitori a.
Rafinesque (1818) in a two-line note in a journal article remarked
"Cl itoria mariana must form a particular genus Vexi 11 aria ." Thus
Vexillaria Raf. was established and recorded by the major standardized
index references. Other botanists did not agree with Rafinesque and
neither. Vexi 11 aria mariana Raf. was used nor were other species of
Clitoria transferred to the genus Vexillaria Raf.
Leandro do Sacramento (1821) in a journal article described a new
species under a new genus Martia, named in honor of the German botanist
Carl F. P. Martius. The description of Ma_r_tia physod_es was detailed
with an illustration provided. The generic characters which indicated
a distinct genus were the lack of a corolla, small stamens with

31
rudimentary filaments, a style in contact with its own anthers, and a
legume with a protruding lateral nerve. The description was based
upon a cleistogamous flowering collection of Cl itoria falcata from
Brazi1.
Schultes (1822) in his “Mantissa in Volumen Primum Systematis
Vegetabilium" published the species of Leandro do Sacramento under the
name Martiusia physalodes. Perhaps Schultes changed the generic name
because of the homonym Marti a Sprengel (1819) which is synonymous with
a non-legume genus.
Humboldt, Bonpland, and Kunth (1824) in their "Nova Genera et
Species Plantarum" described four new species of Neurocarpum. A new
genus was cited in synonymy as "Rhombolobiurn Rich, mss.," probably
based upon a Paris specimen of Richard (P-32!) whose label bears the
generic description of Rhombol obi uni, and less likely based upon a
Geneva specimen of Richard (G-545!) whose label bears the data
"Rhombolobium Richard, genre nouveau qui doit etre place a coti du
Cl itoria. 11 y a un angle que . . . [handwrit ing illegible for this
term] du milieu du Legume du chaqué coti.' Clitoria L. and Ternatea
Tourn. were treated as distinct genera by Humboldt, Bonpland, and
Kunth.
A. P. de Candolle (1825) published the first major revision of
Cl itoria in his "Prodromus Systematis Natural is Regni Vegetabi1is."
Almost published simultaneously was his "Memoires sur la Famille des
Legumineuses" which gave explanations of his reasons underlying the
classification adopted in his "Prodromus." A. P. de Candolle
characterized the genus Cl i tear ia as: caly . of five teeth with a pair
of bracteoles at its base; the corolla and stamens inserted at the

32
calyx base at a point visible to the exterior; a large round standard;
a style more or less dilated at the summit; a resupinate flower; and
a fruit of two valves with seeds separatee by a cellular partition.
The genus Ternatea Tourn. was placed in synonymy. A. P. de Candolle
recognized four sections within the genus, segregated by the calyx
shape, bracteoles, presence or lack of a vexillum spur, and a leaflet
number.
A. P. de Candolle called section I "Ternatea," based upon Kunth's
treatment of Ternatea in 1824. This section was characterized by a
tubular calyx, spurless vexillum, and five or seven leaflets. Two
species were included, Clitoria heterophyl la Lain, and Cl i tori a
ternatea L. A. P. de Candolle named section II "Euclitori a," based
upon his statement that this section included the true clitorias.
Section II was characterized by a tubular calyx, spurless vexillum,
and three leaflets. Six species were included: Cl itoria mariana L.,
Cl itoria mexicana Link, Clitori_a angustí'fol ia H.B.K., Cl itoria
formosa H.B.K., Cl itoria glycinoides DC., and Clitoria poitaei DC.,
the latter two newly described species. It should be noted that the
type species for the genus, C^ ternatea L., is not in this section,
and thus the section cannot contain the "true" clitorias. Section III
was named "Centrosema" by de Candolle, based upon the spurred vexillum,
a unique characteristic in the legumes. Section III was characterized
by a campanulate calyx of five divisions, a spurred vexillum,
longitudinally striated bracteoles, and three leaflets. Three species
were included: (3111 or i a virginiana L., Clitoria brasil iana L., and
Cl itoria pi urnieri Turp. Section IV was named "Glycinops is" by
de Candolle, based upon the resemblance to the genus Glycine.

33
Section IV was characterized by a campanulate calyx of five teeth,
corolla and stamens inserted at the calyx ease, style dilated at
summit, longitudinally striated bracteoles, and three leaflets. One
new species, Clitorija berter iana DC., was placed here. A. P.
de Candolle also noted five additional species that may possibly
belong in the genus Cl i toria, bringing the total number of species
included to seventeen. The five species that were briefly noted, and
not placed in any of the four distinct sections, were: Clitori a
coccinea Schrad., Clitori a spec i osa Cav. , Cl[tori a laurifolia Poir.,
Clitoria vicio ides Nees. & Mart., and Clitoria arborescens Ait.
For each species placed in a section, de Candolle gave a brief
Latin description in his "Prodromus." He included the general
distribution of the species and cited collections he had seen for
the new species.
A. P. de Candolle treated the genus Neurocarpum Desv. as distinct
from Clitoria L. in his "Prodromus" following the tradition of other
botanists of his time. He characterized the genus Neurocarpum as:
calyx tubular with five, subequal, acuminate teeth bearing two
bracteoles at its base; vexillum large, subrotund; stamens diadelphous,
legume stipitate, compressed, subtetragonous, valves bearing a
prominent medial longitudinal nerve; and with seeds in cellular
partitions. Seven species were included: Neurocarpum simplicifolium
Kunth, Neurocarpum angustí' fo I ium Kunth, Neurocarpum guianensis Desv.,
Neurocarpum elliptic urn Desv., Neurocarpum javitense H.B.K., Neurocarpum
macrophyl1um H.B.K., and Neurocarpum falca turn (Lam.) DC. A. P.
de Candolle included a new genus, RhombifoIium Rich., in synonymy
based upon a specimen in Richard's herbarium.

34
A. P. de Candolle treated the genus Martiusia Schult. as also
distinct from Cl itoria and Meurocarpum. He characterized the genus
Marti usia as calyx slightly bilabiate with five teeth, lowermost
longer; corolla lacking; stamens four with two sterile and two fertile,
subciliate anthers; filaments much shorter than the ovary; legume
stipitate, compressed, subtetragonous, bearing a medial longitudinal
nerve. A. P. de Candolle noted that the calyx and legume were those
of Neurocarpum, and that the genus may not be distinct from Neurocarpum.
He included only one species, Martiusia physalodes Schult. The genus
Martia Leandr.-Sacr. was treated in synonymy.
Descourtilz (1826) in his journal article "Sur les Mauchees,
Genre Nouveau dans la Famille des Legumineuses" criticized the use of
the name Clitoria and its reference to an organ from which the name
was derived. He felt that the botanical language should consecrate
the memory of those illustrious persons of importance by adopting
names which recognized their contributions. He proposed the
substitution of the name Nauchea for Clitoria. The genus Nauchea was
named in honor of M. fe docteur Nauche.^ Twelve species of Cl itoria
were transferred to Nauchea and a new species was described. Each
species included a description and a distribution. Descourtilz gave
each species a common name, based upon the specific epithet (i.e.
Nauchea multiflora was "Nauchee a fleurs nombreuses"). Other botanists
did not adopt the generic name of Nauchea.
Desvaux (1826) in his article "Observations sur la Famille des
Legumineuses" proposed a revisionary change in the genus Neurocarpum.
4. A prominent French physician, Jacques-Louis Nauches (1776-
1843), who pioneered work in electrical shock therapy.

35
Two species were placed near those described in 1813. They were
Neurocarpum barbaturn Nees, newly described, and Neurocarpum laurifolium
Desv., a transfer of Poiret's species (thus should have been N.
laurifolium (Poir.) Desv.). Three species were recognized as different.
They were Neurocarpum rubiginosum Desv., a transfer of Jussieu's
species published by Persoon (thus should nave been N. rubiginosum
(Juss. ex Pers.) Desv.), Neurocarpum glycinoides N.,^ a transfer of
de Candolle's species (thus should have been N. glycinoides (DC.)
Nees), and Neurocarpum vijlosum N., newly described. Desvaux proposed
the name Pi Ianthum for this group based upon Poiret's name used for
the type of C. glycinoides DC. (=N. glycinoides Nees). The genus
Pilanthus Puir. ex Endl. (Genera PIantarum p. 1289, 1841), presently
synonymized with the genus Centrosema, mag be based upon Poiteau's
name Pilanthos, although Endlicher's article has not been seen. The
specimen ret erred to by Desvaux and also b/ de Candolle (Pi 1anthum
cited under C_. glycinoides in his Prodromus 2:234, 1825), bears the
name Pi}antnos tetrágonos Poiret. Two new species of Clitoria were
also described, C. laurifolia Nees, a homonym for C^ 1aurifolia
Poiret (1811), and C^ sinuata Nees.
Don (1832) in his "A General History of the Diehlamydeous Plants"
followed de Candolle's treatment of Clitot ia, Neurocarpum, and Martiusia
with minor revisions. Of the five dubious species listed by
de Candolle under the genus Clitoria, Don placed two in the section
Cen tro sen ia (C. speciosa Cav. and C._ arborescens Ait.), one in the
section Glycinopsis (C. coccinea Schrad.), and left one in a dubious
5. Desvaux's use of N. was an abbreviation for Nees.

36
list (C. vicio ides Mees & Mart.)- The filth species was transferred
to the genus Neurocarpum (C. laurifolia Poir.) along with C.
glycinoides DC. from section Euclitoria. He also described two new
species, Clitoria racemosa and Clitoria alba, and placed them both in
the section Euclitori a. No changes were made in the genera Neurocarpum
and Mart i usia with the exception of the transfers already noted.
Bentham (1837) in a journal article on legume genera followed
de Candolle's treatment of these genera with some revisionary changes.
There were three major changes in the genus Clitoria. First, Bentham
segregated the section Centrosema from Clitoria and elevated the taxon
to the level of genus, using de Candolle's name Centrosema. Succeeding
botanists have adopted this revision, and presently, more than a
century later, botanists continue to recognize both genera, Cl i tori a L.
and Centrosema (DC.) Benth. Second, Bentham eliminated the section
Glycinops is by transferring Clitoriana berteriana DC. to a
non-cl i torian genus, Periaridra, and placed Cl i toria coccinea Schrad.
in a dubious status. Succeeding botanists adopted this revision as
the section Glycinopsis disappeared from literature after this date.
Third, Bentham recognized de Candolle's sections Ternatea and
Neurocarpum, and added to them a new section, Bractearia Mart, ex
Benth. Bracteria was described as fruticose species with trifoliate
leaves and bracteoles equal to or longer than the calyx. To this new
section, Bentham transferred Clitoria racemosa Don and Clitoria
poitaei DC. He also described three new species and placed them in
this section, Clitoria amazonum, Clitoria acuminata, and Clitoria
arbórea.

37
Ben tham's treatment of section Terna tea followed de Candolle and
Don with the addition of a new species, Cl itoria lasciva Boj. ex Benth.
Section Euc1itoria underwent minor revision. Clitoria formosa H.B.K.
and Clitoria angustífo 1ia H.B.K. were transferred to the genus
Centrosema. Bentham added a newly described species, Clitoria
pedunculata Boj. ex Benth., to the two remaining species, Clitoria
mariana L. and Clitoria mexicana Link. Bentham also added notes on a
number of miscellaneous species published as Cl itoria, but which had
been transferred to other genera.
Bentham's treatment of the genus Neurocarpum followed de Candolle
and Don with two important changes. First, Bentham placed the genera
Martia Leandr.-Sacr. and Martiusia Schult. in synonymy. Second,
Bentham remarked on the close relationship between Neurocarpum and
Clitoria by stating that Neurocarpum is a Clitoria except that it has
a different legume (i.e. Neurocarpum has a costate legume; Clitoria
has an ecostate legume). Eleven species were recognized and placed in
two unnamed, artificial groups. Group 1 had suberect stems whereas
group 2 had procumbent or cl imbing stems. Five species were newly
described and several species were placed in dubious status.
Bentham (1839), in a publication on British Guianan plants
collected by Schomburgk, substituted the name Dendrocyamus for his
section Bractearja in the genus Clitoria. He reasoned that the name
Bracteria was inappropriate since it had been given to a genus of the
Rubiaceae and had been used as a section of the genus Chaetogastra
(Melastomaceae).
Bentham (1858) published a second revision of the genus Clitoria
in a Journal of the Linnean Society article. This was the last revision

38
of the genus until this work over a century later, and was adopted by
all succeeding botanists in their treatment of the genus. The major
change from his earlier revision (1937) was the uniting of the genus
Neurocarpum with Clitoria. Bentham reasoned on p. 35:
In a systematic point of view I had formerly
endeavoured to render Clitoria more natural, by the
elimination of DeCandolle's section Centrosema; and
I now find it necessary for the same purpose to unite
it with the Neurocarpum of Desvaux, hitherto universally
adopted by other botanists, myself included. This
entails the giving up, as a generic character, one which,
in Leguminosae, is usually considered as absolute, the
raised longitudinal nerve or wing along the center of
each valve of the pod. It is the same peculiarity which
has induced the separation of Tetragonolobus from Lotus
among European plants. But in the division of Lotus, as
well as in that of Clitoria, this purely technical
character is unaccompanied by any other differences, and
I now have instances in Clitoria where it is inconsistent
in one and the same species, and even on the same
specimen.
Bentham continued to recognize three sections, although he changed
their names. Section Terna tea remained unchanged, whereas Neurocarpum
was substituted for Euclitoria, and Clitorianthese was substituted for
Bractearia (=Dendrocyamus in 1839) used in Bentham's earlier revision
(1837). No comment was made on the change in section names, nor was
there any reference to the earlier sectional names. However, a
comparison of the two revisions by Bentham indicated that the sectional
diagnosis agreed (although the 1837 description was expanded upon in
1858), the species placed in each section agreed, the order in
which each section was treated agreed, and the number given by Bentham
for each section agreed.
Section Terna tea was characterized as: stems herbaceous or
rarely fruticose, prostrate, twining or climbing; leaflets 5-9, rarely

39
3; legume valve flat or slightly convex, non-costate, seeds subreniform,
compressed, smooth. Six species were recognized, all known previously.
They included Clitor i a lasciva Boj. ex Benth., Clitor ia ternatea L. ,
Clitoria pilosula Wall., Clitori a heterophyl1 a Lam., Clitoria bif1 ora
Dalz., and Clitoria pedunculata Boj. ex. Benth.
Section Neurocarpum was characterized as herb, twining or
prostrate or short erect; leaflets 1 or 3; legume valve convex, medial
longitudinal costa or rarely ecostate; seeds globose, ovoid, or thick
subreniform, with a viscid coat. This section contained eleven species
placed into three groups based upon the stem habit. The groups were
unnamed and distinguished by one to three symbols. The first group
with twining stems included: Clitoria inacrophylla Wall., Clitoria
mar i ana L., and Clitoria glycinoides DC. The second group with
prostrate stems included two newly described species: Clitoria
flagellaris and Clitoria rufescens. Group three with stems arising
from a lignose rhizome, erect to ascending, included: Clitoria
cajanifolia Presl , Clitoria simplicifolia (Kunth) Benth., Clitoria
guianensis (Aubl.) Benth., Clitoria densiflora (Benth.) Benth., and
Clitoria stipularis (Mart.) Benth., plus a newly described species,
Clitoria nana.
Section Clitorianthes was characterized as erect shrub or tall
climbing; leaflets 3; legume valve flat or slightly convex, coriaceous;
seeds in matured state unknown. This section contained nine species
divided into two unnamed groups based upon the bracteoles. The first
group with narrow bracteoles, or bracteoles much shorter than the calyx,
included Clitoria polystachya Benth. , Clitoria brachystegia Benth. ,
Clitoria arborescens Ait., C1 itoria javitensis (H.B.K.) Benth., plus

40
two newly described species, Clitoria se 11 oi and Clitoria leptostachya.
The second group with ovate, coriaceous bracteoles subequal to the
calyx, included: Clitoria amazonum Mart, ex Benth., Clitoria racemosa
Benth., and C] itoria hoffrnanseggi i Benth.
The treatment of each species included a Latin diagnosis, the
distribution as then known, the citation of one to a few collections,
and synonymy. Characters used in the diagnosis of the species
included: stem habit; leaf shape and pubescence, and leaflet number;
peduncle size, location, and number of flowers borne, calyx;
bracteoles; legume curvature and costa; and sometimes stipules or
vexill urn pubescence, hen t ham occasionally gave additional descriptive
data with a minimum quantity of measurements.
After Beritham's revision, the next century of publications on
Clitoria was mainly of a floristic nature, not revisionary. Isolated
descriptions of newly described species placed in the genus also
occurred. Two exceptions to the lack of revision are worth noting.
First, Baker (1879), in a floristic treatment of legumes in "Hooker's
Flora of British India," elevated Bentham's sections to the level of
subgenus. Subgenus Ternatea was character ized as having flat legumes
without a costa. Clitoria mar i ana L. and Clitoria macrophylla Wall,
were transferred here from Bentham's section fleurocarpum. Subgenus
Neurocarpum was characterized as having a turgid, costate legume.
Bentham's section Clitorianthes was not treated, probably because the
species within this section do not occur in the Paleotropics. Second,
Kuntze (1891) transferred all Clitoria species to the genus Ternatea,
a revision not adopted by later botanists. Occasionally, floras would
include Clitoria and Martiusia as separate genera, the latter genus

41
containing the species of Neurocarpum (e.g., Small, 1933; Britton,
1924).
In the earlier parr, of the twentieth century, research in various
areas began on Clitoria (e.g., agronomy, anatomy, chemistry, cytology,
development). The results of these studies are enumerated in the
chapter "Prior Research on Clitoria."
These selected articles represented a partial, highlighted history
of the genus Clitoria. The important portions of the nomenclatura!
history-of the genus, above i;he level of species, are summarized in
Figures 1 and 2. For both figures, genera were expressed in capital
letters and enclosed in a "box" for better visualization. Genera
included in synonymy were enclosed in parentheses. For the few
polynomials included, which are important in tracing the origin of
some generic names, the key terms are used and underlined, followed
by "..." which represented the missing, nonessential terms. Sections
are numbered and subgenera are preceded by the abbreviation "Subg."
Hi_storical Chrono_logy
A more complete historical chronology is presented below. Each
article is presented in a chronological order, by year, with a brief
synopsis of the publicación, and nomenclatura! notes. Those articles
marked (*) are expanded upon in the "General Flistory" section
previously discussed, and the reader is referred to its coverage.
Those articles marked (R) are expanded upon in the next chapter, "Prior
Research on Cl itoria,11 and the reader is referred to that chapter for
details.

PRELINNEAN TO de CANDOLLE TREATMENTS (1768-1825)
1813, Desvaux
r
I NEUROCARPUM
1825, de Candolle (*)
l
* NEUROCARPUM
IV. Glycinopsis
Figure 1. The history of Clitoria, pre-de Candolle.

REVISIONARY TREATMENTS (1825-1900)
1825, de Candolle (*'
* NEUROCARPUM
1827-9, Bentham (-)
1858, Bentham (0)
¿ NEUROCARPUM
1
* MARTI USIA
(MARTIA)
fMART I USIA 1 !~^
j \
1891. Kuntze
(MARTIA) 1 \
A TERNATEA
\
/
IcjlIu, uescuur l . i i
r
'[NAUCHEA
f ClI iORIa
I. Ternatea
I. Ternatea-
I. Ternatea-
II. Euclitoria-
->II. Euclitoria-
11. Neurocarpum
1879, Baker
•j CLITORIA |
-^-Subg. Ternatea
->Subg. Neurocarpum
T T T n ^ f
-> t t t H
III. Centrosema
IV. Glycinops is
C. berteriana
i? CENTROSEMA
0 CENTROSEMA
# PERIANDRA
P. species
Key:
I! II
Clitoria above line, non Clitoria below line
Figure 2. The history of Clitoria, post-de Candolle.

44
P rjSj Lj nnean (1650-1753: Pol'/nominal Peril; I)
1 667: Sevenhuysen was reported by Bt eyne (1678) to have collected
a specimen i rom Santo Domingo that was related to the Mol ucean specimen
Flos clitoridis terna tensibus. If this specimen is verified as a
Clitoria, this would be the earliest known record of the genus.
1678: Breyne (*) described and illustrated two specimens that are
the framework upon which Linnaeus established two of the original four
species of i 1 i tori a. His Table 31 described FI os clitoridis
ternatensibus ( = Cl_itorJj ternatea L. ) from the island of Ternate,
Moluccan Islands. His Table 32 described Planta leguminosae
Brasiliana , phaseoli facie, flore purpureo máximo (= Clitor i a
brasil iancp L. ).
1688: Rtieede tot Draakestein described Schonga cuspi from Malabar
(=S.W. India) which Linnaeus later cited in synonymy for Clitori a
ternatea. from its description and a photograph of Schongi cuspi as
illustrated in his Table 38, one can easily conclude that this plant
is not a Clitoria specimen.
1691: Plukenet published Foenum graecum phaseoloides Virginianus
flore ampio caeruleo which is the origin of the third original Linnaean
species, CJ itoria virejiniana L.
1 693: Plumier described Phaseol us ampl_e f 1 ore pel tato, sil iquis
nigris et angulosis (=C1itoria plumieri Turp.) based upon a specimen
from Santo Domingo.
1696: In his catalogue of Jamaican plants, Sloane included
Phaseolus minor lactescens flore purpureo (-C1itor i a gal act ia Crantz).
1698: Dr. David kreig brought a specimen to Great Britain that
was collected in Maryland. Ihis specimen is now deposited in the Royal

45
Botanic Garden Herbarium at Edinburgh (E-127, Herb. Dubois!). This
specimen represents the oldest known representative of the genus
Cl itoria (= C. mariana L.).
1700: Plukenet published Phaseolus Indi cus, coeruleus,
Glycyrrhizae foliis alatis, flore ampio Cl i ferio (=Clitor i a ternatea
L.) and Phaseol us fl ore vexil lo, si 1 ique rectis tereti bus (= Cl i toria
brasiliana L.).
1701: Commelin described and illustrated in his Table 24,
Phaseolus Indicus, glycyrrhizae foliis, flore ampio, caeruleo, pleno.
This was the first known example of the unique double flower of
Clitoria ternatea L.
1703: Dr. Edward Bulkley sent a specimen to Great Britain that
was collected from Fort St. George in the East Indies. This specimen
is now deposited in the Royal Botanic Garden Herbarium at Edinburgh
(E-69, Herb. Dubojjs!). This specimen represents the oldest known
specimen of the type species of the genus Cl itoria (=C. ternatea L.).
1704: Petiver (*) published two polynomials in an appendix, the
first to use "Cl itorius" as the leading key term of the polynomial.
His number 54 was Cl itorius Marianus trifol iatus viri_d_i_s which
disappeared from Cl itoria literature. His number 55 was Cl itorius
Marianus trifoliatus subtus glaucus which was the origin of the fourth
original Linnaean species, Clitoria mariana L.
1706: Tournefort (*) established the genus Ternatea based upon
a specimen from Ternate, a Moluccan Island. Three elements were
published which later served as the framework for the three varieties
of Clitoria ternatea L. These were Ternatea flore simp!ici, caeruleo

46
(agreed with Breyne, 1678); Ternatea flore pleno, caeruleo (agreed
with Commelin, 1701); and Ternatea flore s imp!lei albido.
1709: Petiver published Plukenet's name (1691) and illustrated
a specimen in his Table 104, Figure 19.
1732: Dillenius (*) described and illustrated Clitorius trifolius
flore minore caeruleo (=C1itoria virqiniana L.) which he concluded was
segregated from, yet related to the specimens of Breyne (1678) and
Petiver (1704). No reference was made to Plukenet (1691).
1737: Burmann reported Qos clitorius flore coeruleo ( = C1itoria
ternatea L.) from Ceylon (=Sri Lanka presently) citing Breyne (1678)
and Plukenet (1700).
1737: Linnaeus (*) established the genus CJjtoria^ and recognized
two species. The first species was Clitoria foliis pinnatis (=C1itoria
ternatea L.) based upon Breyne (1678), Rheede tot Draakestein (1688),
Plukenet (1700), Commelin (1701), Tournefort (1706), and Burmann (1737).
Three "varieties" were recognized using Tournefort's segregation. The
second species was Clitoria foli is ternatis (=C1itori a brasiliana L.)
based upon Breyne (1678) and Phaseolus flore vexillo amplitismo,
fil iquis rectis teretibus published by Plumier in "Spec. 8."^
1 739: Gronovius published Clitoria foliiji ternatis, calycibus
obiongus, from Virginia (=U.S.A.), citing Petiver (1704) number 55 in
synonymy, lie provided the first record of a specimen citation for
Cl itoria, Clayton 108 (--Cl itoria mariana L.). He also reported
Clitoria foliis ternatis, calycibus campanulatis, based upon Dillenius
(1 732), and Clayton 112 (= Clitoria virqiniana L.).
6. This author was unable to establish the publication title and
therefore was unable to obtain a copy of Pluinier's publication.

47
1740: Royen reported three species from Batavia (=Java). The
first species was Clitoria foliis pinnatis of Linnaeus (1737). The
second species was Clitoria foliis pinnatis, caule decumbente (=non-
Clitoria). The third species was Clitoria foliis ternatis of Linnaeus
(1737).
1747: Rumpf (Rumphius) described "Bongu Biru" from the island of
Amboin, a Moluccan Island, as Flos coeruleus (-C1 itoria ternatea L.).
Vernacular names in several languages were noted.
17'48: Linnaeus (*) established binominal s for his two species
published in 1737. The two species became Clitoria ternatea and
Clitoria brasiliana, respectively.
Post-Linnaean (1753-1824: Descriptive Period
1753: Linnaeus (*) officially established the genus Clitoria,
and recognized four species, Clitori a ternatea and Clitoria brasiliana
based upon his prior works (1737; 1748), Clitoria virginiana based upon
Dillenius and Gronovius, and Cl itoria maraña based upon specimen
number 55, Pet i ver (1 704 ). India, Brazil, Virginia, and northern
America were given as the ranges of the four species.
1754: Linnaeus (*) published a description of the genus
Clitoria. This description is the first description published that is
officially recognized by present botanists through the rules established
in the International Code of Botanical Nomenclature (Stafleu, 1972).
1754: Miller (*) in his fourth edition of “Gardner's Dictionary"
followed Tournefort (1706) in the treatment of the genus Ternatea. Thus
the genus Ternatea Tourn. ex Miller became officially established. In
the seventh edition (1759) Miller adopted the Linnaean treatment. This

48
was the first encyclopedia type of reference work on the genus.
Descriptions and horticultural notes were provided.
1756: Browne reported two species from Jamaica with Linnaean
polynominals cited in synonymy. They were Clitoria major scandens,
foji is subrotundo-ovatis, floribus geminatis (=C1itoria virginiana L.)
and Clitoria minor scandens, foliis subvillosus obiongo-ovatis,
floribus geminatis (=C I itoria ternatea). His Table 32 bore the title
"Clitoria galactia"; however, the plant was described under the genus
Galactia in the text, with Sloane's Phaseolus minor 1actescens cited
in synonymy. From the illustration and description, this plant is not
a Clitoria.
1758: Linnaeus in his tenth edition of "Systema Naturae" cited
two additional species in addition to those established in 1753.
Neither were numbered as the other species were numbered. Clitoria
lactescens disappeared from Clitoria literature after this date, except
for Richter (1840), who cited the species in synonymy for Clitoria
galactia L. The second species, Clitoria zoophthalmum, also disappeared
from the literature until cited by Richter as synonymous with the
non-Clitorian species, Dolichos urens L.
1762: Gronovius published the same two polynominals in the second
edition of his Virginia Flora as he did in 1739. A third non-Clitorian
species published in 1 739 (Clitoria foliis pinnatis, caule decumbente)
was now transferred to the genus Cracca.
1763: Linnaeus described a new species, Clitoria galactia, based
upon Sloane's specimen cited by Browne (1756). No reference was made
to Clitoria lactescens.

49
1766: Crantz followed Linnaeus in his treatment of Clitoria.
He listed tne five species in Linnaeus "Species Plantarum" (1763), but
altered the spelling on two species, published as C1 ternatensium and
C. galacte_a.
1775: Aublet described a new species, Crotalaria guianensis, from
French Guiana. He was the first to use bracteole and stipule
characteristics, and the first to note the prominent costa on the fruit,
a characteristic that was important in the segregation of genera and
section's by later botanists. Aublet had collected several plants from
French Guiana, which included two species similar in gross appearance
that seemingly matched the illustration provided in his Plate 305.
Thus, from the misidentification, the name Crotalaria guianensis Aubl.
would become associated with two distinct :;pecies (i.e. Cl itoria
guianensis vAubl.) Benth. and Clitoria 1 anrifolia Poir.) and provide
confusion in later literature and specimen identification.
1775: Buchoz (*) followed the Linnaean treatment and provided
the first detailed generic description of Clitoria, which included
perianth, stamens, pistil, legume, and seeds. Each species was
described and notes on origin, culture, and economic properties were
included.
(' o
1 775: Forskal (or Forsskal) described a new species, Lathyrus
spectabi 1is (-Clitoria terna tea L.) from Egypt. He was the first to
indicate the villous style typical of the genus.
1786: Lamarck (*) described two new species of Clitoria, C.
heterophyl1 a and C. fa]cata, along with tlie basic five Linnaean species.
He also described Crotalaria 1ongifolia citing Aublet's Crotalaria
guianensis in synonymy. Thus, Lamarck's name is superfluous for
Aublet's name.

50
1786: Scopoli described Clitoria micrantha and segregated Royen's
C. brasiliana from the species and called it Clitoria quadralupensis.
Both species plus C. galactia L. were described with detailed
descriptions and illustrated.
1 788: Swartz described ?C1itoria multi flora with a two-line
Latin diagnosis based upon a plant from Santo Domingo.
1 788: Halter's Carolina flora recognized two species, C. mariana
L. and C_L vircjinjana L., segregated on leaflet number and calyx type.
1789: Browne republished his treatment of Cljtoria^ originally
pub!ished in 1756, still not adopting the Linnaean treatment. His
Table 32, Figure 2 continued to bear the totle "Clitoria galactia"
while described under the genus Galactia.
1791: Gaertner, in his publication on fruits and seeds of various
plants, described and i 11ustrated those of C. ternatea L.
1 792: Richard described Clitori a cap i tata with characteristies
that are too vague for placement of the species without the specimen
in hand. No type was given.
1796: Salisbury published Clitoria spectabilis which was based
upon Lathyrus spectabilis Forssk. and C_. ternatea L. Salisbury's name
was superfluous for the Linnaean species.
1797: Raeuschel published Clitoria aqui1upensis, a nomen nudum,
in his list of Clitoria species.
1798: Roth described Cljtoria amoena with a very detailed Latin
description (clearly marking it as a Centrosema member). It was
published in Roemer's "Archiv fur die Botariisk" and republished by
Roth in his "Catalecta Botánica" in 1800.
1802: Cavanielles described a new species, Clitoria speciosa.

51
1804: Desfontaine in his "Tableau dt l'ecole de Botanique" listed
Cl i tor i a glabella and Clitoria gal actia g 1 .< bel 1 a as Cljitoria_ glabre.
Both C. gjahella Desf. and C. glabre Desf. were without any description
or discussion and thus, nomen nudum.
1806: Salisbury described and illustrated Cl itoria calcarigera.
Three varieties were described based upon leaflet shape. Variety alpha
was the typical representative. Variety beta was based upon Dillenius
(1 732). Variety gamma was based upon Cl i torga virginiana L. Thus
Salisbury's name was superfluous for the Linnaean name, C. virginiana.
1807: Persoon included ten species of CJitoria in his treatment
of the genus. Each species had a Latin diagnosis with habitat. Two
species from Domingo were newly described. Clitoria plumieri Turp. was
based upon a specimen sent to him by Turpin bearing a Latin diagnosis
under the name. The species was named fot Plumier who described the
species in 1693. The second new species was Clitoria rubiginosa Juss.
ex Pers. often cited as C. rubiginosa Pers.
1811: Poiret in a supplement to Lamarck's "Encyclopedie
Method i que Botanique" added seven species to those listed by Lamarck,
the last one, C^ capitata Rich. , lacking a number. Newly described
were Clitoria polyphylla and Clitoria laurifolia from Porto Rico and
Cl i tori a bructeata of unknown origin. The other three species included
the new pair described in Persoon plus Swartz' species. Cl itori a
amoena Roth was synonymized with C^ brasjj iana. Cl itoria gal actia was
reported as the type for a new genus Galac tia Brown which was adopted
by Michaux.
1812: Curtis illustrated C. ternatea in color in his "Botanical
Magazine" in plate 1542.

52
1813: Balbis listed three species ot Clitoria in his "Catalogus
Stirpium." Clitorja broussonetii, named for the French botanist who
sent the specimen to Balbis under a name toy coerulei, was newly
described in a footnote. The footnote contained a Latin diagnosis and
placement of the species near galactia.
1813: Desvaux (*) described the genus Neurocarpum and indicated
that it included Aublet's Crotalaria gajanensis ( published as
C. guianensis by Aublet) and a new species, y el 1 ipticurn, which lacked
a description and is a nomen nudum.
1814: Desvaux (*) described Neurocarpum ellipticum and included
Crotalaria guyanensis Aubl. (Aublet published as: guianensis) and
Crotalaria 1 ongjjfo 1 ia_ Lam. as synonyms of Neurocarpum janensis Desv.
1814: Roxburgh listed five species of Clitoria cultivated in
Calcutta. Cljtoria erecta, a new species, lacked a description and
thus was a nomen nudum.
1814: Brown described a new species, Clitoria arborescens, in
Aiton's "Hortus Kewensis," based upon a cultivar from Trinidad. This
species is usually cited as C. arborescens Ait.
1817: Eaton (*) substituted the name Vexi1laria for Clitoria
because of the criticism regarding the alleged resemblance of the
Clitori a flower to the female reproductive anatomy. Eaton's
superfluous name was not adopted by botanists.
1818: Edwards illustrated and described in detail Clitoria
piumieri Turp., with notes presented on the history of the species.
1818: Rafinesque (*) established the genus Vex i 11 aria Raf. by
his published note "Cl itoria mariana must form a particular genus

53
Vexi11 aria." Neither the generic name not- V. mariana (L.) Raf. were
adopted by botanists.
1820: Curtis published a color plate and notes on Clitoria
heterophylla Lam. (Plate 2111).
1821: Schrader described Clitoria coccinea, and noted it near,
yet distinct from Cl itorja fa]_cata_ Lam.
1821: Mees included a note that Clitoria coccinea Schrad. was
near his species, but uncertain where Clitoria faj_cata Lam. belonged.
This was the source of Clitoria falcata Mees, a name associated with
C. coccinea Schrad. in synonymy in later literature, but the name was
not validly published. Hence, C. falcata Mees is an illegitimate name.
1821: Trattinnick cited fourteen species in his discussion of
the genus Clitoria, and described and illustrated two cultivars,
C_. ter n a tea and C. bras i J i an a.
1821: Leandro do Sacramento (*) described a new genus Martia
based upon a cleistogamous specimen from brazil. The genus was named
in honor of the German botanist Carl Martius. One species, Martia
physodes, was placed in the genus. Martia is a homonym of Martia
Spreng. (1818).
1822: Schultes (*) published Leandro do Sacramento's species
under the name Martiusia physalodes. The change in generic name is
presumed to be due to tiie recognition of Martia as a homonym. The
change in the spelling of "physodes" is unclear.
1824: Kunth described in detail and illustrated two new species
of Neurocarpum, N_. simp 1 icifol ium (plate LIX) and Ik_ angustifol ium
(plate LX). Both species wer-e from Brazil.

54
1824: Humboldt, Bonpland and Kunth (k) described four species,
two of which were published three months earlier by Kunth. The other
two were newly described in detail. They were N. javitense from
Javitam (=Yavita, Amazonas, Venezuela) and N. macrophyllum from
Novo-Granata ("Columbia). A new genus Rhombcflobium Rich, based upon
Richard's manuscript was cited in synonymy. The genus Ternatea Tourn.
was recognized as separate and included T. vulgaris, a new name for the
Linnaean C. ternatea. Under a third genus, Clitoria, two new species,
C. angu'stifol ia and CJ formo sa, were described from Brazil.
1824: Mees and Martins described two new species from Brazil,
Cl itoria vieioides and Clitoria angustifolia. A Latin diagnosis was
given for Clitoria tomentosa which was observed as differing from
C. rubiginosa Pers.
1824: Schrank described Clitoria gladiata, a new species from
Brazi 1.
Post-Linnaean (1825-1858: Revisionary Period)
1825: Hamilton miscited the genus Neurocarpum Desv. as
"Neurocargo 11." He split Desvaux's species (1814) such that
el 1ipticum was based upon a specimen in Desvaux Herb, and N_. guianensis
included Crotalaria guianensis Aubl. and l rota 1 aria 1ongifolia Lam.
He noted the relationship to Clitoria 1aurifolia Poir. and C.
rubiginosa Pers., both bearing the prominent lateral nerve on the
legume.
1825: Velloso described and illustrated four new species,
Cl itoria flumensis (plate 128), Clitoria brasiliana (plate 129),
Cl itoria gemina (plate 130) and Clitoria insulana (plate 131).

55
1825: A. P. de Candolle (*) published in mid November the first
major revision of the genus Cl i tori a in his "Prodromous." The genus
was divided into four sections. Section I (Ternatea Kunth) included
two species. Section II (Euc1itori a ) included six species, C.
glycinoides and C. poitaei newly described. Section III (Centrosema)
included three species. Section IV (Glycinopsis) included only the
newly described C^ berteriana. In discussion of C^ glycinoides
de Candolle referred to Poiteau's Pilanthus. This was an earlier
record of Pilanthus Poit. ex Endl. (Genera Plantarum 1289, 1841). The
genera Martiusia (one species) and Neurocarpum (seven species) were
treated as separate genera. This publication is the source for
N. falcatum DC. which was an error in citation by later botanists.
Lamarck's species was transferred by de Candolle from Clitoria and
should have been cited as N._ falcatum (Lam.) DC.
1826: A. P. de Candolle (*) published his "Memoires sur la
Famille des Legumineuses" nearly simultaneously (early February) with
his "Prodromous," in which he explained his reasons for the classifi¬
cation used in his "Prodromous." This was the source cited for the
origin of C^ berteriana; however, the "Prodromous" predated this work
by almost three months.
1826: Desvaux (*) proposed a new group within Neurocarpum that
would bear Poiteau's name Pi 1anthum. This section included two
transfers cited as N^_ glycinoides N. (C. glycinoides DC.) and
N. rubiginosum Desv. (C. rubiginosa Pers.) plus newly described
N. villosum N. Two species were placed near Desvaux' species described
in 1813, ?L_ laurifol ium Desv. (C_;_ laurifol ia Poir.) and N^ barbatum N. ,
newly described. Two species of Clitoria were newly described,
C. laurifolia Nees, a homonym of Poiret (1811), and (T_ sinuata Nees.

56
1826: Descourtilz (*) substituted the superfluous name Nauchea
for Clitoria and listed thirteen species, one newly described as
N. púdica.
1827: Tussax described and illustrated Turpin's C. plumieri from
the West Indies.
1828: Wallich listed four species in his catalogue. Three were
new species which lacked descriptions, but had specimens cited for
each. They included Clitoria macrophylla Wall. (no. 5345), Cl itoria
acuminata Grah. (no. 5346) and Cl itoria pilosula Wall, (no. 5347 ), each
a nonien nudum.
1829: Descourtilz described the genus Nauchea, and illustrated
and described three species, N. púdica, N. virginiana and NL_ rubiginosa,
each with medicinal notes.
1830: Sweet listed twelve species in his catalogue with Clitoria
occidentalis listed as a substitute for the non-Linnaean "virginiana
B.R. 1047." The species is not described and the citation is unclear.
1830: Guillenim and Perrottel in their "Mora Senegambiae
Tentamen" describe a new species, Clitoria micrantha, a homonym of
Scopoli (1786).
1832: Curtis illustrated M arborescens. This illustration was
a composite of two species, M arborescens and C^ javiten sis which
resulted from additions made to an unpublished drawing of 1822.
1832: Presl described and illustrated Nenrocarpum cajanifolium.
1832: Geel described and illustrated the "PIumier Clitoria."
1832: Zuccarini described Martia mexicana in detail.
1832: Don (*) adopted de Candolle's treatment of Clitoria,
Martiusia and Neurocarpum. He described Clitoi ia racemosa and Clitoria

57
alba and placed them in Sect. Euclitoria. Foui of the five dubious
species listed by de Candolle were placed. Clitoria broussonetii was
transferred to the genus Cologania. Clitoria phyrne Juss., C. mariana
Hoc. and Sesse ined., C. angustifol ia Nees and Mart, and rubiginosa
Nees and Mart, were synonymized with Galactia species.
1835: Pritzel listed Nauchea clypeata Desc. based upon "Desc.
Ant. 8,591," a publication not obtained by this author.
1837: Hooker described and illustrated Clitoria virdiflora Bouton
ex Hook'.
1837: Bojer described Clitoria 1asciva and reported Clitoria
virdiflora Bouton mss. Both Hooker and Bojer cite "Bouton mss. in hb.
nostr." Stafleu's "Taxonomic Literature" (no. 537) indicated Hooker
published sometime within October 7 to November 8. The date of Bojer's
publication was unknown except for the year, although if it preceded
Hooker, then C. virdiflora Boj. was a nomen nudum. Bojer made no
reference to Hooker.
1837: Schlechtendal reported Clitoria mariana L. from Mexico.
This was the source of Clitoria mariana Schlecht. cited by other
botanists. Schlechtendal had misidentified his specimens, which are
Clitoria mexicana Link.
1837: Bentham (*) published a revision of de Candolle's treatment
of Clitoria , Neurocarpum and Martiusia. Two sections of Clitoria were
eliminated. Section Centrosema was elevated to a genus level. Section
Glycinops is was eliminated through the transfer of its only species to
another genus. To the remaining two de Candolle sections, Bentham
added the section Bractearia. Martiusia and Marti a were synonymized
with Neurocarpum. Bentham noted the close relationship of Clitoria and

58
Neurocarpum segregated only by the prominant costa on the legume. New
species included Clitoria amazonum Mart, ex Benth., Clitoria acuminata
Benth., Clitori a arbórea Benth., Clitoria racemosa Benth., Clitoria
pedunculata Boj. ex Benth., Neurocarpum longifolium Mart, ex Benth.,
Neurocarpum frigidulum Mart, ex Benth., Neurocarpum rufescens Benth.,
Neurocarpum densiflorum Benth., and Neurocarpum brae tea turn Mart, ex
Benth.
1838: Sc hiechtendal described Clitoria schiedeana.
1838: Lind ley reported that C^ terna tea roots were emetic.
1839: Bentham's 1837 article was republished with the title
revised.
1839: Bentham reported on Schomburgk's British Guiana collections
and substituted the name Dendrocyamus for his section Bractearia. He
reasoned that the name Bractearia was preoccupied by a genus in the
Rubiaceae and in a section of Chaetogastra, ami thus inappropriate for
CJJ toria.
1839: Bentham described Clitoria polystachya in his enumeration
of Hartweg collections.
1840: Bentham described Neurocarpum flagellare and synonymized
N. frigidulum with N. 1ongifolium in Hooker's "Journal of Botany."
1840: Steudel published a list of Clitoria species and species
transferred. The following binomináis, each a nomen nudum, appeared for
the first time: C^ bra si 1 iana Arrab. , C. graharni Steud. (C^ acuini nata
Grah.), insulana Arrab., C. micrantha Smith, C^. phryne Commers. and
C. púdica Steud. (Nauchea púdica Desc.). The latter name was a
transferred name and legitimate. Clitoria grahann was an apparent

59
substitute name for the homonym C. acuminata Grah. which to this date
was still undescribed.
1841: Steudel's second volume published two new binominals for
Neurocarpum, each a nomen nudum. They included Marti a brasiliensis
Zuccar. (N_. el 1 ipticuin Desv.) and N. mexicanum Steud. (Martia mexicana
Zuccar.), both transferred to Neurocarpum, the latter described
previously, the former undescribed but placed in synonymy.
1841: Hooker and Arnott described Neurocarpum mu!tiflorurn in their
enumeration of the plants collected on the Beechey1s voyage.
1841: Brunbury described Neurocarpum resupinatum from Brazil.
1842: Hasskarl described two new species from Java, Clitoria
obionga (C. virqiniana L. var el 1 iptica DC.) and Neurocarpus retusus.
1842: Bertoloni described and illustrated Clitoria alabamensis.
1843: Martius and Galeotti described Clitoria multiflora and
Clitoria grandiflora from Mexico.
1844: Bentham described Cl itoria brachystegia from Ecuador in his
enumeration of plants collected on the voyage of the Sulphur.
1844: Hasskarl published Neurocarpus, an orthographic variant of
Neurocarpum, and Neurocarpus retusus.
1844: Paxton illustrated in color and described Clitoria ful gens
(Centrosema coccinea).
1845: Voigt transferred Clitoria erecta Roxb. to Neurocarpum. He
listed six species of Clitoria cultivated in Calcutta.
1845: Blanco reported one species, C. ternatea, from the
Phi 1ippines.
1845: Tenore described Clitoria tristis.

60
1847: Burnett illustrated and described a new variety, C.
ter ñatea v a r. major, based upon a cultivar raised from seeds sent from
New South Wales. He also illustrated and discussed Paxton's species.
1848: Hasskarl provided a detailed description of his species,
Neurocarpus retusus, published in 1844.
1848: Schomburgk published three new species and a new genus
based upon a manuscript of Klotzsch. Each was a nomen nudum. The
genus was Macrotrul1 ion Kl. with two species, M. spendens K1. and
M. elegans Kl. The third species was Neurocarpum speciosum Kl.
1850: Dalzell described CJjtoria biflora, a new species, from
India.
1851: Miguel described Clitoria amoena, a new species from the
lower Marrowyne River.
1852: Bentham described two species in a footnote that were
published previously as a nomen nudum. They were Cljtoria macrophyl1 a
Wall, (published 1828) and Clitoria grahami Steud. (published in 1852;
synonym C. acuminata Wall, published in 1828.)
1853: Duchass and Walpers describe Neurocarpum argentum, a new
species from Panama.
1854: Beurling published Clitoria portobc11ensis, a new species
from Panama.
1856: Miguel described a new species from Java, Cl itoria java nica.
1858: Bentham (*) published his second revision of the genus
Clitoria which included Terna tea, Neurocarpum, Marti a and Mart i usia in
synonymy. The major change was the synonymizing of the latter three
genera with Clitoria. Bentham noted that ihe character of the legume
costa broke down. Another change was the name CJitoriathus used for

61
section three instead of Bentham's earlier names Brae tearia (1837) or
Pendrocyamus (1839). New species described were C. f1agellaris,
rufescens, nana, selloi, leptostachya; C_. hoffmanseggi i
(superfluous name for C. arbórea) and C. stipularis. The species
transferred from Neurocarpum included simplicifolia, C. quianensis
and densi flora.
Post-Linnaean (1859-1932): Floristic Period)
1862: Bentham in Martius and Eichler's "Flora Brasi1iensis"
listed 'fifteen species with descriptions, synonymy, and a key to the
species. Tii i s was the first record of a i ey to Cl i tor ia species and
used characters of: habit; leaflet shape, pubescence and number;
legume costa; peduncle length and number of flowers borne; and
bracteole shape and length relative to calyx or pedicel. New
binominal s included C. guyariensis (orthographic variant of C. guianensis)
and cajanjfolia (transfer of species from Neurocarpum). A new
variety described was C._ glycinoides var. uurantiaca.
1864: Griesbach reported four species of Clitori a in his flora
of the British West Indies.
1864: Bentham described a new species in his "Flora Australiensis,"
Cl i tori a austra l_i s .
1866: Griesbach described a new species in his "Catalogus
Plantarum Cubensium," Clitoria glomerata.
1871: Oliver reported one species from tropical West Africa! in
his flora.
1875: Micheli described a new species from Brazil, Clitoria
pedunculata. This name was a homonym for Bojei ex Benth. (1837).
Micheli placed his species in Section Clitorianthes.

62
1878: Lowis provided some observations on the Indian C. ternatea.
1878: Hance reported on rare Chinese plants and listed Clitoria
macrophylla Wall, for a specimen in his herbarium collected by Sampson.
This was the source for C_. macrophyl 1 a Hance cited by later botanists.
1878: In a second Philippine flora, Blanco continued to report
one species of Clitoria as he had in 1845.
1878: Vatke described Clitoria zanziharensis, a new species from
East Africa.
1879: Baker (*) reported five species of Clitoria in his flora of
British India. He made a revisionary change that elevated Ternatea
and Neurocarpum to the subgenus level, the first botanist to do so.
Section Clitorianthes was not reported, presumably because the species
within this group did not occur in India. A new variety was reported,
C. ternatea var. pi 1 osula (C. pilosula Wall.).
1882: Sagot reported five species of Clitoria in his French
Guiana flora. A new variety, C. javitensis var. glabra, was described.
1887: Watson described two new species from Mexico. Clitoria
triflora was placed near C. guianensis in the section Neurocarpum.
Clitoria sericea was described from a fruiting specimen. Watson did not
place this species in any section.
1887: Hemsley described Clitoria Hanceana based upon a specimen
in Hance's herbarium collected by Sampson. He noted that this was
equivalent to C. macrophylla Hance (cf 1878), but not C. macrophyl!a
Wall.
1887: Sesse and Mocino described two new species from Mexico.
Clitoria racemosa (homonym of Don, 1832, and Bentham, 1837) was first
reported in synonymy by Don (1832) and probably by de Candolle (1825)

63
as a synonym of CL plumieri based upon a Manuscript of Mocino and Sesse
cited as "fl. Mex. ic. ined. ' The second species, Clitoria uirginiana,
was probably an orthographic error (cf. 1693) because Gronovius'
“Flora Virginiana" was cited (also typographic error with "U" for "V"
and corrected by Sesse and Mocino in the second edition of their
publication.
1889: Defiers reported one species from Yemen.
1893: Sesse and Mocino published their second edition of "Plantae
Novae Hispaniae" and cited C. virginiana. This corrected the
orthographic error of their first edition (cf. 1887).
1891: Kuntze (*) transferred twenty-nine species to the genus
Ternatea because Clitoria was a homonym of Clitorius Dillenius
(1732).
1894: Taubert in Engler and Prantl's "Die Naturlichen Pflanzen
Familien" adopted Bentham's treatment. They indicated about thirty
species in the genus were found in the tropical zone, but their species
were reported by sections and totaled to only twenty-four. They placed
ca. six species in Section Ternatea (Asia and Africa), ca. fifteen
species in Section Neurocarpurn (Asia, Australia, and America), and ca.
three species in Section Cl itorianthes (tropical America).
1894: Nairne reported one species oi CJJtojia in his Western
India flora.
1894: Robinson described the flowers of C. sericea (cf. Watson,
1887) and noted that the style was similar to Centrosema, yet had other
flower characters of Clitoria.
1894: Sesse and Mocino's "Flora Mexicana" followed their 1893
treatment ot Mexican plants.

64
1894: Siebert and Voss reported Clitoria coelestis as a horti¬
cultural synonym of terna tea. This new binomial is a nomen nudum.
1897: Britton and Brown reported one species of Clitoria in their
flora of the northern United States.
1897: Duss reported two species from Martinique that occur
natively plus one cultivated species. A new variety, C^ glycinoides
var. ecostata, was a nomen nudum.
1897: Micheli described a new species from East Africa, Clitoria
tanganicensis.
1899: Rose reported nine species of Clitoria in Mexico with six
additional species as doubtful and excluded. Two new species were
Clitoria humilis and CJitoria subsessi1is. Rose indicated that Clitoria
sericea may be a Cracca and that Clitoria grandiflora Mart. & Gal.
and Clitoria schjedeana suggested Centrosema.
1900: Bailey reported two Australian species of Clitoria in his
Queensland flora.
1901: Huber described a new species from Brazil, Clitoria cearensis
and noted its affinities witli C_. glycinoides, C. guyanensis and
C. cajanifolia.
1903: Merill reported vernacular names of C. ternatea in his
Philippine Dictionary.
1904: Chodat and Hassler described a new variety from Brazil,
Clitoria cajanifolia var. latifolia.
1907: Harms reported the first occurrence of cleistogamy in
Clitoria in a detailed article. Cleistogamous flowers in three species
were described and illustrated. They were C. glycinoides, C.
gujanensis and C. cajanifolia. Harms discussed the history of Marti a

65
physodes, a species described based upon a cleistogamous flowered
specimen from Brazil (cf. 1821).
1908: Mattei described a new species from northern Africa,
Clitoria al bi f 1 ora. He noted its affinities to C_. ternatea and
tanganicensis.
1909: Bailey listed the same two species of Clitoria in his
Queensland catalogue that he reported in his flora (1900).
1909: Gibbs reported one species of Clitoria in his Fiji flora.
1909: Herzog described a new species from Bolivia, Clitoria
nervosa. He noted its affinities to C_^ hoffmanseggii and C. racemosa
Benth.
1909: Huber described a new species from Brazil, Clitoria
obidensis.
1910: Hassler described a new variety of Clitoria from Paraguay,
C. dens iflora Benth. var. mucronata.
1911: Dunn added C. mariana to the list of species found in
China.
1911: Hayata reported one species of Cl itoria from Formosa.
1912: Merrill reported one species of Clitoria from Manila in
his flora.
1912: Compton (R) described the seedling morphology and vascular
structure of C. teniajtea. This was the first non-floristic, non-
nomenclatura!, non-horticultural research undertaken on the genus
Clitoria.
1913: Harms described a new species of Clitoria from the Congo,
Cl itoria kaessneri.

66
1914: Baevecke reported one species of Clitoria from the Eastern
United States in his flora.
1915: Gagnepain described a new species from Laos, Clitoria
1inearis, and placed it near hanceana.
1916: LeCompte reported six species of Clitoria from Indo-China.
A key, descriptions, distributions, and citations were included.
Cl i tori a 1 i nearis was illustrated. A new variety was described,
C^ hanceana Hems 1 . var. laureol_a.
1917: Ewart reported one species of Clitoria in his Australian
flora on the Northern Territory.
1918: Britton reported one species of Clitoria from Bermuda in
his flora.
1918: Kirtikar and Basu reported the Indian vernacular names and
medicinal properties of C. terna tea.
1918: Pittier described and illustrated the keel and wings of a
new Venezuelan species, Clitoria dendrina.
1919: Brandegee described a new species from Baja California,
Cl itoria monticol_a, and segregated it from C. mariana.
1920: Britton reported one species of Clitoria from the Bahamas
in his flora.
1 920: Fawcett reported two species of Cl itoria from Jamaica in
his flora. They were segregated by leaflet number and flower color.
1921: Harms described a new species from Brazil, Clitoria
brachycalyx and reported its affinities with C. brachystegia.
1922: Ducke described a new species from Brazil, Clitoria
snethlageae, and noted its affinities with C^ 1eptostachya.

67
1 922: Ridley reported two species or Clitori a from the Malayan
Peninsula in his flora and noted that C. cajanifo1ia was introduced
into Indonesia on Brazilian cattle which had the viscid seeds in their
hair.
1922: Standley reported six species of Clitorian trees and shrubs
from Mexico in his flora. A key and brief diagnosis were included.
1923: Merrill reported one species of Clitoria from the
Philippines in his flora.
1924: Britton reported three species from Puerto Rico and the
Virgin Islands in the survey conducted by the New York Academy of
Science. One species was reported under the genus Clitoria. Two
species from the Neurocarpum section were reported under the genus
Martiusia. Even through the original species N. el 1ipticum Desv. was
listed in synonymy, Britton did not use the older generic name
Neurocarpum. The two transferred species were Martiusia rubiginosa
(Juss. ) Britton and Martiusia laurifolia (Poir.) Britton. This was the
first example of a botanist diverting from Bentham's treatment of 1858.
This trend of recognizing two genera, Clitoria and Martiusia, was
followed by a few American botanists.
1925: Fragoso and Ciferri (R) described several ascomycete
species from branches and legumes of "siccis Clitoriae triandrae"
collected in the Dominican Republic. This was the source for two
names, Clitoria triandra and Clitoria riandra.
1925: De Wild described a new species from British East Africa,
Cl itoria mearnsi.
1926: Small described a new species from Florida, Clitoria
fragrans and noted its affinities with C. mariana.

68
1 928: Standiey reported four species of Cl itoria from the Panama
Canal Zone in his flora. A key and habitats were included.
1929: Baker reported four species of Clitoria in his treatment
of Tropical Africa legumes. A key and diagnosis were included. He
noted that C^ al biflora and C_^_ mearnsi were equivalent to C^ ternatea.
A new variety that had been nom. in sched. was described as Clitoria
ternatea var. angustifolia Hochst. ex Bak. Baker published Clitoria
tanganyicensis Micheli as an orthographic correction for C^
tanganicensis Micheli (1897).
1929: Hutchinson & Dalziel transferred (C racemosa Don to the
genus Vigna.
1929: Standley transferred C. sericoa Wats, to the genus
Tephrosia.
1930: Barker and Darbeau reported four species of Clitoria from
Haiti which included C. piunrieri Tuss., a nomen nudum.
1 930: Ducke described a new species from Brazil, Cl itoria
grandji folia, and noted its affinities to C. arborescens.
1930: Macbride described a new species from Peru, Clitoria
pozuzoensis. He noted that some of the characters that Bentham used to
distinguish the species of Section Clitorianthes broke down since
Macbride's new species keyed to C_^ javitensis from which it was totally
different.
1931: Gleason reported a species of Clitoria collected on the
Tyler-Duida expedition to Colombia.
1931: Malme described a new species from Brazil, Cl itoria
chapadensis and noted its affinities with C. guyanensis. He also
described a new variety, Clitorja glycinoides var. megapotamica.

69
1931: Sandwith reported his observations on the type of
nomenclatural problems of C. javitensis.
1931: Standley reported two species of Clitoria from the
Lancetilla Valley of Honduras.
1931: Williams reported four species of Clitoria_ from Trinidad
and Tobago. A key is included.
Post-Linnaean (1 932-present: Research Period)
1932: Tschechow and Kartaschowa (R) reported the first chromosome
count for the genus with 2n = 16 for ternatea. The chromosomes were
illustrated in a camera lucida drawing. Relationships of C^ ternatea
with other species in tne Phaseoleae tribe were discussed.
1932: Rydberg reported one species of Clitoria in his flora of
the Central North American plains and prairies.
1933: Small reported one species of Clitoria and two species of
Martiusia in the Southeastern United States. Ill's floristic study
adopted Britton's treatment (1924). Small's species (1926) was
transferred as Martiusid fragrans Small.
1933: Holland and Joachin (R) reported their results on soil
erosion experiments with C. cajanifolia in Ceylon (=Sri Lanka).
1935: Burkill reported the economic uses of two species of
Clitoria of the Malayan Peninsula.
1936: Brown reported one species of Clitorija in his flora of
Southeastern Polynesia.
1936: Standley reported two species of Clitoria from British
Honduras. Neither a key nor descriptions were included.
1936: Stahl included three species of Clitoria from Puerto Rico.
1937: Standley reported four species of Clitoria from Costa Rica.

70
1937: Stehle repotted on the colonization of some plants in
Martinique and included Cl itoria rubiginosa Juss. var. ecostata (Urb.)
Stehle, a noinen nudum.
1939: Amshoff reported eight species of Clitoria in Pulie's
Suriname flora. A key was included. Amshoff adopted Bentham's
classification (1858).
1940: Standiey described a new species from Panama, Clitoria
velutina.
19-41: Burkart described and illustrated a new species from
Argentina, Clitoria cordobensis. He discussed cleistogamous flowers
in the species and its affinities with C^ mariana, C. guyanensis,
C_. multif 1 ora and C. rubiginosa.
1943: Kok, Machado and Meirelles (R) reported their results of
utilizing C ternatea hay for sheep digestion.
1 943: Mai.bride reported eight species of Clitoria from Peru. A
key, descriptions, and citations were included. However, this treatment
was of poor quality because of the numerous misidentified specimens that
Macbride examined and the difficulty within the group Clitorianthes.
Macbride reported variability in several species, and along with his
additional notes, indicated his awareness of the problem. Some species
reported do not occur in Peru. Other species not reported do occur in
Peru. Peru has several endemics not recognized by Macbride.
1 943: Mocoso reported four species of Cl itoria from the Dominican
Republic in his flora. Clitoria polyphylla was included in synonymy
with a Bjirbieria species.
1943: Pittier described a new species from Venezuela, Clitoria
glaberrima. He noted its affinities with C. arborescens and C. dendrina.

71
The description included "Pedicel 1i 4-5 cm1 which was a typographical
error for "mm."
1944: Pittier published a second description of C. dendrina
(cf. 1913) and illustrated the perianth.
1945: Hassell (R) reported his results of experiments to
establish C. ternatea in pastures of Queensland, Australia.
1946: Standley reported five species of Clitoria_ in his flora of
Guatemala. A key was included.
1946: Strang reported the ornamental properties of C_. raceinosa
Benth. in Brazil.
1948: Robyns reported one species ot Cl i tori a in his flora of
Albert Natiunal Park in Africa.
1948: Stehle and Quentin reported two species of Clitoria from
Guadeloupe in their flora. They included C. rubiginosa var. genuina
(illegitimate name for the typical variety) and rubiginosa var.
ecostata (Urb.) Stehle, a nomen nudum (cf. 1937).
1949: Burkart described and illustrated a new species from
Argentina, c1 itoria epetiPlata. He noted its affinities and compared
the species to C_. guianensis. The variety C. epetiolata var.
latiuscula was described.
1949: Chopra, Nath, Badhwar, and Gho^h reported the economic uses
and problems of C. ternatea in their "Indigenous Drugs of India."
1 949: Henderson reported C^ laurifol ia as an established Malayan
wildflower introduced from Brazil early in the nineteenth century.
1949: Johnston reported C L portobel1onsis from San Jose Island,
Gulf of Panama, in his flora and discussed its confusion with Clitoria
arborescens, the name the Panama plants were given by botanists.

72
1 949: Merrill in íiis "Index Rafinesque" i eported that C. parviflora
Raf. was C. terna tea.
1950: Metcalfe and Chalke (R) included the genus Cl itoria in their
descriptive anatomy of the Papi1ionaceae. Leaf, axis and wood
characteristics were included. The species studied were not reported.
1950: Velez and Overbech included C. ternatea as a desirable
cultivar in their book.
1951: Krapovickas and Krapovickas (R) reported a chromosome count
of 2n=24 for C. cordoboensis and discussed its relationship to C.
ternatea.
1951: Leon reported three species of CJjtoria from Cuba in his
flora. Cl i loria glomeiata was included in synonymy with a Galactia
species.
1951: Quisumbing reported the medicinal properties of C^ ternatea
in the Philippines.
1951: Ran (R) described endosperm development in C. ternatea.
1952: Andrews reported one species of Cjjtoria from the Sudan in
his flora.
1 952: Burkart reported five species of Clitori a from Argentina.
A key was included.
1952: Lemee reported six species of Clitoria in his French Guiana
flora.
1 952: Schery described a new species, Clitoria coriácea, from
Brazil.
1953: Frahm-Leliveld (R) reported a chromosome count of 2n=l6 for
C. ternatea.

73
1954: Berhaut reported two species of Clitoria from Senegal in
his flora. Within his key he published C. ternatea var. alba.
1954: Wilcez reported three species of Clitoria in his Belgium
Congo flora. A key and descriptions were included. He recognized
Baker's variety (1939) of C. ternatea and included C. tanganicensis
and C^ iiiearnsj as synonyms.
1955: Sirdeshmukh (R) reported a double flower occurring within
C_._ terna tea.
1956: Bunting and Lea (R) reported their observations on the
introduction of three strains of C^_ ternatea in Sudan pastures.
1956: Fidalgo (R) reported his anatomical study of the Brazilian
tree ornamental, C. racemosa Benth. (=C^ rairchiIdiana). This was the
first and only anatomical study of a Clitoria species, except for the
survey of Metcalfe and Chalk (1950). Twenty plates of illustrated
anatomical structures were presented.
1956: Gardner and Bennett included C. teiñatea as a poisonous
plant in their book "Toxic Plants of Western Australia" and reported
the curariform properties of the seeds of C. atborescens.
1957: Cowan described a new species from Venezuela, Clitoria
cerifera, and noted its affinities to C^ a -borescens. The species was
illustrated in his Figure 65.
1 958: Chopra, Chopra, Honda, and Kapur included C_^_ ternatea in
their book 1 Indigenous Drugs of India" and reported the species' use and
effects as: a cathartic, a remedy for snake bites and scorpion sting,
an anti-dysentaric drug, a diuretic drug, a laxative, a drug for gonor,
and a skin jelly for eczema, prurigo and impetigo.

74
1 958:
Santapau reported one species of Clitoria in his flora of
Purandhar,
India.
1 959:
Rizzini described a new specif, from Paraguay, Clitoria
burkartii.
1 959:
Tiwari and Gupta (R) reported the first chemical study of a
Clitori a species and reported a new chemical "aparajitin" extracted
from the leaves of terna Lea .
1960: Sinha (R) identified a sterol from the yellow oil of the
seeds o'f C. terna tea. In a second article, Sinha (R) reported a
chemical and its properties extracted from the leaves of C. mariana.
Although unnamed by Sinha, the author of the present study found the
properties 10 agree with those of "aparajitin."
1961: Hundley reported two species of Clitori a from Burma and
their vernacular names.
1961 :
Sen (R) reported and illustrated binucleate pollen mother
cells in C.
ternatea.
1961 :
Sen and Krishnan (R) reported their genetic experimental
results with the double flowers of C^ ternatea and concluded that the
double flowers originated through a dominant gene mutation.
1961: Saroja (R) reported a chromosome count of n=8 for the
double flowered C_. ternatea. He also reported his results on the
fertility of its pollen grains with those grains of the papilionaceous
flower type. Saroja concluded that the mutant gene was pleiotropic.
1962:
Lind and Tallantire reported one Clitoria species from
Uganda.
1 962:
White reported one species of Clitoria from Northern
Rhodesia (^Zambia).

75
1963: Backer and Bakhuizen von Briril reported four species of
Cl itoria from Java within a detailed descriptive key. They synonymized
C. javanica with rubiginosa.
1963: Maheshwari reported one species of Clitoria from Delhi.
1963. Rizzini reported twenty-three species of Clitoria from
Brazil. His treatment included keys, descriptions, morphological notes,
extensive citations, and illustrations of leaflets and some germination
diagrams, lie reported that for the few species studied, seed
germination type was an excellent sectional character. He found that
Sect. Terna tea and Sect. Cl i tor ianthes v/as of the epigean type and that
Sect. Neurocarpum was of the hypogean type Newly described were
Clitoria rubiginosa f. longjfolia, C. rubiginosa f. latifolia, C.
guyanensis r. chapadensis (Muline) Rizz., C. amazonum f. rotundifolia
and C. racemosa f. obovatifoI ia_. Despite i few nomenclatura 1 errors,
Rizzini ‘s floristic treatment was the best of any seen by the author of
the present study.
1965: Angley reported three species of Clitoria from Parana,
Brazil without naming the species.
1965: Gooding, Loveless,and Proctor reported one species of
Cl itoria from Barbados.
1 965: Liogier transferred two species from Mart i usia in Britton
and Wilson's flora (1 924) to Clitoria, which corresponded to Bentham's
treatment.
1966: Chavan reported two species of Clitoria from Gujarat, India.
1966: Hutchinson and Dalziel reported two species of Clitoria
from West Tropical Africa.

76
1966: Lourtig discussed the specimens within the herbarium of
Johan Burman and reported th.it Flos_ Cl_ito> ius flore coeruleo Burm was
Clitoria ternatea (cf. 1737).
1966: Padmanabhan (R) reported the first tissue culture studies
on a Clitorian species. His results on the excised embryonal axes of
ternatea were discussed.
1967: Boulos reported one species of Clitoria in his weed flora
of Aswan, Egypt.
1967: O'Arcy reported one species of Cl itoria from the Virgin
Islands.
1967: Howard reported his observations on the woody Clitoria
species and substituted the name Clitoria fa ire hi 1 diana for the homonym
CM raceinosa Benth.
1 967: Mu IIic k, Prakash and Chatterji (R) reported their results
on seed germination of C. ternatea under various ecological and
physiological conditions.
1 967: Mu I lick and Chatterji reported their results of the effects
of sodium cyanide on seed germination of C. ternatea.
1967: Mitra and Datta reported a chromosome count of n=8 for
C. ternatea.
1968: Baum discussed binominals used by Linnaeus in his "Systema
Naturae, ed. ID" and reported C. 1 actescens was galactia and that
C. zoophthaI mum was Dolichos urens.
1968: Gupta and lal (R) reported their isolation of a hexacosanol ,
a sitosterol and an anthoxanthin glucoside from the seeds of C. ternatea.
1968: Lakshmann and Padmanabhan (R) reported their results of the
antibiotic iscuchitine on the growth of C. ternatea in vitro.

77
1968: Radford reported one species of Clitoriep from the Carolinas.
1968: Uphof reported the economic importance of three Clitoria
species in his dictionary.
1 970: Dat.ta and Saha (R) reported the floral vasculature of
C_. terna tea and concluded that it represented an advance stage in the
Phaseolae.
1970: Katiyar, Ranjhan and Shukla (R) reported their experimental
results on the yield and nutritive of C. ternatea for sheep.
1970: Oakes (R) reported his results on C. ternatea grown for
livestock consumption.
1970: Cornell and Johnston reported one native species and one
cultivated species of Clitoria in Texas.
1970: Vei'dcourt reported one species of Cl itoria from Tropical
East Africa and described a new variety C_. rubiginosa var. glabrescens.
He proposed that the following species of the "rubiginosa-mariana"
complex be reduced to subspecific levels under the two species
C. mariana and C. rubiginosa based upon stipule shape. His complex
included C. mariana L., C. grahami Steud. ex Benth., C. javanica Miq.,
C. australis Benth., C. fragrans Small, C. mexicana Link and
C. rubiginosa Pers.
1971: Cameron reported C^ ternatea reacted favorably to a small
extent in becoming estahi ished where trifluralin was used to control the
black pigweed Trianthema por tul aseastrum.
1971: D'Orey and Liberato reported one species of Cl itoria from
Portugese Guinea.
1971: Gil left, Poll ill and Verdcourt reported three species of
Cl itoria from Tropical l ast Africa. Clitoria ternatea was concluded

78
to be a polymorphic species that included C. zanzibarensis, C. mearnsii,
C. tanganicensis and C. ternatea var. angustifolia, and no variety was
recognized. Refer to tiie African members of C. ternatea in this study.
1972: Adams reported two species of Clitoria from Jamaica.
1 972: Liberato reported one species of Cl itoria from S. Tome and
Principe.
1973: Aiyar, Narayanan, Seshadri and Vydeeswaran (R) reported
three glycosides of kaempferol extracted from the leaves of C_. ternatea.
1974: Croat reported six species of Clitoria in Panama, with the
new combination C. guianensis var. subsess¡lis. Croat correctly
provided the first explanation of the plants known by the name C.
arborescens as represented by two species, a liana C. javitensis and
the tree, C. glaberrima.
1 975: Fantz sent a manuscript on Cl itoria for inclusion within
the Leguminosae in the Flora of Panama series, which was accepted and
approved by the editor-. Six species were reported and described. They
represented the same six species identified by Croat, but with some
changes in the species names to correct nomenclatura! errors to conform
with the Code. Croat's variety was not recognized as his segregating
characters broke down. The publication will appear when the entire
family has been treated.
1976: Fantz presented a preliminary paper comparing diagnostic
characters that segregate Clitoria from Centrosema. This was in
response to some individual suggestions and herbaria proposals that the
two genera should be recombined, a proposal he does not agree with as
the genera are distinct entities.

PRIOR RESEARCH ON CLITORIA
Most of the prior research on Clitoria that is non-floristic and
non-nomenclatura 1 occurred after 1 950. The amount of research on
Clitoria has been relatively small given the size of the genus. Most
of the work has involved the species Clitori a terna tea. This species
is commonly cultivated in gardens and often escapes, becoming
naturalized, giving it the largest distribution (pantropical and
subtropical) in the genus. Because of its easy accessibility,
C. terna tea was usually selected as the representative of the genus
Clitori a when research problems included several legume genera.
Anatomical Research
The few anatomical studies on Clitoria have resulted from surveys
including a number of genera. Only Fidalgo (1956) provided a detailed
anatomical study of a species within the genus.
Metcalfe and Chalk (1 950) included Clitori a as one of the genera
in their descriptive anatomy of the Papi1ionaceae. No reference was
made as to which Clitoria species were used in their study. They often
described an anatomical characteristic followed by a list of genera
that included species with this characteristic. The following
anatomical characteristies were noted for Clitori a: Leaf: Non-glandular,
hooked hairs, with short basal cells, and a larger bent, terminal cell
(p. 505); 1ower surface papillose or subpappilose (p. 506); epidermis
79

80
often including a proportion of mucilaginous cells (p. 506); rod-shaped
crystals (styloids) in the palisade tissue (p. 511). Axis: Young stem
with cork arising in the middle of the cortex (p. 513); tanniniferous
cells, colored brown in dried material, and sometimes containing
proteins, mucilage, and other materials as well, often occurring in
groups situated in the pith and phloem or in the primary cortex
(p. 515). Wood: Parenchyma very abundant, in broad, moderately
regular bands 4 cells or more wide in some species (p. 521); rays
exclusively uniseriate or with only occasional biseriate rays in some
species (p. 624); rays up to 1 mm in height in some species (p. 524);
sheath cells present in some species (p. 525); fibres forming small
islands on the cross-section owing to the abundance of wood and ray
parenchyma (p. 525).
Datta and Saha (1970) included Clitoria ternatea as one
representative member in their study of the floral vasculature of the
tribe Phaseoleae. Their findings include that the common bundles for
perianth, stamen and disc remained undivided for a short distance; that
the disc traces are feebler, branching immediately after origin in the
swellings at the bases of the stamens for a short distance; that the
common bundles for the perianth and stamens are elongated; that the
style is hollow having one dorsal and two marginal traces, with the
upper part solid having a single trace; that the vascular trace
continues almost up to the stigma tip. A camera lucida drawing of a
transection of the floral axis is illustrated in Figure 1 and a graphic
diagram of the floral vasculature is illustrated in Figure 17.
Fidalgo (1956) did an anatomical study of Clitoria fairchildiana
(published under the synonym Cl itoria raceim sa), a tree species

81
cultivated in parts of Brazil as an ornamental, and infrequently
cultivated elsewhere. He provided 43 page... of his techniques and
descriptive anatomical results on the root. , stems, leaves, flowers,
fruits, and seeds. He included 20 plates (¿2 figures) of drawings that
illustrate the anatomical structures. He concluded that there is a
great occurrence of calcium oxalate, monoclinal, principally in the
cortical parenchyma; small particles of starch were found in woody
parenchymous cells, in great abundance in the pith; tannins occur
frequently, not from internal differentiation of cell, but from
secretory pockets or canals; mucilaginous canals commonly not in
parenchymatous pith rays; external epidermis provided with uni seriate
protective hairs (not peduncle) with 1 or 2 basal cells, partially or
totally caducous with developing organ; secretory hairs occur in hidden
parts of peduncle, with a single basal celi and a terminal swelling;
development of 2° structures of stem emerges between layers of apical
collenchyma, strata of suberized cells occupy various positions;
petioles invaginated from bark by activity of cambium and isolated
woody elements forming a secondary internal woody ring of bark;
pulvinus and rachis express an interesting evolution of cortical weave
which permits movement of the leaves; leave present numerous
adaptations to xerophytic conditions, with great cavities in the lower
epidermis, obliterated by long protective hairs over large, scanty
stomata; fruit with two patterns of sclerenchymous layers orientated
in different directions; seed coat formed of numerous weave patterns
in the external epidermis, not containing toxic proteins, presenting
great advantages for their use through cattle alimentary tracts.

82
Chemical Research
Chemical research has been conducted by Indian investigators on
the seeds and leaves of Clitoria species that occur natively in India
and are reported by the Indians to be of medicinal value.
Tiwari and Gupta (1959) reported a new chemical they named
"aparajitin" that was obtained from an alcoholic extract of the dried
leaves of C. ternatea. The chemical name was derived from the plant's
Indian vernacular name of "aparajita." Based on experimental results,
aparajitin (^26H5o02) was determined to be a o-lactone of 2-methyl-4-
hydroxy-n-pentacosanoic acid with the chemical structure:
ch3(ch2)i8*CH2CHCH2CH2CH*CH2C0
0
The properties of aparajitin were described as insoluble in water,
cold alcohol, chloroform, ether, and petroleum ether; sparingly soluble
in acetone; soluble in hot alcohol, benzene, ethyl acetate and
pyridine; did not decolorize bromide water or alkaline potassium
permangate solution and gave no acetyl derivative; did not reduce
Fehling's solution nor respond to any keto group test; laevo-rotatory
with M.W. of 391 and m.p. of 92-93°; treatment with hydroiodic acid
and hydr'obromic acid produced 2-methyl-4-iodo-n-pentacosanoic acid
^26^51^2^ anc* methyl-4-bromo-pentacosanoic acid (C^gHgiC^Br)
respectively; esterification produced methyl acid, 2-methyl-4-hydroxy-
n-pentacosanoic acid, which formed the acetyl derivative of 2-methyl-
4-hydroxymethyl-n-pentacosanoate; oxidation produced arachidic acid
and B-methylglutanic acid.

83
Sinha (1960) identified the yellow fixed oil (yield 18.78%)
obtained from the seeds of C. ternatea as a ysitosterol based upon
the preparation of acetate and benzoic derivatives and comparing their
comparative compositions, melting points and rotations with those of
known sitosterols. The sterol's properties were described as:
obtained as colorless shiny plates; molecular formula of C^qH^O;
negative test for N,P,S, and halogens; positive Salkowski reaction
and digitonin test; coloration assumed purple to blue, then to green
in Liebermann Burchard reaction; purple coloration in Steinkle
Kehlenberg reaction, which turned cobalt blue on exposure to light.
Sinha (1960) described a a-lactone compound from alcoholic extract
of the dried leaves of Clitori a mariana with a molecular formula of
^26^50^2' froni thie nearly identical propei ties described and
treatments of the new extract, the chemical is "aparajitin," although
Sinha has no reference to the work of Tiwari and Gupta.
Gupta and Lai (1968) reported the isolation of hexacosanol,
B-sitosterol and an anthoxanthin glucoside from the seeds of
C. ternatea. On acid hydrolysis, the anthoxanthin glucoside yielded
quercetin and glucose. Chromatography of the amino acid composition
enabled the identification of the essential amino acids lysine, valine,
methionine, phenylalanine and isoleucine in the protein hydrolysate,
along with aspartic acid, serine, glycine, alanine, glutamic acid,
tryosine, proline, argisine and histidine.
Kulshrestha, Kumar and Kliare (1968) detected cersulfate positive
compounds by thin layer chromatography in the butanol-soluble fraction
of the seeds of ternatea. Four of the six crystalline compounds
were identified as adenosine, kaempferol-3-rhamnoglucoside,

84
p-hydroxy-cinnamic acid and ethyl-alpha-D-galactopyranoside. The
remaining two compounds were a polypeptide and a phenol glycoside.
Aiyar, Narayanan, Seshadri, and Vydeeswaran (1973) reported three
glycosides of kaempferol from the leaves of ternatea. The ethyl
acetate soluble part of the ethanol extract gave three compounds on
chromatography identified as follows: Compound A (m.p. 198-200°) as
kaempferol-3-monoglucoside; Compound B (m.p. 215-7°) as kaempferol-3-
0-rhamnosyl (1^6) glucoside; Compound C (m.p. 188-90°) as kaempferol-
3-rhamnosyl (1^6) galactoside. Light petroleum ether and ether
extracts on chromatography yielded waxy matter, chlorophyll and
8-sitosterol.
Cytological Research
Very little research has been done on the chromosomes of Clitoria.
From the few counts reported, the species appear to fall into two
distinct groups which correlate with the morphological characters which
indicate two subgeneric lines for the species involved. The subgenus
Cl itori a has x=8, and is based on one species. The subgenus
Neurocarpum has x=12, and is based on four species. The chromosome
counts reported for Clitori a species are tabulated in Table 2.
Tschechow and Kartaschowa (1932) included C. ternatea as one of
the species in their cytological survey of the tribe Phaseoleae. They
provided the first chromosome count for the genus (2n=16). A camera
lucida drawing of the chromosomes was provided. They remarked that
C. ternatea differs uniquely from the other species in the tribe both
by the number of x=8 (versus x=ll, 12, 19, 20 in other species, the
latter pair of counts pertaining to other species in the same subtribe)

85
Table 2. Reported chromosome numbers in Clitori a species.
Subgenus CLITORIA
Species
n
2n
date
reporter
ternatea
16
1932
Tschechow & Kartaschowa
ternatea
8
1938
Senn
ternatea
16
1953
Frahm-Leliveld
ternatea
16
1957
Frahm-Leliveld
ternatea
16
1962
Shi bata
ternatea
8
1967
Mitra & Datta
ternatea (double)
8
1961
Saroja
Subgenus NEUROCARPUM
Species
n 2n
date
reporter
cordobensis
24
1951
Krapovickas & Krapovickas
falcata*
24
1960
Frahm-Leliveld
laurifolia
24
1957
Frahm-Leliveld
macrophyl1 a
12
1971
Larsen
*Chromosome
number reported
under the
synonym C. rubiginosa

86
and by the two distinct chromosome sizes, four long chromosomes and
four short ones. Even among the two groups, these eight chromosomes
are not equal in size, although there appears to be a 3:1 size
relationship between the long and short chromosomes.
Krapovickas and Krapovickas (1951) provided a count (2n=24) and
an illustration of the chromosomes of Clitoria cordobensis. They
concluded, after comparison to C. ternatea, that there are either two
distinct groups of species with x=8 and x=12 respectively, or that
C. cordobensis was a triploid.
Sen and Krishnan (1961b) reported that binucleate pollen mother
cells occur at a very low frequency in C. ternatea. One of the two
nuclei was always smaller. Meiosis was not observed in these cells
beyond early diakinesis. An illustration of a binucleated cell with
both nuclei at pachytene was presented.
Developmental Research
Developmental research has been undertaken almost entirely upon
C. ternatea, primarily by Indian researchers. Studies have been
conducted on endosperm, embryo growth, seed germination, and seedling
growth.
Rau (1951) described the endosperm development in C. ternatea
along with four other selected papilionaceous species. The endosperm
is of the nuclear type, with the free endosperm nuclei arranged along
the sides of the inner layer of the inner integument. As the seed
cavity at the chalazal end widens in later stages, the massive nucellus
projects into it. The antipodal end of the embryo sac is wedged into
this nucellar mass. Free endosperm nuclei lying in dense cytoplasm

37
gather here. Cytokinesis of endosperm nuclei occurs when the embryo is
in an advanced stage of development. The cell formation is restricted
to just one or two layers around the embryo; the remaining part retains
free nuclei.
Padmanabhan (1966) reported growing excised embryonal axes in
vitro on Nitsch's basal medium supplemented with vitamins. He
concluded that the course of axis growth was remarkably similar to
natural germination events, namely; (1) growth of root (day 2-3),
(2) elongation of hypocotyl (day 4), and (3) activation of shoot bud
(after day 20). Elongation was not affected by removal of the apical
bud. From these results the observation that upper hypocotyl curves,
he concluded that the embryonal organs are capable of independent
development when isolated, but follow the same developmental pattern
as in the entire embryo.
Lakshmann and Padmanabhan (1968) reported the effect of the
antibiotic ascochitine on in vitro growth of Cl i tori a ternatea. They
concluded that, in general, the toxin exerted its effects on the
embryos grown in vitro at much lower concentrations than is required
to cause perceptible effects on the germinating seeds. Ascochitine
produced total inhibition of root growth, presumably because root
proteins were easily denatured. A low percentage of embryos did
produce callusing of the root apex and the hypocotyl. The shoot
meristem was not affected by the toxin, presumably because the shoot
meristem converts a large quantity of the toxin to dihydroascochitine,
rendering this toxin benign.
Mullick and Chatterji (1967a) reported on seed germination
of C_. ternatea. Imbibition and germination behavior was

88
studied under various eco-physiological conditions in the laboratory,
conditions similar to those found in the wild. They concluded: light
had no effect on either behavior; seeds germinated mostly at a
temperature of 35° ± 2°C; better imbibition and germination behavior
were obtained when seeds were subjected to physical and chemical
treatments of impaction with sand, impaction without sand, hot water,
concentrated sulphuric acid, absolute alcohol and normal potassium
hydroxide. Seed imbibition and germination behavior were about fifteen
per cent under laboratory conditions of light and temperature.
Mu 1 lick and Chatterji (1 967b) reported the effects of sodium
cyanide on seed germination of terna tea. They concluded that seeds
soaked for 24 hours in 100 ppm of aqueous sodium cyanide solution
accelerated imbibition, germination, and promoted early growth of the
seedlings. Increased concentrations produced inhibitory effects
leading to the suppression of normal seed germination and growth.
Light and temperature did not influence the chemical effect on the
seeds.
A. P. de Candolle (1825) described seedling development based upon
three species of Clitori a and illustrated various stages in Plate IX
(Figures 33-35). Two of these species are now placed in the genus
Centrosema. The illustrations show plants with 3-foliate leaves, but
the leaflet shape tends to indicate de Candolle's Figure 33 as
C. ternatea by elimination. Clitoria ternatea typically is 5- and
7-foliate; however, Rizzini's study (1963) clarifies the apparent
discrepancy in leaflet number. With the data representing a mixture
of two genera, de Candolle's treatment becomes a non-reliable source
of seedling development information in Clitoria.

89
Compton (1912) described the seedling morphology and vascular
structure of C. ternatea in early developmental stages. The cotyledons
are 21 x 14 mm, remain enclosed in the thick, foliaceous, oblong testa,
cordate at the base. Germination is epigeal. Hydrocotyl gradually
tapers to a long, much branched primary root. The hypocotyl exhibits
scattered hairs with a transverse wall and recurved pointed tip. The
root is tetrarch with a central pith far below the collet region. Each
phloem strand contains a group of fibers. As one passes up the
hypocotyl, the xylem breaks into groups. Midway up the hydrocotyl,
each xylem quadrant is represented by three groups, a medium protoxylem
tissue and two lateral metaxylem tissues. Higher up the hypocotyl, the
four root protoxylem strands divide and join their metaxylem strands,
producing eight bundles of xylem. Each cotyledon receives four of
these bundles along with two bands of phloem.
Rizzini (1963) reported that Clitoria seeds exhibit one of two
types of germination. Hypogeal germination proceeds quite slowly.
The cotyledons remain within the soil inside their testa, are non-green,
and maintain their volume, not growing. The primary leaves are simple.
The second pair of leaves have the normal three leaflets. Species
exhibiting hypogeal germination were C. laurifolia, C. rubiginosa
(=C. falcata), and C. quianensis. Epigeal germination proceeds
rapidly. The cotyledons shed their testa and are elevated above the
soil level, are green, and grow well. The primary leaves appear above
the cotyledons, and are distinct in shape and number of leaflets from
following leaves. For these species, the second pair of leaves is
similar to the normal 3-foliate leaves. An exception was C. ternatea
where the second pair of leaves (3 leaflets) were dissimilar to typical

90
cauline leaves (5 and 7 leaflets). Species which exhibited epigeal
germination were C. ternatea, C. amazon uni, and C_. racemosa
(=C. fairchildiana). Rizzini concluded that germination type was
constant and could be used as a sectional character, with hypogeal
germination characteristic of section Neurocarpum (=subgenus
Neurocarpum) and epigeal germination characteristic of sections
Ternatea (=subgenus Clitoria) and Clitorianthes (=subgenus Bractearia).
Morpho1ogic:a! Research
Morphological research on Clitori a has been on the double flowers
of C. ternatea, primarily by Indian investigators, and on the
cleistogamous flowers of several species, by a German investigator.
Both topics will be taken up in more detail elsewhere (double flowers,
p. 523; cleistogamy, p. 133). A brief summary of their work is
presented here.
Sirdeshmukh (1955) reported an actinomorphic flower of C. ternatea
with five petals, all standard-like, with the other flowers on the
plant being papilionaceous. The calyx, androecium, and gynoecium were
normal.
Sen and Krishnan (1961 a) reported their genetic research on double
flowers of C_. ternatea. Double flowers are non-papi 1 ionaceous with ten
free stamens in contrast to normal flowers of papilionaceous corolla
and diadelphic androecium. Double flowers were crossed with blue and
white varieties of the normal flowers. The generations indicated a 3:1
frequency of double to single flower types. They concluded that the
double flowers originated through a dominant gene mutation, changing
the organs vital to taxonomic classification, and at the same time,

91
maintaining the viability and fertility of the plant through the
production of large numbers of fruits.
Saroja (1961) reported his findings on the double flowers of
C. ternatea. He found the chromosome number and size in agreement
with single flower results reported by earlier investigators. However,
examination of pollen grains showed differences. The pollen grains of
double flowers were larger and the percentage of fertile pollen was
lower. He concluded that the action of the mutant gene is manifested
in different ways and therefore may be considered pleiotropic.
Harms (1907c) described the occurrence of cleistogamous flowers
in three species of Clitori a, namely laurifolia (reported under
synonym C. cajanifolia), C_. falcata (reported under synonym C.
glycinoides) and C. guianensis. All three species were illustrated.
He described and compared the calyx, corolla, androecium, and gynoecium
of cleistogamous and chasmogamous flowers.
In describing a new species, Burkart ( 1941) reported the
occurrence of cleistogamous flowers in C. cordobensis. He compared
these flowers with those reported by Harms.
Agronqmic Research
Most of the agronomic research has utilized C. ternatea along with
other legumes in pasture surveys and experimental plots to determine
its usability as a forage crop. Nutrient studies have been reported
using sheep, although the species is reported in a number of floras and
economic botany references, to also be edible by cattle and goats.
Holland and Joachim (1933) conducted soil erosion experiments for
six years on a steep slope planted with tea in Ceylon (-Sri Lanka).

92
Of the two legumes used, one was the shrubby C_. cajanifol ia
(~C. laurifolia) planted as a hedge. The l!itori a hedge grew well
from the beginning and proved most satisfactory as a contour hedge,
reducing erosion in its plots by 92 per cent.
Kok, Machado, and Meirelles (1943) reported the results of
digestion trials with sheep utilizing the hay of several legumes,
including C. ternatea. Cuttings of the species were made 74 days
after sowing, in full blossom, at a medial height of 70 cm, and fed
to three sheep. Chemical results analyzing sheep feces, data on
quantity consumed, digestibility coefficients, and nutrient values
were presented. They concluded that C. ternatea hay compares in
quality with that of alfalfa in early bloom.
In another study with sheep, Katiyar, Ranjhan, and Shukla (1970)
reported their experimental studies on yield, chemical composition,
and nutritional value of C. ternatea. They noted that: the plant is
suitable for growing in association with grasses, provided it is
protected from summer drought and winter frost; the sheep preferred
grazing on the leaves, flowers, and pods, but not on the harder woody
stems; the sheep were able to consume sufficient quantity because of
heavy plant production of edible structures; the chemical composition
data of ternatea compared favorably with most of the cultivated
legume crops; the nutrients were equally digestible in both green and
hay fodder; and C. ternatea alone or in association with other
perennial grasses proved promising for forage.
Hassell (1945) reported his experimental results of several
legumes established in pastures in Queensland, Australia. He noted
in a brief descriptive note of the species, that cattle are particularly

93
fond of C. ternatea. Also noted was the difficulty of eradicating the
species because of its deep-rooting habit and heavy seeding. Hassell
planted seeds of all native legumes found growing in the Central
district in experimental grazing plots. Two introduced legumes were
also planted. He concluded that germination was very poor for all
species (which would include C. ternatea) except for the two introduced
legumes.
Bunting and Lea (1956) reported their observations on legumes and
grasses of natural and established pastures in the Sudan. Three
strains of C. ternatea were sown on plots and observed for three years.
Two strains were affected by viral disease, the third apparently
resistant strain growing well. This strain, along with Rhodes grass,
was established in a low lying area subjected to rain floodings, and
appeared promising in providing a useful means of associating some
stock with arable development in the region.
Oakes (1970) evaluated various legumes in field trials in
St. Croix, U.S. Virgin Islands, to determine which legumes possess
characteristics which make their utilization desirable and advantageous
in supplying forage and improving forage quality at critical times
when feed supplies are low, in poor quality, and in maximum demand.
Forage legumes were established by seedling on Fredensborg clay, in
pure and mixed stands, (C1itori a ternatea grown with grass Panicum
maximum in alternating rows), and without the addition of fertilizer.
Oakes concluded that C^ ternatea would make a significant contribution
to browse and pollard available to livestock.

94
Palynological Reseat ch
No description of pollen grains has been
absent from standard palynological references
found. Clitoria is
(e.g. , Erdtman, 1952).

MORPHOLOGY
A number of Clitori a species exhibit great morphological variation
in size, form, and pubescence of the vegetative, flowering, and
fruiting structures. These variations often can be observed in the
same population or on the same individual plant. Yet despite the
variations, an array of morphological facies permits recognition of a
constant pattern. Patterns of variation can also be observed above the
species level. Observed patterns of variation for a structure will be
noted, both within the subgenus and in comparison to other subgenera.
Comparisons with the related genera of the Glycineae will be noted
using the descriptive treatment of Hutchinson (1964) as a model.
Habi t
The habit of Clitori a has the greatest variability to be found
within the Glycineae, and perhaps one of the most variable of the
Papilionaceous legume genera. Genera of the Glycineae are perennial
herbs or shrubs with many genera having climbing properties. The
members of Cl itoria are trees, shrubs, lianas, subshrubs (or woody
herbs), perennial herbs, and herbaceous twining vines, with some
species exhibiting intermediate habits.
Three woody habit types are associated with the subgenus
Bractearia and absent (one exception, C. lasciva, a liana of subgenus
Clitoria) from the other subgenera. These plants are found mostly in
95

96
tropical forests and exhibit tree, tall shiub, or liana habits. Field
observations on these plants are sparsely i oponed. Trees are recorded
up to 30 m tall and 60 cm d.b.h. Clitoria rajrchi 1diana is noted for
its arching virgate branches that droop, touching the ground to form a
canopy. Tall erect shrubs (sometimes reported as treelets) are usually
3-8 m tall. Occasionally reported is a subclimhing property of their
upper branches (e.g., C. amazonum). Woody vines or lianas climb by
their twining stems. No specialized climbing structures (e.g.,
adventitious roots) have been noted. Occasionally a liana species has
an individual reported growing as an erect shrub in open areas
(e.g., C. javitensis), or a species of disjunct distribution exhibiting
the liana habit in one region and a shrub habit in another region
(e.g., C. 1eptostachya, a liana in Guyana, a shrub in Brazil).
The plants of the subgenus Neurocarpum are usually associated with
drier, water-deficient, open habitats. They exhibit habits of shrubs,
subshrubs, perennial sutfrutescent herbs, and herbaceous twining vines.
Shrubs are low-growing, or to 5 m tall, often associated with
mountainous terrain. Clitori a polystachya is the only shrub species
with a report of a climbing apex (one Mexican population). Subshrubs
(or woody herbs) are plants with one to a few erect stems above ground,
usually 0.5-1 m tall, the stems well 1ignified nearly the entire length
(e.g., C. guianensis). They are usually found in savannas and open
grasslands. During periods of water stress, all structures are dropped
leaving only the main stem axis above ground, which may persist for a
long period of time, is fire-tolerant, and eventually is aborted. When
water becomes available, new stein(s) are produced. Suffrutescent herbs
are usually erect, the stems becoming more or less lax with age

97
(e.g., C. fragrans). The herbaceous stem apices may elongate and
become twining vines (e.g., C. mariana), ot prostrate creepers
(e.g., C. mexicana), or procumbent (e.g., linaria). Herbaceous vines
(e.g., C. falcata) are usually associated with forests, and are high
climbers, although they will creep about the forest floor in more open
areas. Vines of subgenus Clitoria grow in more open areas, often
forming dense tangled mats, but will climb when conditions are suitable
(e.g., C. ternatea). The base of the plant is suffrutescent.
The few species of the subgenus Cl i toi ha are suffrutescent herbs
(similar to those of the subgenus Neurocarpurn) or vines as noted above,
with one liana species.
Stems and Branches
Steins and branches of most species are longitudinally striated.
In non-climbing plants, one can follow these striations and observe a
slight twisting of the axis. A gradient of pubescence usually exists
along these axes, from densely pubescent near the tip, to glabrate
towards the base. Juvenile parts often exhibit a subquadrangular or
angular cross-section, becoming rounded to terete with age. Internodes
are often hollow in the center of the woody plants with a few species
having a sol id pith. The entire plant of the more herbaceous members
and some subshrubs are mounted on herbaria sheets; thus the pith
condition is inconspicuous, although the same pattern occurs as in the
woody plants. Bark is rarely noted in field observations, although it
tends to be somewhat smooth with numerous cracks on the trunk when
reported. Only wood samples of C. nervosa were seen. Bark of smaller

98
diameter stems and branches is grayish to dark brown, smooth, splitting
in longitudinal strips to expose light-colored wood beneath.
Subterranean stems (usually called xylopodia) of the subshrubs and
perennial erect herbs are poorly collected. Usually the xylopodium is
woody, unbranched, often knobby at the apex where the aerial portions
sprout, rootless or bearing one to a few short roots of 0.5-2 mm
diameter, glabrate or bearing scattered microscopic uncinate hairs.
Xylopodia are short with a distinct elongated axis that parallels the
surface, and rarely collected with an elongated root-like appendage at
its distal end. Field observations of C. mariana and C. fragrans
indicate that the larger rhizome-like structure lies near the surface
of the soil with the distal portion moving deeply down into the soil at
a thirty to forty-five degree angle. The distal portion is 2 m or more
long (and is often broken during the digging process) and 2-5 mm thick.
Occasionally there are swollen portions to 10 cm long and 1 cm
diameter in this distal portion.
Pubescence
Trichomes are present on most plant structures, and along with
other characteristies, can be useful in distinguishing two superficially
similar species (e.g., C. javitensis and C. sagotii). They occur in
two sizes, microscopic (must be viewed at 25-30X) and macroscopic
(viewed at 1 OX or seen with naked eye). Microscopic trichomes are
uniseriate, usually bicellular to multicellular, and uncinate. These
hairs are inconspicuous and often difficult to view even when one
knows that they are present. They are best observed by looking along
the edge of the surface rather than viewed in a face view. The hairs

99
are also usually more readily seen along major nerves. Uncinate
trichomes are often found along the various axes of the inflorescence,
on the calyx, on the outer vexillum surface, on leaf surfaces and
nerves, on bracts and bracteoles, on legume valves and sutures, on
ovaries, and infrequently on the ala, carina, and staminal tube.
Macroscopic hairs, if present, often mask t.he uncinate trichomes.
Macroscopic trichomes are usually 0.3-0.8 (1.0) mm long, with only a
few species exhibiting longer hairs. They are often appressed or
subappressed, less frequently falcate to erect. The term "minute" will
be used for trichomes 0.1-0.5 mm long. The term "long" will be used
for trichomes greater than 1 mm long.
Juvenile structures are often densely pubescent, becoming
moderately pubescent to glabrate with age. Those structures lacking
trichomes are: the inner surface of the vexillum, calyx, stipules,
stipels, bracts and bracteoles; the ala and carina (usually); the
androecium (except the tube apex occasionally or the free filaments);
and the seed. The vegetative structures (except as noted), the
gynoecium, the outer surfaces of the vexillum, calyx, bracts, and
bracteoles, and the fruit possess some degree of pubescence. Fruits
and calyx often will appear glabrous because of the inconspicuous,
microscopic trichomes. The upper leaf surface becomes glabrate to
glabrous early, with vestigial hairs sometimes persisting on the midrib
and major nerves. Those species which retain the pubescence on the
upper surface are worth noting.

100
Leaves
Leaves are alternate, odd-pinnately compound, usually 3-foliate,
less commonly 5- or 7-foliate, infrequently uni foliate, rarely 9- or
11-foliate. The terminal leaflet is usually slightly larger to much
larger than the lateral leaflets. Leaves have been often cited as
"trifoliate," an ambiguous term which can apply to palmately compound
leaves. When used with the genus Clitoria, the term refers to the
common condition of three pinnate leaflets. Leaves are usually
stalked, although nearly half the members of the subgenus Neurocarpum
have subsessile leaves in which the petiole is shorter than the rachis.
Unifoliate leaves are sessile with the exception of the Asian
C. cordi form is and C_._ macrophylla. In those plants bearing both
trifoliate and unifoliate leaves, the trifoliate leaves are similar in
appearance to the terminal leaflet, although often broader. The
unifoliate leaf is attached by its petioluie and appears to be a
terminal leaflet in which the petiole, rachis, and lateral leaflets
were lost. In C. cordiformis and C. macrophylla, the leaf's petioluie
is attached to a long stalk which is somewhat rugose laterally at its
apex, a condition observed only in petiole bases and petiolules of
other species. This suggests a terminal leaflet and petiole, with the
rachis and lateral leaflets lost. In C. pozuzoensis, a variation
occurs in which some leaves appear palmately compound due to the
reduction and near elimination of the rachis. Leaflets of C. epetiolata
are retained while the petiole and rachis are both lost, the leaf thus
becoming digitately compound.

1Ü1
Leaflet
Leaflets are polymorphic in shape and size within many species.
Leaflets are entire with revolute margins occurring only in C. coriácea
and C_^ tunuhiensjs. The apex is more or less obtuse, usually abruptly
to rapidly forming a short to long acumen in the more ligneous species.
In advanced species, the acumen is lost, the leaflet apex being obtuse
to retuse, or occasionally emarginate. The extreme apex is often more
or less mucronately bristled. The base is cuneate to rotund, rarely
cordate. Nerves are conspicuous, prominently raised below with the
midrib and major nerves impressed to weakly raised above. The midrib
sometimes appears raised (will catch a fingernail) because of a shallow
groove on each side of the midrib. Primary nerves are ascending,
usually arcuate to falcate toward the margin, anastomosed with the
nerve above, and fairly consistent in number. Secondary nerves are
conspicuously reticulate, especially below. The upper surface is
typically green to dark green, glabrate to glabrous when matured. The
lower surface is green to pale green, often pubescent, and glaucous to
glaucescent in a number of species of the subgenus Neurocarpum. Minute
wax globules are present on the lower leaf surface of coriácea.
Stipules and Stipels
Stipules and stipels are persistent to deciduous, rarely caducous,
longitudinally striated, and paired. Stipules are erect, appressed,
often ciliate, usually 2-12 mm long and 1-4 (6) mm wide. They are
caducous in andrei, C. dendrina, and brachystegia. Stipules are
conspicuously elongated in plumosa (2-3 cm) and kaieteurensis
(1-2 cm), and become minutely foliaceous in stipularis (13-19 mm
x 7-13 mm). Stipule length can be variable in some species (even on an

102
individual plant) by half a centimeter or more (i.e., C. obidensis,
C. guianensis). Stipels are subequal to or shorter than the stipules
in length, and much narrower. Terminal stipels are often conspicuously
shorter than the lateral stipels. Stipels are typically less than
1.5 mm wide except in £. kaessneri (1.5-2 mm) and C. stipularis
(2-3 mm). Elongated stipels occur only in plumosa (1-2 cm).
Petiole and Rachis
The leal stalk is longitudinally striated, quadrangular to terete;
its pubescence is usually similar to that of stems or branches, and it
sometimes bears one or two caniculi on its adaxial side. The rachis is
similar to the petiole; sometimes it is weakly compressed laterally,
and usually shorter, except in the subsessile leaves of many members of
subgenus Neurocarpuin where it is longer than the petiole, or in the
subgenus Clitori a where the combined rachis segments exceed the
petiole. Subsessile leaves have petioles 0.2-2 cm long.
Petiolule
The petiolule is usually subquadrate, dark-colored, rugose, and
more densely pubescent than the petiole and rachis.
Observations
Measurements of each leaf structure form nearly a continuous
gradient within the genus. Stipules and stipels of most species are
similar in shape with overlapping sizes. Leaflet size, shape, and apex
form gradients within the total variation continuum. Petiole and
rachis size is usually variable within a species, except for those
species with subsessile leaves. Leaflet number and petiolule length
appear more reliable as diagnostic characteristics.

103
At the subgeneric level, leaf structures are best used as
supportive characters, but there are exceptions. These structures are
summarized in Table 3. In general, leaves of the species of subgenus
Bractearia are large, long-stalked, and 3-foliate, with the leaflets
acuminate and large petioluled. Leaves of subgenus Clitoria species
are small, short-stalked, and 5- to 11-foliate, with short-petioluled
leaflets having obtuse to emarginate apices. Leaves of species in
subgenus Neurocarpum are highly variable and difficult to characterize
But they are usually distinguished from Clitoria species by a smaller
leaflet number and from Bractearia species by a generally smaller size
Inflorescence
The inflorescence is axillary, usually solitary, or terminal in
some tree and tall shrub species, or cauliilorous in some liana
species. Cauliflorous inflorescences are usually few to fascicled,
located below the leaves. Inflorescences are basically racemose,
infrequently paniculate, or commonly reduced to one or two flowers in
advanced species. Inflorescences are typically chasmogamous or
occasionally cleistogamous (subgenus Neurocarpum), very rarely
chasmocleistogamous. Flowers are one to many, but paired at the nodes
The central axis is angular to terete, longitudinally striated, often
bearing moderate to dense, short or microscopic trichomes, and more or
less nodose. This inflorescence is similar to other genera of the
Glycineae which typically have axillary, racemose, chasmogamous, and
more or less nodose inflorescences. The genus Clitoria is unique in
possessing cauliferous, cleistogamous (except the genus Amphicarpa),

Table 3. Comparison of leaves among the subgenera.
CHARACTER BRACTEARIA CLITORIA
NEUROCARPUM
Leaflet number
3
5, 7, 9, 11
(C. kaessneri = 3
& 1)
3, 3 & 1, 1
Petiolule
Large, 4-10 mm
Small, 1-3 mm
Small to medium, 1-5 (6) mm
(C. stipularis = 4-8 mm)
Leaf apex
Acumen
Acuminate (obtuse)
0.5-3 cm
Obtuse to emarginate
Obtuse (Acuminate)(Emarginate)
0.5-1 (2) cm
Petiole
Long, 4-10 (16) cm
Subsessi1e, 0.2-2
Short, 2-6 cm
cm
Subsessile, 0.2-2 cm
Short to medium, 2-6 (10) cm
Absent
Petiole vs rachis
Much longer
Much shorter
Much longer in stalked leaves
to shorter in subsessile
1 eaves
Leaflet size
Length
Width
Large
(6) 8-20 cm
3-9 (18) cm
Small (Medium)
2-7 (10) cm
1-3 cm
Vari abl e
2-10 (15) cm
1-9 cm

105
and chasmocleistogamous inflorescences. Oilier genera of the Glycineae
lack lianas, thus lack cauliferous inflorescences.
Chasmogainous Infloresce11ces
Inflorescences which bear the typical cross-pollinated,
papilionaceous flowers are usually lignificd to some degree, but
become highly woody in those species of the subgenus Bractearia.
These inflorescences bear numerous flowers which abort at the lower
nodes. The inflorescence nodes are thickened, knobby, projecting
outward from the main axis, a condition in legume axes called "nodose."
There are four main inflorescence types which can be recognized in the
subgenus: the panicle, the subpanicle, the nodose raceme, and the
flexuous raceme. Examples of inflorescence types are illustrated in
Figure 3.
The woody panicle is associated with tree species bearing the
primitive bracteoles (cf. pp. 113-114), and it is probably the
primitive inflorescence type within the genus. The woody panicle has
a central main axis with primary lateral branches (5-30 mm long)
bearing a pair of flowers at its apex (rarely three flowers). Flowers
abort first at their pedicel base, leaving the primary lateral branch
with its persistent bracts. Then the primary branch will abort
leaving a nodose axis. Examples of this inflorescence type is found
in C. ainazonum and C. arbórea.
The woody subpanicle is derived from the woody panicle and
superficially appears racemose, although the primary branches are
present, minute, usually 1-5 mm long. Examples of this inflorescence
type is found in C. fairchi1 diana and C. nervosa. Both the woody
panicle and subpanicle are found only in the section Bractearia.

Figure 3. Generalized inflorescence types in Clitoria. Top row, woody
inflorescences: (a) woody pani>le, a, found in C. arbórea;
(b) woody subpanicle, as found in C. fa ire hi!diana; (c) nodose
raceme, as found in C. javitens is; (d) flexuose raceme, as
found in C. f1exuosa. Bottom r«,w, herbaceous inflorescences:
(e) panicle, as found in pol ystach./a; (f) few-flowered
raceme, as found in C. mariana var. oriental is; (g) biflowered
or pedunculate raceme, as found in epet iolata; (h) uni-
flowered raceme, as found in C_. terna tea. ("Dots on woody
inflorescence axis represent woody knobs, or "nodose"
condition. Flowers and primary branches commonly abort, thus,
inflorescences illustrated reflect this and were drawn from
specimens. )

107
a

108
The flexuous raceme is limited to the section Flexuosa, and is
also associated with primitive bracteoles. This inflorescence has a
weakly to conspicuously zigzag main axis, with primary lateral branches
1-2 mm long or lacking. Examples of this inflorescence type are found
in flexuosa and C. pozuzoensis.
The woody raceme has a conspicuous nodose axis from which a pair
of pedicels is borne. This is the most common inflorescence type in
the subgenus, and is associated with the advanced bracteole type. It
can become extremely elongated (as in leptostachya and C. pendens,
to 1 m long) or very short, nearly subsessile, with the nodose aspect
inconspicuous (as in C_. sacjotiJ_ and coriácea, subsessile to 0.5 cm
long).
Herbaceous inflorescences usually beat a reduced number of
flowers, often two, less commonly few to many flowers. Other genera of
the Glycineae with their herbaceous inflorescences typically are several
to multi-flowered. Panicles are uncommon, but herbaceous panicles are
found in a few shrubby Mexican species (e.y., C. polystachya and
C. monticola). The primary lateral branches of the panicle are
variable in length, typically with two to three flowers at its apex,
infrequently further branched in large panicles. Racemes have a
peduncle with a pair of flowers (one may abort) at its apex, rarely
having a short internode segment which bears an additional two or four
flowers. This inflorescence type is typified by falcata and
C. guianensis. The few-flowered raceme commonly bears (two) four to
eight flowers, as is typically found in C. mexicana and C. australis.
The ultimate reduction occurs in uni flowered racemes as in C. ternatea,

109
where the solitary flower is borne laterally from the peduncle apex,
or sometimes terminally. Rarely a second tlower is produced.
Cl eistogamo u I njfl or esc t.- nee s
Cleistogamous inflorescences occur only in some species of the
subgenus Neui ocarpum. Tliey are similar to the chasmogamous inflores¬
cence type of the species, usually occurring at the lower nodes.
Inflorescences are ascending and remain so, never pushing the fruits
towards or into the soil. Occasionally these inflorescences appear at
the upper nodes above the chasmogamous inflorescences. Although both
chasmogamous and cleistogamous inflorescences may occur at the same
time on a plant (plant chasmodichogamous), plants are usually collected
with only one inflorescence type.
Chasmocl ei stogamous I rrf 1 orescences
ChasmocIeistogamy is extremely rare in the genus, observed only
once in two species, and reported here for the first time in the genus.
In both cases, the inflorescence was a two-flowered raceme in which one
pedicel bore a chasmogamous flower while the other pedicel bore a
cleistogamous flower in fruit. Chasmocleistogamy occurred in a
specimen of C. guianensis from Minas Gerais, Brazil (Irwin 2384,
US 2324452). The cleistogamous flower bore a dehiscent fruit. The
chasmogamous flower bore the persistent calyx, the petals aborted.
Chasmocleistogamy occurred in a specimen of C^ mariana from
Pennsylvania, United States of America (J.J. Carter s^n^, "Culley's,
Lan. Co., 1 Aug 1906," PH). The cleistogamous rlower was in a juvenile
fruiting stage. The chasmogamous flower was in full bloom. In both
collections, all other flowers were cleistogamous on the plant, as well
as in the duplicate inflorescences seen (Irvin: NY, UC; Carter: no

no
duplicates seen). These chasmocleistogamous inflorescences are
illustrated in Figure 4.
Bracts
Bracts are stipule-like, longitudinally striated, pubescent
outside, often ciliate, and found in three series below the base of the
pedicels. Ttie inner pair of bracts is caducous, usually the smallest,
and located between the pair of pedicels. The middle pair of bracts is
the most conspicuous, persistent, usually the largest, and located
opposite the pedicels, each bract concave and more or less appressed
around the pedicel, becoming spreading to reflexed in age. The outer
pair of bracts is deciduous to semipersistent, usually nearly as large
as the middle pair, and located between the pedicels. Unless otherwise
noted, descriptions of bracts will refer to the middle pair (i.e., the
bracts usually found).
Bracts are commonly 3-7 mm long and 1-4 mm wide. Elongated bracts
occur in flexuosus, C. woytkowskia, C_. obidensis, C. plumosa,
C• stipularis, and C. densiflora, reaching 16 mm in length. Broadened
bracts occur in C. amazonum, C_. f1exuosus, and C. woytkowskia.
Clitori a amazonurn has bracts 3-6 mm wide, whereas the bracts of the
latter two species are very large, conspicuous, and foliaceous, up to
9 mm wide. Conspicuously reduced bracts of 1-3 mm long and 0.5-1.5 mm
wide occur in the woody species C. dendrina, C. brachycalyx,
C. brachystegia, C^ 1eptostachya, and C^ pendens. Minute, inconspicuous
bracts of 1-2 mm long and 0.2-0.4 mm wide occur in the nonrelated herb,
C. heterophyl1 a.

Figure 4. Chasmocleistogamy in Clitoria. Leaves removed from
illustration. C. guianensis (Irwin 2384, US 2324452).
C. mariana (J. J. Carter s.n., Culley's, Lan. Co. Penn.
Aug 1. 1906, PH).

112
CLITÃœR ] A .L4RL4NA
CH = Chasmogamous Flower
CL = Cleistogamous FI ower
.UIANENS IS

113
Pedicels
The pedicels are paired, longitudinally striated, usually
pubescent, conspicuously twisted, and less than 1 cm long. They become
thicker in fruit, and woody in the subgenus Bractearia. When the
flower aborts, the abcission zone occurs below the base of the pedicel.
Bracteoles
The bracteoles are paired, persistent, longitudinally striated,
pubescent outside, often ciliate, inserted on the pedicels between
0.5-1.0 mm below the calyx, rarely longer. In two sections of the
subgenus Bractearia, the bracteoles are conspicuously large (more
1 cm long and 0.6 cm wide), coriaceous, subequal to longer than the
calyx and protecting the flower in its bud state. Hutchinson (1969)
indicated that these petaloid bracteoles are primitive and common in
the subfamily Brachystegioideae of the Caes.a 1 piniaceae and lacking in
the Fabaceae. In other species of the subgenus, the bracteoles are
typically 3-12 mm long and 1-3 (6) mm wide. Elongated bracteoles occur
in C. plumosa (22-25 mm) and obi densis (20-35 mm), both also 3-6 mm
wide. Reduced bracteoles of 1.5-3 mm occur in dendrina, C.
brachycalyx, brachystegia and a variety of C. 1eptostachya, the same
species with reduced bracts.
In the subgenus Clitoria, bracteoles are pellucid, translucent to
semitransparent between the veins, and similar to the calyx. They
shrink in size as the fruit is formed, unlike the coriaceous to
cartilaginous bracteoles of the other subgenera which retain their size
in fruit. In some members of ternatea, the bracteoles will become
subequal to the calyx and enclose the flower in bud, although there are

114
reduced bracteoles (ca. half the calyx) which do not protect the bud
even on the same plant. Minute bracteoles occur in C. heterophyl1 a.
In the subgenus Neurocarpum, large bracteoles occur in
C. stipularis and densiflora, but are not the petaloid protective
bracteoles of the primitive type. These two species are noted for the
large stipule-like structures, the stipules, bracts and bracteoles.
Observations
After the segregation of the subgenera by fruits and seeds, the
inflorescences and bracteoles can be used as supportive characters.
The differences have been noted and are summarized in Table 5.
Bracteoles and inflorescence types are good characters to distinguish
large groups of species, and sometimes serve as sectional characters.
Bracts and pedicels are less reliable and best used to differentiate
among small numbers of species.
Chasmogamous FI owei’s
Chasmogamous flowers are resupinate (a characteristic unique to
the Glycineae, segregating it from other legumes), showy, papilionaceous,
and blue to violet, pink to rose, or white, occasionally white fading
yellow in the dried state. Flowers are 0011,1)1 icate before opening, and
are often mounted this way. In some members of C. ternatea,
"double-flowers" occur which are 4- or 5-merous, actinomorphic, with
all petals banner-like. A generalized chasmogamous flower is
illustrated in Figure 5.
Chasmogamous flowers are of three distinct sizes in the subgenus
Bractearia. Small flowers are uncommon, consistently 2-4 cm.
Medium-sized flowers are 4-6 cm and nearly as common as the large-sized

Figure 5. A generalized chasmogamous flowi r. (a) flower, longitudinal
section, x 2; (b) flower, x 1; (c) flower with petals aborted
and early development of fruit, x 1; (d) vex i 11 uni, x 1;
(e) ala and carinum, x 1; (f) aodroecium, x 1; (g) gynoecium,
x 1; (h) apex of style, x 5; (i; peduncle; (j) pedicel;
(k) bract; (1) bracteole; (in) aodroecium; (n) blade of keel
petal; (o) vexillum claw; (p) siyle; (q) blade of wing petal;
(r) blade of vexillum; (s) claw of ke. 1 petal; (t) claw of wing
petal; (u) calyx lobe; (v) caly: throat; (w) swollen calyx
base.

116
q

117
flowers of 6-8 cm. In two species, flexuosus and pozuzoensis,
individual flowers sometimes reach 9.5 cm in length. In the other two
subgenera, the same variation pattern is present except that the size
range of medium and large flowers is half a centimeter less. Thus for
subgenera Clitori a and Neurocarpum, small flowers are typically
2-3.5 (4) cm, medium-sized flowers are typically 3.5-5.5 cm, and large
flowers are 5.5-7.5 cm. Flowers of 8-9.5 cm are absent in these two
subgenera.
Flower color is variable and often different collectors report
different colors for the same species. In general, flower color refers
to the inner surface of the vexillum. The outer surface is whitish
with the inner surface bearing the conspicuous pigmentation. The alae
and carina are paler in color to whitish. The inner surface of the
vexillum usually has a narrow patch of white or pale to deep yellow in
the center near and extending down into the throat. The veins are
numerous, dark-pigmented compared to the pigmentation of most of the
vexillum surface, usually dark blue to violet or reddish, converging
at the throat. This color generalization is based upon the few
herbarium sheets with detailed field data on flower color and
observations of live flowers in Florida populations (C. mariana,
C. fragrans, and cultivated ternatea). In these populations, flower
color was variable from light to darker shades.
In general, those species with blue to violet or rose to lilac or
pink shades are associated with the medium-sized to large flowers, only
occasionally with small flowers (e.g., C. dendrina). In general,
species with white flowers have small flowers, although occasionally
medium-si zed (e.g., C. falcata). Yellow is a rare flower color in the

118
genus. Occasionally a species with white flowers has the flower color
change from white to pale yellow with age, becoming a deeper yellow in
the dried state (e.g., C. falcata).
Calyx
The calyx is infundibular, persistent in fruit, conspicuously
nerved, and 5-lobed. The upper two lobes (opposite vexillum, located
dorsally due to resupination) are subconnate and the lowermost lobe
(opposite carina, located ventrally due to resupination) is narrower,
often longer. The calyx is usually pubescent, although often seemingly
glabrate because of the microscopic hairs and few, appressed to
suberect, inconspicuous macroscopic trichomes. The calyx is reported
as green or violet-tinged to purplish towards the base (base darker-
colored in di'ied state).
The tubular calyx is infundibular. The base rapidly expands
outward, then gradually with a tapering expansion becomes broader up
to the throat, where the lobes ascend but spread slightly outward.
The swollen, expanded, usually dark-colored base indicates the location
of the internal disc to which the claws of the petals and the
androecium are attached. The lobes are broad, subequal, usually
rapidly tapering above tlie middle to the apex, except for the narrower
and usually longer ventral lobe. The lobes are shorter than to
occasionally nearly subequal to the tube. An exception occurs in the
Asian species of subgenus Neurocarpuin, where the lobes are subequal to
slightly longer than the tube (e.g., C. macrophylla). A shortened
tubular calyx with reduced teeth occurs in section Brachycalyx of the
subgenus Bractearia. The calyx becomes cup-shaped at maturity in

119
C. brachycalyx and C. brachystegia, superficially resembling some
campanulate cálices, although the lobes are different.
The calyx nerves are numerous and difficult to count in the
subgenus Bractearia. In many species of the other two subgenera, there
is a more or less prominently 10-nerved calyx. Five nerves ascend
directly into the apex of the lobes. Five nerves ascend to the throat
and beneath the sinus, dichotomously forked with the branch nerves
radiating into each adjacent lobe, more or less near the margin, and
ascending to the lobe apex. The lobes thus appear prominently
3-nerved. These nerves which arise from the primary nerves are more
or less prominent in individual flowers.
Calyx texture is difficult to characterize. It shows no apparent
shrinking as it persists during the fruiting stage of the plant in
subgenera Bractearia and Neurocarpum, and is more coriaceous in the
former subgenus and more cartilaginous in the latter. In the subgenus
Clitori a, the calyx is subpel lucid between the nerves and shrinks
slightly during the fruiting stage.
Calyx measurements are often lacking in species descriptions of
Clitori a. When reported, measurements include the entire calyx length,
and are seldom indicative of the calyx parts (tube length, lobe length,
tube width at throat). Tube lengths are variable, although patterns of
variation are discernible. The large flowers of species in Bractearia
typically have a calyx tube 1.5-2.5 cm long, with a few species to 3 cm
(C. flexuosa, C. pozuzoensis) which nearly approximates the tube length
of medium-sized flowers (1.3-2 cm). The large flowers of subgenus
Neurocarpum agree with those in the subgenus Bractearia whereas the
medium-size flowers of subgenus Neurocarpum (tube 0.7-1.6 cm) agree

120
with the tube length (0.7-1.3) of small-size flowers. In the subgenus
Clitori a médium-size flowers agree with those in subgenus Neurocarpum
whereas the small flowers possess tubes of 0.4-0.8 cm. Tube length
often varies from one-half to one centimeter in large flowers and is
less variable in small and medium-sized flowers.
The length of the lobes is highly variable, arid does not correlate
with flower size. The lobes of the majority of species in the subgenus
Bractearia are 3-7 mm long whereas most species in the subgenus
Neurocarpum are 7-12 mm long. Each subgenus has a few species in the
opposite range of variation. Those species in subgenus Clitori a are
split between these two ranges. Extremely short lobes (1-3 mm) occur
in C^ brachycalyx, C. brachystegia, C^ glaberrima and C^ heterophyl1 a.
Extremely long lobes (13-18 mm) occur in £. plumosa and C^ densiflora,
with lobes of 20-27 mm in C^ obidensis. Calyx lobes are commonly
2-5 mm wide near their base with broad lobes (5-8 mm) occurring in
C. obidensis and C . dens if1 ora, and very narrow lobes (0.4-1 mm) in
C. heterophyl1 a. The ventral lobe of C. plumosa is unique in that its
length (25-30 mm) greatly exceeds the length (13-18 mm) of the other
four lobes.
The total length of the calyx (tube plus lobes) is variable,
although large flowers generally have a calyx length of 2-4 (5) cm
whereas small and medium-sized flowers have a calyx length of 1-3 cm.
There is no distinct pattern along subgeneric lines. An extremely
short calyx length (0.6-1.0) cm occurs in C. heterophyl1 a and a very
long calyx length (4-5.5 cm) is found in C. obidensis and C. plumosa.
The width of the calyx tube at the throat is variable, but in
large flowers, calyx throats generally are 8-13 mm wide, whereas the

121
calyx throat width in small flowers is 3-6 mm (except for C.
brachycalyx, 8-12 mm, and dendrina, 5-8 mm). Medium-sized flowers
exhibit two calyx throat widths, the larger width of 8-13 mm and an
intermediate width of 5-8 mm. Species of subgenus Bractearia are
7-13 mm wide (except dendrina, 5-8 mm). Species of the subgenus
Neurocarpum are 4-8 mm wide (except simplicifolia, laurifolia,
and C. falcata, all (6) 7-11 mm wide). Species of the subgenus
Clitoria are divided between these two sizes. The smallest calyx tube
throat width is 3-4 mm in (_ heterophylla. Calyx tube widths at the
calyx base are typically 2-8 mm, and show no apparent distinct patterns
of variation. The base width is distinctly large in C. plumosa
(8-10 mm).
Typical calyx pubescence is composed of microscopic, uncinate
trichomes with a few, short, appressed macro-trichomes, or less
commonly erect. A glabrous calyx (which may sometimes bear widely
scattered uncinate trichomes) occurs only in C. amazonum and
C. glaberrima. Densely silky trichomes occur only in section FIexuosa
(subgenus Bractearia). Long reddish, shaggy trichomes, highly
conspicuous to the eye, occur in C. plumosa. A dense mat of puberulous
hairs and some uncinate trichomes occur in C^ dendrina, C. brachycalyx
and C_^ brachystegia.
Vex i 11 uin
The vex ilium is complicate, large and U-shaped when expanded,
erect, obovate to orbicular, more or less emarginate, usually large and
showy, short-clawed, appendageless, with the pigmentation of the inner
surface darker towards periphery and light-colored center with veins

122
promi na tel y darker-colored, converging into the throat, and is pubescent
on the outer surface, often minutely ciliolate near apex.
The vex ilium of Cl itoria is similar to those that are described in
other members of the Glycineae, although usually larger. A few genera
bear auricled appendages and Centrosenia is unique in possessing a spur.
Centrosema also differs from Clitori a by having an incurved claw. Yet
the similarity in appearance of the vex illum apparently is confusing to
some botanists (e.g., the large number of generic misidentifications
despite the incurved claw and spur of Centrosema). When open, the
Clitoria vexillum forms a narrow "U-shaped" to broad "V-shaped"
platform wi th an open throat, the alae and carina elevated above the
vexillum surface. Each half of the vexillum is concave on its inner
face. The Centrosema vexillum forms a narrow "V-shaped" platform with
the lateral edges reflexed outward. The throat is closed as the alae
and carina are in close contact with the vexillum surface. Each half
of the vexillum is convex on its inner face, weakly so near the base
where the halves join and strongly so near the apex.
The size of the vexillum correlates with the flower size discussed
previously. The length of the claw is typically 4-8 mm in small
flowers (except C^ heterophylla, C. biflora and C. australis, 1-3 mm)
and medium-sized flowers of subgenera Neurocarpum and Clitoria.
Medium-sized flowers of the subgenus Bractearia and some large-sized
flowers are 8-12 mm with some large flowers possessing claws of
variable lengths to 20 mm (C. amazonum, C. i a vitensis, C. densiflora).
The claws are typically 1-5 (8) mm in the subgenus Clitori a, 4-12 mm in
the subgenus Neurocarpum (except C^. densiflora, 16-20 mm, and
C. australis, 2-3 mm), and variable in the bubgenus Bractearia.

123
The pubescence is of very dense, short, appressed trichomes in
half the species of subgenera Bractearia and Neurocarpum. In
Bractearia, the silky appearance of the vexillum is associated with
species exhibiting arboreal and tall, erect shrub habits. In
Neurocarpum, no pattern is evident. The second typical pubescent type
consists of uncinate trichomes with a few, scattered, macrotrichomes.
This type of pubescence is generally associated with the liana species
of subgenus Bractearia. There is no evident pattern in subgenus
Neurocarpum, although all species of the subgenus Clitori a possess this
pubescent type.
A1 ae
The alae are oblong to oblong-falcate, spatulate, shorter than the
vexillum, long-clawed, adherent in the middle of its blade to the
carina, and extending beyond the carina.
The size of the blade, claw, and distance the alae extend beyond
the carina are highly variable. No correlation with subgeneric lines
nor with flower size can be drawn. The closest pattern observed was
that the alae extended past the carina: 3-6 mm in small flowers
(except C. cordobensis, 8-9 mm); 5-8 mm in medium-sized flowers (except
C. brachystegia, American members of mariana, C. ternatea, and
C. lasciva, 6-12 mm; and in C. 1eptostachya, 3-4 mm); and 7-13 mm in
large flowers (except C_^ pendens, 4-6 mm, and epetiolata, 13-15 mm).
The alae blade of most species is 12-30 mm long with variable
widths. Six species have shorter blades, 7 12 mm, while only
C. plumosa has an elongated blade (30-35 mm). The blade width forms a
continuous gradient, although most species of subgenus Bractearia have
a width of 5-12 mm while most species of subgenus Neurocarpum have a

124
width of 3-8 mm. The claws are variable, although most species of
subgenus Bractearia have a claw length of 10-25 mm (C. amazonum,
26-33 mm), whereas those species of subgenera Neurocarpum and Clitoria
are 8-15 mm.
Carina
The carina is subfalcate to falcate, incurved, acute, shorter than
the alae to which it adheres, and long-clawed.
The size of the carina blade and claw form a pattern associated
with flower size, not subgeneric lines. The measurements form a
continuum with a distinct break between large flowers and small
flowers. Medium-sized flowers overlap the lower dimensions of large
flowers and larger dimensions of the small flowers. The carina blade
is 10-20 mm long in large flowers, 5-10 mm long in small flowers
(except C. brachycalyx, 10-14 mm), and 7-15 mm long in medium-sized
flowers. The carina blade is 4-8 mm wide in large flowers, 2-4 mm wide
in small flowers (4-5 mm in C. nervosa, C. glaberrima and C.
brachycalyx) and 3-6 mm wide in medium-sized flowers. The claw is
2-4 cm in large flowers, 1-2 cm in small flowers, and 1-2 cm in
medium-sized flowers of subgenera Neurocarpum and Clitoria and 2-3 cm
in medium-sized flowers of subgenus Bractearia (except C. fairchildiana,
1.5-2.5 cm).
Androecium
The androecium is diadelphous, persistent in fruit, with the
vexillary stamen free or occasionally coherent below and easily
separated. The stamina! tube is glabrous, straight to weakly
arcuate and incurved at tip. The free filaments are filiform. The
anthers are uniform and hasifixed.

125
The stamens are similar to those of Hie other genera in the
Glycineae. A unique stamen arrangement occurs in the double-flowered
variety of ternatea. The stamens are all free, a primitive
condition rare in the Fabaceae, or the stamens are subpolyadelphus,
with some stamens free and others coherent, in 1-3 bundles of 2-3
stamens near their base.
The length of the stamina! tube is highly variable and correlates
somewhat with flower size. Large flowers have a stamina! tube 3-5 cm
long, whereas small flowers have a tube 1-2.5 cm long. Medium-sized
flowers have variable tube lengths, but those species in subgenus
Bractearia are 2.5-4 cm long, those species in the subgenus
Neurocarpum slightly shorter at 2-3 cm, and those species in subgenus
Clitoria are 1-2 cm long. In general, species of subgenus Bractearia
are mostly 25-42 mm long, whereas those ir, Neurocarpum are mostly
14-23 mm, with each subgenus having a few species in the size range of
the other subgenus. Subgenus Clitoria has tubes 10-20 mm long with
C. heterophylla possessing a short tube ot 8-10 mm. Extremely long
tubes are found in C^_ pozuzoensis, C. pendens, C. obidensis and
C. plumosa (40-55 mm).
The length of the free filaments forms a continuum from 1-7 mm and
shows no patterns in the variation.
The anthers are consistent in size, typically 1-2 mm long and
0.5-0.8 mm wide. Six species in subgenus bractearia possess anthers
2-3 mm long, all lianas except the shrub C amazonum. Several species
of each subgenus have anthers 0.7-1.0 mm wide.
Uncinate trichomes appear near the apex of the staminal tube in
a few species of subgenus Bractearia. The trichomes occur in

126
C. nervosa, andrei, C arborescens, and in some members of
C. dendrina and C_. brachycalyx.
Gynoecium
The gynoecium is enclosed within the stamina! Lube. The ovary
is stipitate, linear, compressed, many-ovuled, and densely pubescent.
The style is elongated, twisted, geniculate near the tip, dilated above
the geniculate point, and conspicuously bearded lengthwise from near
ovary to the stigma. The style base is persistent in fruit. The
gynophore is subsessile to long, usually densely uncinate. The stigma
is capitate, often short-pubescent at its base.
The key characteristic that segregates Clitoria from all other
genera of the Glycineae is the geniculate, bearded style. Sometimes
the style of Centrosema is described as bearded at the apex, but the
trichomes are short and congregated below the stigma, never found
through the entire length of the style. In Clitoria, the trichomes are
short to long, dense, originating mostly from the margins of the
flattened style, and extending the entire length. The "beard" has
often shorter', less conspicuous trichomes below the point of
geniculation.
The ovary has two distinct pubescent types. Some are densely
uncinate, becoming scattered on the legume valves or glabrous. Other
ovaries may bear uncinate hairs, but are masked by short to long,
densely appressed, white to yellowish-white trichomes. These hairs may
be retained or lost in fruit. The gynophore has a pubescence similar
to the ovary near its apex and is uncinate to glabrate at its base.
Ttie length of the gynophore fits no distinct pattern of variation,
although species in the .ubgenus Cl itoria usually have a subsessile

127
ovary, the gynophore being 1-1.5 mm long. Those gynophores in the
other species are 1-8 mm with the gynophore elongated in C. pendens
(8-13 mm).
Ovary variation correlates along subgeneric lines. Those species
in the subgenus Bractearia have ovaries 10-20 mm long, whereas those
species in the subgenera Neurocarpum and Cl itori a have ovaries 5-9 mm
long (except (h_ ternatea with 10-13 mm). The ovary is elongated in
C. amazonum (28-32 mm long). The width of the ovary is usually 1-2 mm
wide in subgenera Bractearia and Clitoria, whereas the ovary in
subgenus Neurocarpum is 0.7-1.0 mm wide.
Variation in style length does not seem to correlate with
subgeneric lines nor flower size. The majority of species have styles
10-21 mm long, while several species in the subgenera Bractearia and
Neurocarpum have styles to 35 mm long. The style length above the
geniculation point varies from 4-14 mm, the variation seemingly not
forming any pattern.
Observations
The chasmogamous flower is often glued to the herbarium sheet.
It is usually mounted on the herbarium sheet in the complicate (i.e.,
folded) position. Few flowers accompany collections of species in the
subgenera Clitoria and Neurocarpum in packets, as is the case more
frequently in subgenus Bractearia. Thus internal examination and
measurement of the flower structures become difficult without
destruction of the flower. This author offers the following suggestions
for the examination of dried flowers with minimal damage to the
material. First, avoid opening the flower at all, and make measure¬
ments from the outside. The calyx persists in fruit and can be

128
measured at any time, although it is best to measure those of subgenus
Clitori a in flower as the subpellucid calyx shrinks in fruit. Stamina!
tubes are best measured in fruit as they are persistent, project
above the calyx, and shrink slightly. In early fruiting stages, the
petals are often absent and one can find the filaments still attached
and occasionally with an anther. The base of the stamina 1 tube (and
claw attachments for the petals) can be determined from the outside.
The abrupt widening of the calyx at its base indicates the internal
disc to which these structures are attached. The carina blade often
forms an impression in the vexillum which can be easily observed and
measured. The style can be measured in specimens with immatured
fruits. Second, when necessary, use a needle probe and carefully tease
the vexillum halves apart. Pick another flower when resistance is met,
as pressure applied against the resistance will tear the vexillum. By
using forceps or a probe to gently hold the vexillum halves apart, one
can measure the alae and carina blades, and measure the claws down to
the swollen calyx base (-disc area). If necessary, one can gently
tease the carina blades apart from the base to examine the androecium
and gynoecium, although one must dissect the calyx to observe the ovary
and gynophore. Third, place the flower contents into protective
packets when dissected, and affix the packets to the herbarium sheets.
Glasseine envelopes were used in this study for dissected floral parts,
for both the flowers this author dissected and for the loose floral
parts this author found in packets. Dissection need not occur,
however, for most floral structures; the open vexillum can be reclosed
with a slight pressure, and the parts preserved. This author has tried
an alternative method suggested by others on locally collected, pressed

129
Clitoria flowers. This method entailed the removal of the flower from
the specimen, placing it in water to moisten the petals, partially
drying the flower to remove most excess water, then dissection of the
floral parts. This method is not recommended as it leads to more
destruction than the method this author suggested above.
A number of flower structures showed patterns of variation along
subgeneric lines and correlated with flower size. The variations are
summarized in Tables 4 and 5. Those characters which correlate with
flower size are best used to segregate small groups of species. Those
characters which correlate with the subgenera are best used as
supportive data since measurements often overlap or the character could
be misinterpreted. These characters may be useful in segregating
groups of species into sections. In the descriptions, measurements of
those structures that are more difficult to view without flower
dissection (e.g., ovary, gynophore), may be less reliable as fewer
measurements were made, especially in species of one to few
collections. However, measurements were highly consistent in those
species with larger collections of flowering material.
Flower color is somewhat unreliable as only a small number of
species descriptions include references to color, and these typically
reflect the color observed in the field which can be interpreted
differently by each collector. Pale blue, lilac, and rose to pink
shades dried to various shades, sometimes appearing blue to lilac-
shaded, and at other times appearing a darkened dull-yellow to
orangish-yel low. VJhite flowers appear unpigmented in the dried state,
or sometimes darkened, and sometimes they dry between a pale to a
bright yellow shade.

Table 4. Comparison of flower structures by subgenera. Variation patterns of these characters
correlate along subgeneric lines. All characters refer to chasmogamous flowers with the
exception of the character "cleistogamy."
CHARACTER
BRACTEARIA
CLITORIA
NEUR0CARPUM
Calyx nerves
Numerous
10-nerved
10-nerved (numerous)
Calyx texture
Subcoriaceous
Subpel1ucid
Carti1aginous
Calyx shrinkage
Retains size in fruit
Shrink in fruit
Retains size in fruit
Calyx lobes
Generally 3-7 mm
(7-12)(13-18)(20-27)
3-7 mm or 7-12 mm
Generally 7-12 mm
(3-7)
Calyx tube
width throat
Generally 7-13 mm
(1 spp. = 5-8 mm)
4-8 mm and 7-13 mm
Generally 4-8 mm
(3 spp. = 6-11 mm)
Alae blade width
Generally 5-12 mm
7-13 mm or 3-6 mm
Generally 3-8 mm
Alae claw length
10-25 (33) mm
8-15 mm
8-15 mm
Staminal tube
2.5-4 cm
1-2 cm
1.4-2.3 mm
Ovary length
10-20 mm
5-9 mm (1 sp. = 10-13)
5-9 mm
Cleistogamy
Absent
Absent
Present or Absent

Table 5. Comparison of flower structures by flower size. Variation patterns of these characters cut
across subgeneric lines except where indicated. (B = Bractearia; N = Neurocarpum;
C = Clitoria).
CHARACTER
LARGE FLOWERS
MEDIUM-SIZED FLOWERS
SMALL FLOWERS
Subgeneric size:
Bractearia (B)
6-8 (9.5) cm
4-6 cm
2-4 cm
Neurocarpum (N)
5.5-7.5 (8) cm
3.5-5.5 cm
2-3.5 (4) cm
Clitoria (C)
Absent
(1 sp.-lasciva)
3.5-5.5 cm
2-3.5 (4) cm
Composite size
5.5-8 (9.5) cm
3.5-6 cm
2-4 cm
Flower color
Blue, violet, pink, lilac
(rarely white)
Blue, violet, pink, lilac
(rarely white)
Whi te
(rarely pigmented)
Calyx tube length
15-25 mm
13-20 mm (B)
7-1C mm (C,N)
4-8 mm
Total calyx length
2-4 (5) cm
1-3 cm
1-3 cm
(iube + lobes)
Calyx tube width
8-13 mm
8-13 mm or 5-8 mm
3-6 mm
(at throat)
Vexi Hum cl aw
8-12 (20) mm
4-8 mm (C,N)
8-12 mm (B)
4-8 mm (1-3)
Carina length
10-20 mm
7-15 mm
5-10 mm
Carina width
4-8 mm
3-6 mm
2-4 mm

Table 5 - continued
CHARACTER
LARGE FLOWERS
MEDIUM-SIZED FLOWERS
SMALL FLOWERS
Carina claw
2-4 cm
1-2 cm (C,N)
1-2 cm
1.5-2.5 cm (B)
Stamina! tube
3-5 cm
2.5-4 cm (B); 2-3 cm (N)
1-2.5 cm
1-2 cm (C)

133
Ciéistogamous F1 owers
Ciéistogamous flowers are small, inconspicuous, apetalous, or
sometimes retaining minute, translucent vestiges within the calyx.
These flowers are self-pollinated. Cleistogamous flowers are rarely
reported for the genus, and are known in literature (Harms, 1907;
Burkart, 1941) from only five species. Ample examination of herbarium
collections has revealed that such flowers are confined to eleven
species .in the subgenus Neurocarpum. These species occur in dry
habitats of the Americas. Two of these species (C. falcata and
C. laurifolia) have been introduced into and naturalized in Africa and
Southeast Asia-Indonesia. Cleistogamous flowers which occur in the
American variety of mariana are absent from the Asian variety.
Cleistogamous flowers may occur on a plant at the same time as
chasmogamous flowers (chasmocleistogamy), but usually the plant
specimens are collected with only the cleistogamous flowers, and these
are usually in various stages of fruiting. A generalized cleistogamous
flower is illustrated in Figure 6. Cleistogamous flowering and
fruiting structures are summarized by species in Table 6.
In general, the inflorescence, bracts, bracteoles, fruits, and
seeds are similar to the chasmogamous structures, although sometimes
of smaller sizes. Their morphology will not be reiterated here. The
calyx, corolla, androecium and gynoecium exhibit marked differences and
will be included.
Calyx
The calyx is much smaller than the chasmogamous calyx, narrowly
infundibular-tubular, persistent in fruit, conspicuously nerved, and

r
Figure 6. Generalized cleistogamous flower, (a) cleistogamous flower,
longitudinal section, x 5; (b) gynoecium, x 8; (c) androecium,
x 3; (d) cleistogamous flower, x 3; (e) cleistogamous flower
with juvenile fruit, x 3; (f) cleistogamous flower with mature
fruit, x 3; (g) translucent vestigial petals, x 15.

135

Table 6. Comparison of cleistogamous
SPECIES
BRACTS
BRACTEOLES
CALYX
TUBE
polystachya
2-3 mm
3-4
mm
4-6 mm
mexicana
2-3 mm
3-5
mm
4-7 mm
monticola
2 mm
o
J
mm
4-5 mm
mariana (Amer.)
2-3 mm
3-5
mm
4-5 mm
fragrans
2 mm
2-3
mm
3-4 mm
cordobensis
4-5 mm
4-5
mm
5-6 mm
falcata
3-4 mm
4-7
mm
5-8 mm
guianensis
3-7 mm
3-7
mm
4-8 mm
v. macrocleistogamous
5-6 mm
5-7
mm
10-11 mm
eoet.iolata
4-5 mm
5-7
mm
6-9 mm
nana
1.5-3 mm
4-5
mm
6-7 mm
laurifolia
2-4 mm
3-5
mm
5-7 mm
structures by species.
CALYX
LOBES
GYNOPHORE OVARY
STYLE
2-3
mm
1
mm
6
mm
3-4
mm
2-4
mm
1
mm
5
mm
3-5
mm
5-2
mm
1
mm
D
mm
4
mm
2-3
mm
1
mm
4-5
mm
4-6
mm
2-3
mm
1-2
mm
4
mm
5-6
mm
2-3
mm
1.5-2
mm
3-4
mm
3-4
mm
4-7
mm
1
mm
4-5
mm
3-4
mm
3-4
mm
2-3
mm
4-5
mm
5-6
mm
4-6
mm
?
?
?
4-6
mm
2
mm
6
mm
5
mm
5-6
mm
3
mm
5
mm
4
mm
2-4
mm
1.5-2
mm
3-4
mm
3-4
mm
OJ

Table 6
SPECIES
STAMINAL
TUBE
FREE
FILAMENTS
polystachya
4-5 mm
1.5-2 mm
mexicana
4-5 mm
1-1.5 mm
monticol a
4-5 mm
0.5-1 mm
mariana (Amer.)
0.1 mm
1 - 3 mm
fraarans
0.1 mm
1-2 mm
cordooensis
3-4 mm
1-1.5 mm
falcata
0.5 mm
3-4 mm
guianensis
G. 5 mm
3-4 mn i
v. macrocleistoqamous
0. 5 mm
4 mm
epetiolata
0.5 mm
2-4 mm
nana
0.5 mm
2-3 mm
1aurifolia
0.5-1.5 mm
1.5-2 mm
continued
STIPE
COSTA
6-7
mm
-
6-7
mm
-
6-7
mm
-
5-10
mm
-
9-14
mm
-
8-16
mm
-
6-12
mm
+
8-14
mm
+ (-)
16-18
mm
+
8-12
mm
-
7-9
mm
-
6-13
mm
+
LEGUME LEGUME
LENGTH WIDTH
2.5-4
cm
7-8
mm
2-4
cm
6-7
mm
2.5-3
cm
5-7
mm
3-4
cm
6-8
mm
2-4
cm
6-8
mm
2-4.5
cm
3-6
mm
2.5-4.5
cm
7-11
mm
2.5-5
cm
7-11
mm
4-4.5
cm
8-11
mm
3-6
cm
8-11
mm
3.5-4.5
cm
7-9
mm
2.5-4.5
cm
8-11
mm
co
^0

138
with its apex five-lobed, shorter than the tube, upper lobes
subconnate, lowermost lobe narrower, and subequal to slightly longer.
The calyx has the appearance of a narrow tube, although the width at
the throat (2-4 mm) usually is double the basal width (1-2 mm) where
the internal disc is located and to which the androecium and petals are
attached.
The calyx is inconspicuous, the tube usually 4-8 mm long, the
lobes 2-6 mm long. This is quite in contrast with the chasmogamous
flowers of these species which have tubes of 8-23 mm long and lobes of
4-15 mm.
Coro! 1 a_
The petals are variable, often absent, sometimes present as
vestigial remnants, whitish-transparent, minute, more or less clawed.
Petal number varies from 1-5; petals usually are 1 mm long, sometimes
to 2-3 mm, erect, similar in shape, with a short claw and oblong blade.
Petals may be present or absent in flowers on the saíne plant.
Androecium
The filaments are subsessile or connate in a short tube,
diadelphous, and enclosed within the calyx tube. The anthers are
similar to those of chasmogamous flowers. The staminal tube has more
variation than other cleistogamous structures. Two species appear to
have free filaments as the basal tube is highly reduced, circa 0.1 mm
long. Five species have subsessile tubes, circa 0.5 mm long, with one
of these species having a tube 0.5-1.5 mm long. Four species have
conspicuous staminal tubes, although short, at 3-5 mm. The staminal
tubes and filaments are persistent in fruit, but hidden by the calyx

139
tube. The anthers often detach and adhere to the ovary as it is
developing into the fruit.
Gynoecium
The ovary is enclosed in the calyx tube and similar to the
chasmogamous ovary in pubescence and width, but shorter, stipitate to
subsessile, with the style abruptly bent back toward the ovary at its
base. The stigma is often in contact with the anthers. Gynophores are
usually 1-2 (3) mm. Ovaries are 3-6 mm long, 0.7-1 mm wide. Styles
are 3-6 mm long, but may be longer or shorter than the ovary, even
though the range of variation agrees with the range of variation for
the ovary.
Observations
Cleistogamous flowers are known from collections that are
predominantly in fruit or with chasmogamous flowers. Because of their
inconspicuous size and the general lack of collections with only
cleistogamous flowers, one can presume that in most cases the collector
was unaware of the presence of the cleistogamous flowers and collected
the plant because of its showy chasmogamous flowers or its fruit. Some
closely related species (e.g., C_^ tri flora, C. simpl icifol ia) lack
collections with cleistogamous flowers. Plants bearing cleistogamous
flowers should be searched for and reported in these related species.
Cleistogamous structures are unreliable characters for segregation
as the range of variation is too similar to those of other species.
Where there are some distinctions, these may be used as weak supportive
characters for segregation. An exception would be the length of the
stamina! tube, which seemingly is consistent among the flowers

140
examined within a species, and whose length correlated into three
distinct groups (subsessile=0.1 mm; short=G.5-1.5 mm; long=3-5 mm).
It is important to note that in several species (e.g., C.
polystachya, C. mexicana, C. monti col a, C. cordobens is, C. nana,
C. epetiolata), the quantity of cleistogamous collections is low,
limiting the number of c1eistogamous flowers examined. In addition,
some structures (e.g., gynophore, ovary) were rarely seen. The range
of variation for some structures is thus questionable and subject to
expansion with further data.
Legume
The legume is stipitate, rarely subsessile, linear, compressed and
ecostate, more or less sometimes raised around the seeds, or less
commonly turgid and costate, becoming tetragonous. The valves are flat
or convex, slightly thickened on the upper or on both sutures, often
bearing uncinate trichomes, and usually beaked by the persistent style
base. Dehiscence occurs by a splitting along the sutures followed by a
spiral twisting of the valves. Legumes are similar to those of other
species in the Glycineae with only the genus Centrosema (two-nerved
near the margins) and some members of the genus Clitori a (one lateral
nerve) costate.
The legume is one of the major structures used tor segregation of
the species into subgenera. Three basic morphological types occur with
a fourth intermediate type found in two subgenera. Fruits of subgenus
Bractearia are long-stipitate, compressed, ecostate, with the valves
thickened on the sutures, flat, and very large in size. In some
members, the pod becomes weakly turgid around the seels and depressed

141
between the seeds. These fruits are referred to as 1omentaceous"
in this study, a term based upon Dayton and Jackson (1928). Fruits of
subgenus Clitoria are subsessile, compressed, ecostate, with the valves
flat, slightly thickened on the sutures, intermediate in size, and its
base is enclosed within the calyx. Fruits of subgenus Neurocarpum
are short-stipitate, compressed and ecostate, or costate and turgid,
subtetragonous at maturity. Valves are convex, slightly thickened on
the sutures, with the base elevated above the calyx or enclosed within
it. Some members have lomentaceous legumes. Fruit and seed characters
are summarized by subgenus in Table 7. A generalized fruit and
representative seeds are illustrated in Figure 7.
Sti pe
The stipe is mostly enclosed within the persistent calyx,
longitudinally striated, more or less uncinate pubescent, sometimes
bearing macroscopic trichomes near its apex, and persistent after fruit
dehiscence and abortion of the valves. It is usually slender, lignose,
abruptly expanding at the apex to form a shallow concavity upon which
the legume base sits.
Stipe length is variable. In the subgenus Bractearia, the stipe
is elongated, exerted from the calyx, typically 12-33 mm long. A
shortened stipe (10-12 nun) occurs in C_^ glaberrima. Highly elongated
stipes occur in (_ amazonum (25-50 mm), brachycalyx (28-45 mm),
C. pozuzoensis (34-35 mm), and C. flexuosa (39-40 mm). The latter two
1. Lomentaceous legume: a type of legume which superficially
resembles lomeuts iri that the valve is contracted between the seeds
transversely tu seemingly appear to be a segmented fruit, but whose
dehiscence occurs by a longitudinal splitting along the sutures,
typical of a legume.

Table 7. Comparison of fruits and seeds by subgenera.
CHARACTER
BRACTEARIA
CLIT0RIA
NEUR0CARPUM
Sti pe
Long; 12-33 (50) mm
Subsessile; 1-4 mm
Short; 4-14 mm
Valve l^noth
(7) 10-30 cm
7-11 cm or 2-5 cm
3-6 (7) cm
Valve width
10-40 mm
(4) 8-15 mm
6-11 (13) mm
Valve curvature
Flat
Flat
Convex
Valve nerves
Ecostate
Ecostate
Ecostate or costate
Valve base
Exerted
Enclosed by calyx
Exerted or enclosed
Dehiscence
1/4-1 (2-4)
1-2 turns
1/4-2 turns
Lomentaceous legume
+ or -
-
+ or -
C^ec! coat
Smooth
Smooth
Vi sc i d
Seed shape
Orbicular
Subreniform
Subreniform (Orbicular; ovoid)
Seed length
7-13 (17) mm
2-4 min
3-5 (6) mm
Seed width
7-13 (16) mm
4-6 (8) mm
3-6 mm
Seed thickness
Flat (1-3 mm)
Turgid centers (3-5 mm)
Flat (1-2 mm)
Turgid (3-4 mm)
Seed germination
Epigean
Epigean
Hypogean

Figure 7. Generalized legumes and seeds. Subgenus Bractearia: (a) flat
legume, with cross section illustrated; (b) subloinentaceous
legume, flat, depressed between seeds, portion of longitudinal
section illustrated; (c) three views of typical seed.
Subgenus Clitoria: (d) legumes, with cross section illustrated;
(e) three views of typical seed. Subgenus Neurocarpum:
(f) ecostate, lomentaceous legume, with portion of
longitudinal section illustrated; (g) costate, subtetragonus
legume, with cross section of turgid legume illustrated;
(h-j) three views of different types of viscid seeds occurring
within subgenus. Generalized longitudinal section of legume;
(k) view of inner face with seeds attached; (1) length of
seed as defined in this study; (w) width of seed as defined
in this study.

144
c
L~j.

145
species have rarely been collected in fruit;, and may have greater range
of variation. Within the typical range of length, species tend to be
either 12-22 mm long or 23-33 mm. In the subgenus Clitori a, the stipe
is subsessile, 1-2 mm long. Stipes are variable in length in subgenus
Neurocarpum, but typically 4-14 mm. Species tend to be either 4-8 mm
long or 6-12 (14) mm. Elongated stipes occur in the American variety
of C. mariana (12-17 mm) and C. fragrans (15-21 mm).
The stipes are typically 1-3 mm thick and rapidly expanding to
3-7 mm wide at the apex. In the very large fruits of some members of
the subgenus Bractearia the stipe is typically 3-5 mm thick and
expanded to 6-10 mm wide at the apex, as in arbórea and C. nervosa.
Valves
The size of the legume valves is variable. In general, fruits of
the subgenus Bractearia are large, those oi the subgenus Neurocarpum
are small, and those of the subgenus Clitot ia are of intermediate size.
The fruits of the subgenus Bractearia are 10-30 cm long and 1-4 cm
wide. Shorter fruits occur in C. coriácea (6-12 cm), C. obi dens is
(7-9 cm) and C^ kaieteurensis (8-13 cm). The fruits are typically
1-2 cm wide with those 2-4 cm wide associated with the primitive
bracteoles, such as C^ arbórea (2.3-4 cm), C. fairchildiana (2.2-3.3 cm)
and C. juinensis (2.5-3.2 cm). Often these latter fruits have
thickened sutures, almost thick winged-like. The fruits of C.
brachycalyx (1.6-3 cm) are not associated with the primitive
bracteoles, an exception to this generalization. The fruits of
subgenus Neurocarpum are typically small, with a length of 3-6 (7) cm
and a width of 6-11 (13) mm. Most of the species have fruits that span
the measurement ranges, whereas fruits with a more restricted range of

146
variation are found in species of a few collections. The fruits of
several species in both of these subgenera are unknown, but their size
is expected to correlate with those species of close relationship
(e.g., fruits of andrei should correlate with those of C_^ juinensis
when collected). The fruits of the subgenus Clitori a are intermediate
in size. Longer fruits of 7-11 cm belong to lasciva and ternatea,
with the other species correlating with Neurocarpurn species. The width
of the valves agrees with Neurocarpum species except for C. kaessneri
(1-1.5 cm).
Valves are flat in the subgenera Bractearia and Clitori a whereas
those of Neurocarpum are convex. In fruits associated with the
primitive bracteoles, valves are thick coriaceous with prominently
raised striations. They are less prominent in other species of the
subgenus Bractearia and in the subgenus Neurocarpum. Veins in the
subgenus Clitori a are slightly dark-colored, with the striations
impressed to subimpressed, not prominently raised. These striations
are densely packed, often forming transverse diagonal lines with some
connecting striations.
In some species of the subgenera Bractearia and Neurocarpum,
valves become conspicuously raised around the seeds and depressed
between them, sublomentaceous to lomentaceous. The surface is
flattened in the former subgenus and convex in the latter subgenus.
Sublomentaceous fruits are associated with the short calyx tube in the
subgenus Bractearia (section Brachycalyx). In Neurocarpum species,
this fruit type is associated with the Mexican, United States, and
Southeast Asian members.

147
Some species of the subgenus Neurocarpum develop fruits which have
a conspicuously raised, prominent, lateral nerve. Mature fruits become
subtetragonous. This unique character is rare in the Fabaceae.
Costate fruits were historically considered worthy of generic
recognition, thus this one character led to the establishment of the
genera Marti a and Neurocarpum as segregated from Clitoria. Bentham
(1858) noted a breakdown in this character, and correctly combined
these two genera with Clitori a. The costate fruit is prominent in a
number of South American, Central American, and Caribbean members.
A number of specimens, particularly in the species C. guianensis and
C. falcata have incomplete nerves. The nerve when completely formed,
extends one-third to three-fourths the length of the fruit. Both
costate and ecostate fruits may occur on the same plant. An incomplete
costa is prominently raised at one end, and rapidly submerges into the
surface near the other end. Costate fruits are sometimes mistaken for
those fruits found in the genus Canavalia (cf. Fantz, 1976, and
C. javitensis, p. 385), but this genus can easily be distinguished from
Clitoria by its nonresupinate flowers and unique calyx (cf. Sauer,
1964).
The dehiscence of the fruits is variable in the number of spiral
twists that occur. In general, most species have one-fourth of a turn
to two spiral turns with greater twisting in C^ 1eptostachya (2.5-3
turns) and C. brachystegia (3-4 turns). In several species, no
dehisced fruits were observed. Those species with known fruits were
evenly split between those of 0.25-1 turn and those of 1-2 turns.
This character is more descriptive than diagnostic.

148
Fruits from cleistogamous flowers occur only in the subgenus
Neurocarpum. They correlate with the fruits from chasmogamous flowers
except that the size is sometimes smaller.
Observations
Althougti a number of species have yet to be collected in fruit,
the fruits and seeds are major diagnostic characters for the
segregation of the subgenera. Leaves, habit, inflorescences, seed
germination and cytology all provide supportive data for this
segregation. Those species without known fruits can be placed
subgenerically largely by the supportive data.
One has to be particularly careful of the fruits of liana species.
A number of specimens had CVitoria flowers with non-Clitoria fruits.
The fruits were nearly always separate. Those connected were generally
leafless branches, but inflorescence, bracteoles, calyx, and other
characters distinguished them from Clitori a. It is presumed that the
collector inadvertently collected two lianas thinking that they were
the same plant, or picked the fruit off the ground with reason to
believe that they dropped from the liana. Mixed collections with
Clitori a flowering branches and non-Clitori a fruiting specimens were
conspicuously frequent in Suriname collections.
Seeds
Seeds are dark reddish-brown to black, orbicular to subreniform,
sometimes oblong, compressed or thickened, glabrous. The seed coat is
smooth or viscid. Seeds are somewhat variable, but reliable as
diagnostic characters for subgeneric delineation. Seeds of the
subgenus Bractearia are large, smooth, orbicular to oblong, flat to

149
lenticular. Seeds of the subgenus Clitori a are small, smooth,
subreniform, flattened. Seeds of the subgenus Neurocarpum are small,
viscid, orbicular-oblong to subreniform, thickened.
Seed color is variable in a number of species. Both reddish-brown
and black seeds are found in the same specimen. Sometimes darkened
reddish-brown seeds with black spotting to blotching occur. Some seeds
are reddish-black. Where both colors occur in a species, the black
seeds are usually larger and seemingly more mature. This suggests a
continuum where the seeds are brownish, darkening upon maturity and
appearing black. Seed color is a descriptive character, but not
reliable as a diagnostic character.
Seed size is reliable along subgeneric lines, large seeds occur
in the subgenus Bracteatia, whereas small seeds are found in the
subgenera Clitori a and Neurocarpum. Since the longer axis can occur
in two directions, measurements taken and labeled "length" were in the
same direction as the funiculus, whereas those labeled "width" were
perpendicular to the funiculus. Seeds in the subgenus Bractearia were
7-13 (17) mm long and 7-13 (16) mm wide. Most seeds were orbicular to
subquadrate with rounded corners, flat to lenticularly thickened, with
both dimensions subequal. Orbicular-oblong to oblong seeds occurred in
several species with the length being the larger dimension. The
smallest seeds (6-8 mm, orbicular) occurred in C. sagotii and
C. obidensis. Seeds with larger dimensions (length 11-17 mm; width
10-15 mm) occurred in C. arbórea, C. fa i rclii 1 di ana and C. pozuzoensis
with the latter oblong (width 10-12 mm). Seeds in the subgenus
Clitori a are small, 2-4 (6) mm long, (3) 4-8 mm wide. All are
subreniform with the width being the longest axis. The widest seeds

150
(6-8 mm) occur in C. lasciva and C. kaessncri with the latter having
the longer seeds (6 mm). The smallest seeds belong io C. heterophylla
(length 2-3 mm; width 3-5 mm). The seeds of subgenus Neurocarpum are
small, thickened, subreniform with the width slightly longer than the
length. They are typically 3-5 mm long and 4-6 mm wide. The thickness
is typically 3 mm, almost as large as the other dimensions, giving the
seeds a cuboidal-rectangular shape. In a few species (e.g.,
C. guianensis, C. epetiolata), the seeds are orbicular-oblong with the
length the slightly longer axis (length 3-5.5 mm; width 3-4 mm), and
with the thickness appear globular. In the species stipularis and
C. densiflora, the seeds appear ovoid as the length is longer and the
width is slightly longer than the thickness (length 4-6 mm; width
3.5-4.5 mm; thickness 3-4 min).
The seed coat is viscid in subgenus Neurocarpum and smooth in the
subgenera Bractearia and Clitoria. The sticky coat is conspicuous when
the seed adheres to an herbarium sheet, packet, or to plant parts.
When the sticky fluid solidifies, it produces a shinier, glassy coat
around the seed, similar to the solidification of a caramel apple's
sugary coat. The smooth seed coat is dull, non-glassy, and scratch
marks are left when it is scraped with a needle probe. The viscid coat
around the seed will shatter into various sized plates around the site
of probing. The sticky nature of the coat will persist for a number of
years as specimens from the early 19501s (ca 25 years) were observed to
still be sticky. Occasionally older seeds are found in this condition.
The number of seeds that occur per legume is variable, although
most species produce fruits with 4-12 seeds. A few species produce
few-seeded fruits, such as C. biflora (2-4 seeds) and C. kaessneri

151
(1-3 Js) of the subgenus Cl itoria and C. polystachya (1-5 seeds),
C. hanceana (2-6 seeds), and the Asia variety of C. mariana (1-4 seeds)
of subgenus Neurocarpum. Most species in these two subgenera produce
4-8 seeds per fruit with C. stipularis producing more seeds (7-11
seeds). In the subgenus Bractearia, the typical fruit produces 7-12
seeds. Two species with many-seeded fruits are C. amazonum (10-17
seeds) and C. pozuzoensis (13-14 seeds). There are several species
which average 5-8 seeds per fruit. These are all liana species with
the exception of C^ nervosa.
Seed Germination
Seed germination has been inadequately studied, with only the
types for six species known. From the few reported species, Rizzini
(1963) noted that epigean germination occurred in the subgenera
Cl i tori a (C. ternatea) and Bractearia (C. fai rchi 1 di cana reported as
C. racemosa Benth.; C. amazonum) whereas hypogean germination occurred
in the subgenus Neurocarpum (C^_ 1aurifoli a; C. guianensis; C. falcata
reported as C^ rubiginosa).
Observations
Seeds are unknown for a number of species because the fruits are
unknown, or because fruits were collected but seeds were immature or
the fruits were unopened. A viscid coat indicates Neurocarpum species,
large size indicates Bractearia species, with small, nonviscid seeds in
Clitoria species.
A number of seeds exhibited to varying degrees, one or more
depressions in the seed face of various sizes. This condition was not
consistent in a species, and occurred typically in immature seeds. It

152
is not reliable as a diagnostic character and may be an artifact of the
seed histology.
Summary
A number of characters were examined in order to more fully
describe and compare species. Those characters of diagnostic value
have been noted. Most characters can be used to segregate small
species groups.
Species Descriptions
The description of each species will be slightly long, so key
terms (e.g., leaves, inflorescences, flowers, fruits) will be
underlined and flower type (chasmogamous flowers, cloistogamous
flowers) will be highlighted in capital letters. Descriptions will
vary somewhat for individual species, but each will follow the general
pattern that follows in the generalized description.
Clitori a species. Habit, size, general aspect, general
pubescence. Xylopodium or roots. Stem (branch) shape, pith, diameter,
pubescence, juvenile structures; bark; buds; scars. Leaves pinnation
number, texture; leaflet shape, apex, base, margin, midrib above,
primary nerve number, upper surface, lower surface, length, width.
Petiole shape, pubescence, ratio to rachis, length; rachis length,
internodal segments. Petiolules color, shape, texture, pubescence,
length. Stipule duration, striations, shape, apex, pubescence, length,
width; stipel duration, striations, shape, apex, pubescence, length,
width, terminal versus lateral. Inf 1 orescence type, location, number
per locality, number flowers borne, axis pubescence, length, primary
branches. Peduncle; rachis, internodal length, length primary branches

153
bearing pedicels; pedicel pubescence, length. Bract number, shape,
apex, pubescence, inner bracts, middle bracts, outer bracts. Bracteole
duration, shape, apex, pubescence, length, width, insertion below
calyx. CHASMOGAMOUS FLOWERS size, color, odor. Calyx shape,
pubescence, nerves, tube length, width at base to width at throat;
lobes shape, apex, pubescence, length, width at base, ventral lobe
length. Vexillum color, pubescence outside, shape, length, width,
claw. Alae color, shape, extension beyond carina, blade pubescence,
length, width, claw. Carina color, shape, pubescence, length, width,
claw. Staminal tube pubescence, length, free filaments length;
anthers length, width, connective. Gynophore length, pubescence;
ovary pubescence, length, width; style length, geniculate length;
stigma. CLEISTOGAMOUS FLOWERS size, location. Calyx pubescence, type,
nerves, tube length, width at base to width at throac; lobes shape,
apex, pubescence, length, width base, ventral lobe. Corolla presence,
shape, size. Starninal column length, filament number, length; anthers
length, width. Gynophore pubescence, length; ovary pubescence, costa
presence, length, width; style length; stigma. LEGUME stipitate,
color, shape, pubescence, sutures, costa presence, valves texture,
curvature, length, width, beak length; dehiscence turns; stipe
pubescence, diameter, apex, length. Seeds color, shape, pubescence,
viscidity, length, width, thickness; hilum length, width. Seed
germination type; cotyledons shape, length, width; primary leaves.
Characters Occurr i ng jjn Rel at vve 1 y Few Spec i esy
For most structures there are characteristics limited to a small
number of the species within the genus. These characteristics can be

154
useful in quickly identifying a species. A list of these character¬
istics follows below.
LEAVES: 3-foliate, digitately compound, C. epetiolata; 1-foliate
only, (h_ simp! icifol ia, C. cordiformis, arid occasional stems of
C. kaessneri; upper leaves 3-foliate, lower leaves 1-foliate,
C. guianensis, C. densiflora, C. kaessneri, C. australis, C.
macrophylla, C. hanceana; 5- and 7-foliate, biflora, C. ternatea,
C. heterophylla; 7-, 9- or 11-foliate, lasciva, C. heterophyl1 a.
LEAFLETS: pubescence on upper surface moderately dense, minute,
sandpapery rough texture when finger is rubbed from apex to leaflet
base (i.e., retrosely scabrous), C. dendrina, C. bra<:hycalyx,
C. froesii, hermannii; pubescence on upper surface moderate to
scattered, subappressed macrotrichomes, (h biflora, u ternatea,
C. lasciva; pubescence on upper surface rufus, short to long trichomed,
C. plumosa, C. obidensis; leaflet width less 1 cm, C heterophylla,
(h_ fragrans, some (h ternatea; leaflet length less 2 cm,
heterophyl 1 a; midrib strongly raised on upper surface, C. obidensis,
C. sagotii, C. leptostachya; primary nerves greater than 16, C. nervosa
and occasional leaflets of C. fairchildiana and C. juninensis; leaflets
ceriferous below, C. coriácea; leaflet margin revolute, C. coriácea,
C. tunuhiensis. Leaflets large, 10-25 cm long by 4-9 cm wide, subgenus
Bractearia.
PETIOLE: petiole elongated, 10-25 cm, C. dendrina, C. brachycalyx,
C. brachystegi a, C. froesii, and some C. javitensis (NOTE: species
with petioles commonly less than 10 cm occasionally have some leaves
long-petiolate; the species noted have all mature leaves long-petiolate)
petiole subsessile, 0.2-2 cm long, shorter than the tachis, C. biflora,

155
C. heterophylla, and eight species of subgenus Neurocarpum; petiole
absent, epetiolata.
STIPULE: stipules highly elongate, 20-30 mm, C. plumosa; stipules
elongated, 10-20 mm, kaieteurensis; stipules broad, 4-6 mm,
C. arborescens, C. kaieteurensis, 6-8 mm, C. den si flora, 7-13 min wide
and foliaceous, C^ stipularis.
STIPEL: stipels elongated, 10-20 mm, C. plumosa, and occasionally
to 15 mm long in obidensis, C. macrophylla, C. stipularis; stipels
broad, 2-3 mm wide, stipularis and some kaessneri.
INFLORESCENCE: highly elongated, over 20 cm, C. pendens,
C. leptostachya, C. fairchi1 diana, occasionally some C. arbórea,
C. juninensis; inflorescence paniculate, primary lateral branches over
5 mm long, C_. arbórea, C. ainazonum, C. polystachya, C. monticola,
occasionally C. juninensis; inflorescence conspicuously zigzag,
section FIexuosa, C. woytkowskia, C^ flexuosa, C. potuzoensis;
inflorescence cauliferous, section Caul if1orae; inflorescence a
peduncle wi tli 1-2 flowers at apex, most of subgenera Neurocarpum and
Clitoria.
BRACTS: bracts elongated, 10-16 mm long, C. plumosa, C. flexuosa,
C. stipularis, occasionally 7-11 mm long, C. densiflora, C. woytkowskia,
C. obidensis; bracts broaden, 4-6 mm wide, C. amazonum, C. woytkowskia,
and 6-9 mm wide, C. flexuosa.
BRACTE0LES: bracteoles highly elongated, 25-41 mm long,
C. flexuosa, C. woytkowskia, C. obidensis; bracteoles minute, 1-3 mm
long, C. dendrina, C. brachycalyx, C. brachystegia, C. hermannii,
C. canescens, C. heterophylla, C. leptostachya; bracteoles broad (6 mm
or more wide) and long (10 mm or more long), subequal the calyx and

156
hiding it, sections Bractearia and Flexuosa; bracteoles broad (3-6 mm
wide) and not hiding calyx, C. plumosa, obidensis, bracteoles
inserted 2-5 min below the calyx, C. coriácea, C. sagotii, C. lepto-
stachya, C. tunuhiensis, and some javitensis.
CALYX: calyx tube elongated, 2-4 cm, C. plumosa, C. obidensis,
C. flexuosa, C. woytkowskia, C. pozuzoensis, occasionally C. amazonum
(NOTE: some species have tubes of 1.5-2.3 cm); calyx tube cup-shaped,
C. brachycalyx, C. brachysteqia; calyx pubescence dense, silky,
C. flexuosa, woytkowskia, C. pozuzoensis, C. australis; calyx tube
width at base broad, 8-10 mm wide, plumosa; calyx nearly glabrous,
C. amazonum, C_. moyobambensis, C. glaberrima; calyx lobes elongated,
14-18 mm, C. plumosa, C. densiflora, highly elongate'!, 20-27 mm,
C. obidensis; calyx lobes minute, 1-2 mm, C. brachysi.egia; calyx lobes
short, 2-4 inm, dendrina, C. brachycalyx, C. glaberrima, C. froesii,
C. canescens , C._ hermam i, C. nervosa, polystachy.i, C. monticola,
C. heterophyl la; calyx lobes broad, 5-8 mm, C_. obidensis, C. densiflora
ventral lobe elongated, 15-20 mm, C. densiflora, C. . tipularis, greatly
elongated, 25-35 mm, obidensis, C. plumosa; ventral lobe pubescence
stiff trichomes, laterally, feather-like, C. plumosa
ANDROECIUM: stamens all free or some filaments free with others
connate in 1-3 bundles, terna tea; staminal tube greatly elongated,
4-5.5 cm, plumosa, C. pendens, C. javitensis, C. obidensis,
C. pozuzoensis; staminal tube short, 8-11 mm, C. biflora, C.
heterophylla; staminal tube uncinate pubescent neai apex, C. nervosa,
C. hermannii, C. arborescens, occasionally in C. dendrina, C.
brachycalyx; staminal tube pubescence uncinate near base, C.
kaessneri; anthers large, 2-3 mm long, C. amazonum, 0. flexuosa,

157
pozuzoensls, obidensis, C^_ pendens, C. plumosa f roe si i,
C. canescens, C. hermannii; anther connective apiculite, C. andrei,
C. sagoti i, brachycalyx, C. plumosa, C^. hermanni i, C. froesi i.
GYNOECIUM: ovary elongated, 28-32 mm, C_^ amazonum; style highly
elongated, 27-35 mm, plumosa, C. pendens, C. obidensis, C.
pozuzoensis, C_. flexuosa, C. cavalcantei, occasionally C. quianensis,
C. densiflora; style very short, 7-10 mm, C. heterophylla, C.
cordobensi s, 10-11 nun in C. monti col a.
VEXILLUM: vexillum long-clawed, claw 15-20 mm, C_. amazonum,
C. javitensis, C. cavalcantei, C. densif1 ora; flower occasionally very
large, 8-9.5 cm long, C. pozuzoensis, C. obidensis, 1 . plumosa.
ALAE: ala blade elongate, 3-3.5 cm long, C. plumosa; ala claw
elongate, 26-33 mm, C. pozuzoensis; alae scarcely exeeding carina,
2-3 mm longer, heterophylla.
CARINA: carina broad, 6-8 mm, flexuosa, pozuzoensis,
C. woytkowskia, C_. obidensis, C. pendens, C.. plumosa, C. guianensis.
LEGUME: legume costate, stipularis, C_. simpIicifolia,
C. densiflora, C. falcata, CA laurifolia, C. guianensis; legume
pubescence dense, macrotrichomes, C. javitensis, C. dendrina,
C. brachystegia, brachycalyx; legume broad, 2-4 cm, C^_ arbórea,
fairchildiana, C^_ juninensis, nervosa, C_. bracnycalyx,
C. javitensis, C. pozuzoensis; legume width very narrow, 4-5 mm wide,
C. heterophylla; dehiscence 2-4 turns, 1eptostachya, C. brachystegia;
legume convex, subgenus Neurocarpum; stipe subsessile, 1-4 mm,
C. lasciva, C. biflora, C_. terna tea, C. heterophyl 1 a, C. kaessneri;
stipe highly elongated, 2.5-5 cm long, C. amazonum, i. flexuosa,

158
C• pozuzoensis, C. javitens 1s , C. sagoti i, C. 1eptos Lachya, C. pendens,
C. dendrina, C. brachycalyx, occasional brachystegia.
SEEDS: seeds viscid, subgenus Neurocarpum; seed germination
hypogean, subgenus Neurocarpum; large number of seeds per legume,
12-17 seeds, C. amazonum; few seeds per legume, 1-4 ^eeds, C. kaessneri,
C. biflora, C. polystachya, Asian mariana.
CLEISTOGAMY: eleven species of subgenus Neurocarpum, character¬
istics summarized in Tables 7 and 8; chasmocleistogamous inflorescences,
bearing one chasmogamous flower and one cleistogamous flower (extremely
rare), once in C. guianensis, twice in C. mariana.

DISTRIBUTION OF CLITORIA
Cl itoria is a tropical genus with forty-six species native to the
neotropics and subtropics, eleven species native to the paleotropics,
and one species native to the neotemperate and tropical Southeast Asian
regions. Figure 8 illustrates the generalized disti i but ion of the
genus which is found within the zone marked by black lines. Figures 9-
13 illustrate the distribution of each subgenus and its sections.
Subgenus Bractearia has twenty-nine species distributed primarily
in northern South America and Panama, with two species having isolated
populations occurring as far north as southern Mexico, one species
from Bolivia, and another species from eastern Brazil. The most
primitive section, Bractearia, composed of arboreal and tall fruticose
species, is distributed primarily along the Amazon Basin and has one
species known from Bolivia (Figure 9). Section FIexuosa, a primitive
section of predominately lianas, is distributed in tcuadorian and
Peruvian tropical forests (Figure 9). Section Brack/calyx, a slightly
more advanced section composed of mostly arboreal arid fruticose species,
is distributed in northern South America and Panama (Figure 10).
Closely related section Caul1florae, consisting of liana and a few
fruticose species, is also commonly distributed in northern South
America, with a few species found in Peru, a few species known from the
refugia forests along the Amazon River, and one species located in
eastern Brazil (Figure 10).
159

Figure 8. Generalized distribution of the genus Clitoria.

r
L

Figure 9.
Distribution of two sections of subgenus Bractearia.
Section Bracte¿y^ia, diagonal lines; section FIexuosa, dots.

163

Figure 10. Distribution of two sections of subgenus Bractearia.
Section Brachycalyx, dots; section Cauliflorae, diaqonal
lines.

165

166
Subgenus Clitoria has five species native to the paleotropics of
Africa and Southeast Asia, with one anthrogenic species nearly
pantropical in its distribution. Figure 11 illustrates the distribu¬
tion of the subgenus with dotted areas representing the probable native
distribution of the subgenus. Cl itoria terna tea is commonly cultivated
throughout the tropics, and often escapes, flourishes, and becomes
naturalized. It is commonly found within the areas bordered by black
lines where the species had been introduced by man.
The more advanced subgenus Neurocarpum has twenty-four species and
exhibits a disjunct distribution, with most species native to the
neotropics or subtropics. Section Mexicana has a widely disjunct
distribution with most species found from central Mexico to northern
Panama. One species is endemic to northern Argentina and one temperate
species is known from the eastern United States and also is native to
Southeast Asia (Figure 12). Section Tanyst.yloba is distributed in
Southeast Asia from western India to China (Figure 12). Section
Neurocarpum is widespread in the neotropics, distributed from southern
Brazil to southern Mexico, and also common in the West Indies (Figure 13).
Two species have been introduced into Africa and become naturalized in
small populations within the dotted region. One species has been
introduced into portions of Southeast Asia and Indonesia where it has
become established.

Figure 11. Distribution of subgenus Clitoria. Dotted areas represent
probable native distribution; areas bounded by black lines
represent regions where the anthrogenic species, C.
ternatea, has escaped cultivation and become naturalized.

r
168

Figure 12. Distribution of two sections of subyenus Neurocarpum.
Section Mexicana, dots; section Tanystyloba, horizontal
1ines.

170

Figure 13. Distribution of section Neurocarpurn, subgenus Neurocarpum.


SYSTEMATIC CRITERIA AND TAXONOMIC NOTrS
The monographical study of the genus Clitoria was based upon the
classical criteria of morphology, distribution, and comparison of
specimens to the types. Additional evidence available from the
literature, and limited field experience with the Floridian species,
were used for supportive data. From this study a more comprehensive
understanding of the genus was obtained which will provide a basic
conceptual foundation that can later be supplemented by further
research employing biosystematic techniques.
During 1973-1974, requests for loans of Clitoria specimens were
mailed to nearly 70 herbaria (those believed to havi. type collections
or known to have specialized in certain geographical areas). Nearly
7000 sheets of specimens were received and examined and have been
matched to the types. Herbarium abbreviations when.ver used herein
follow Holmgren and Keuken (1974). Specimens examined are listed
below by the loaning institution's abbreviations.
A: Arnold Arboretum, Harvard University, Cambridge,
Massachusetts, U.S.A.
BA: Museo Argentino de Ciencias Naturales "Bernardino
Rivadavia" e Institutio Nacional de Investigaciones
de las Ciencias Naturales Avda, Buenos Aires,
Argentina.
BM: British Museum of Natural History, London, Great Britain.
BR: Jardin Botanique National de Belgique, L.ruxelles, Belgium.
CGE: Botany School, University of Cambridge, Cambridge, Great
Britain.
173

174
E: Royal Botanic Garden, Edinburgh, Scotland, Great Britain.
F: John G. Searle Herbarium, Field Museum of Natural History,
Chicago, Illinois, U.S.A.
FLAS: Herbarium, Agricultural Experiment Station, Gainesville,
Florida, U.S.A.
G: Conservatoire et Jardin Botaniques, Geneve, Switzerland.
GH: Gray Herbarium of Harvard University, Cambridge,
Massachusetts, U.S.A.
HAL: Martin-Luther-Universitat, Sektion Biowissenschaften,
Wissenschaftsbereich Geobotanik und Botanischer
Garten, Halle, German Democratic Republic, DDR.
K: Herbarium and Library, Royal Botanical Gardens, Kew,
Richmond, Surrey, Great Britain.
LA: Botanical Gardens-Herbarium, Department of Botany,
University of California, Los Angeles, California,
U.S.A.
M: Botanische Staatssammlung, Menzingerstrasse, München,
Federal Republic of Germany, BRD.
MICH: Herbarium of the University of Michigan, Ann Arbor,
Michigan, U.S.A.
MO: Herbarium of the Missouri Botanical Garden, Saint Louis,
Missouri, U.S.A.
MPU: Instituí de Botanique, Université de Montpellier,
Montpellier (Herault), France
NY: Herbarium, The New York Botanical Garden, Bronx Park,
Bronx, New York, U.S.A.
P: Museum National d'Histoire Naturelle, Lcboratoire de
Phanérogamie, Paris, France.
PENN: Herbarium of the University of Pennsylvania, Philadelphia,
Pennsylvania, U.S.A. (Acquired by PH in late 1974.)
PH: Dept, of Botany, Academy of Natural Sciences, Philadelphia,
Pennsylvania, U.S.A.
PR: Botanicke Oddeleni Prirodoved muzea Narodniho muzea v
Praze, Praha, Czechoslovakia.
RB: Jardin Botánico do Rio de Janeiro, Rio de Janeiro, Brazil.

175
S: Section for Botany, Swedish Museum of Natural History
(Naturhistoriska riksmuscet), Stoctholm, Sweden.
SI: Instituto de Botánica Darwinion, San Isidro, Prov. Buenos
Aires, Argentina.
U: Institute for Systematic Botany, Tweede Transitorium,
Heidelberglaan, Utrecht, Netherlands.
UC: Herbarium of the University of California, Department of
Botany, University of California, Berkeley,
California, U.S.A.
UMO: Herbarium, University of Missouri, Columbia, Missouri,
U.S.A.
US: U.S. National Herbarium, Department of botany, Smithsonian
Institution, Washington, D.C., U.S.A.
VEN: Instituto Botánico, Caracas, Venezuela.
VSC: Valdosta State College Herbarium, Valdosta State College,
Valdosta, Georgia, U.S.A.
W: Naturhistorisches Museum, Burgring, Wein, Austria.
WIS: Herbarium, University of Wisconsin, Madison, Wisconsin,
U.S.A.
Measurements have been made from structures that were mature or
as nearly so as possible. Ranges of variation are given with
measurements encountered infrequently enclosed in parentheses. Leaflet
size is often quite variable, but was generally taken as representative
when found with flowers and fruits, and the smaller end of the range
may be indicative of less mature leaves. In flower size, the
complicate vexillum and mounting techniques used, make it difficult to
determine the maturity. Ranges represent apparently mature flowers,
and the smaller end of the ranges may represent flowers not fully
mature.
Rarely did the original authors designate holoiypes for the
species of Clitoria. In many cases one to a few collections were

176
cited. These "type" collections were treated as syntypes. Most
definitions of syntype include the situation in which an author cited
two or more collections (i.e. C. multiflora: Galeotii 2323 and 3290),
each collection becoming a syntype. In a number of cases only one
collection was cited, but specimens of this collection are found in
several herbaria (i.e. C. densiflora var. mucronulata: Hassler 9831,
BM,F,G-5 sheets,MPU,NY,S,UC,W). It is often in doul t whether the
original author examined only one specimen or more than one, and which
specimen(s) was that on which the description was bused. With no
holotype designated, this second example illustrated a second type of
syntype collection, one implied but not described in most syntype
definitions.
When holotypes were not designated, a lectotyp. was selected from
a syntype specimen. Lectotypes were not selected for those binominals
treated in synonymy. In selecting a lectotype, fir t priority went to
the specimen that was believed to have been examined by the original
author who erected the species. Next priority went to the specimen
which appeared to best match the description as originally published.
Last priority went to the specimen which would be the best repre¬
sentative specimen. Priority was not given to the institution from
which the specimen was deposited, thus lectotypes will be scattered in
various institutions on various continents, rather than localized at
one or two institutions near the author of this stud/.
For new species and new subspecific taxa, holot/pes were
selected. In a number of cases few to numerous collections were
examined. Technically, all will become paratypes when published with
the holotype. Although each paratype contributed to the description

177
as a whole, the type concept (one or more specimens upon which the
name of the taxon is based) would be abused by having a large number
of "type collections" (e.g., a new variety, (R poly .tachya var.
congesta, has 9 cited collections and a total of 24 sheets examined).
Therefore, a representative collection(s) was selected and labeled as
the paratype collection, with the remaining collections treated as
non-type collections.
In the citation of types, accession numbers were used after the
institutional abbreviations (e.g., UC 83907) to identify more
specifically a specific specimen. A number of institutions lacked
accession numbers; however, in the process of prepaiing the specimens
for loan, some lending institutions placed the loan number and a loan
accessional count on the herbarium sheet. These aci.essional numbers
were cited as substitutes for regular accession numbers. The loan
number also present on the herbarium sheet is listed below, but was
not cited for each individual use because of its repetitiveness. For
example, the type specimen of arbórea (cited as 1;R-8) indicates
that the type specimen is deposited at Bruxelles anti bears the loan
number "74/2521" and a loan accessional number of "b" within the stamp
as described below. A list of these institutions and their loan
numbers are as follows (xx is used in place of any accessional
number):
BR: Bruxelles specimens have inked, circular stamp with "Herbier
Jardin Botanique De l'Etat, Bruxelles" around the
margin, and two horizontal lines in the center. The
upper line, in pencil, has "xx"; between the lines, in
pencil, is the loan number "74/2521"; and below the
bottom line, in ink, is "FLAS." Accession numbers
range from 1-11. Specimens are cited as "BR-xx."

178
E: Edinburgh specimens have an inked, circular stamp with
"Royal Botanical Gardens, Edinburgh" around the
margin. In the center of the circle, in pencil, is
the loan number "60/74" with "xx" below this number,
ranging from 1 to 150. Specimens are cited as "E-xx."
G: Geneva specimens are mounted to a white Cardboard strip and
pinned to the inside of a folder, occasionally with one
or two additional sheets within the I older. The lower
right hand corner on the outer front of the folder
bears the loan number "3538/xx" where xx ranges from
1 to 588. Specimens are cited as "G xx" where "xx" may
be a multiple number to include the additional sheets
upon which the specimen is mounted.
K: Kew specimens have an inked, elliptical stamp with "Royal
Botanical Gardens, Kew" at the upper and lower margins.
In the center of the ellipse, in pemil, is the loan
number "HO 847/74" with "xx" placed below this number
and ranging from 1 to 453. Specimen, are cited as
"K-xx."
M: Muchen specimens have a penciled loan number "596/xx" either
above or below the data label, but usually placed in
the right hand corner. Some specimens bear a gummed
strip taped around the specimen, with an inked
accession [?] number. Specimens will first be cited by
the inked number on the gum strip attached to the
specimen, then secondly by the loan number (i.e. as
"M-xx") when the gummed strip is lacking.
P: Paris specimens have an inked, circular stamp with "Herb.
Mus. Paris, Phanerogamie" around the margin. In the
center, in ink, is "Pret No." with the number "59" next
to it, and the number "xx.75" below it. Both sets of
numbers are in pencil with xx ranging from 1 to 61.
Specimens are cited as "P-xx."
Nontype collections cited for a specific purpose (i.e., an illustration
documentation, or a specimen cited which lacks identifiable data) will
follow the same style for institutional citation as types.
A small number of specimens were encountered which were labeled
as type collections (i.e., "nov. gen.," "sp. nov.") that bore names
which apparently are unpublished (e.g., C. pinetorum McFarlin) or
were uncited when the name was published in synonymy (Rhombi1 obi urn
Rich.). Unpublished names will be treated as nomina ija scheda (nom.

179
in sched., an illegitimate name known only from an herbarium label).
The collections cited for these names (in both case.) will be the
specimens from which the name was obtained, the probable type
collections.
Specimen citations include locality data as th. y appear on the
label to avoid a possible change in the data by this author. All
months are cited in 3-lettered abbreviations, not by numerals, except
in those few instances in which the data for the month versus the date
are unclear, then the exact numerals are used. The number of sheets
examined from a particular institution are noted; thus "A,BM,G-5 sh.,
K-2 sh.,W" indicates that this author examined and annotated during
this study, five sheets from Geneva, two sheets from Kew, and one
sheet from each of the other institutions for this collection.
Specimen citations are arranged geographically. The more modern
geographical name for a country is included along with the older name.
Major provinces were obtained from field data and from maps contained
within the Comprehensive Times Atlas.

SYSTEMATIC TREATMENT
Clitoria L., Sp. PI. ed. 1, 2: 753. 1753.
Ternatea Tourn., Mem. Math. Phys. Acad. Roy. Sc i. , p. 103, Tab. 5.
1 706; nom. in val. (Based on Flos c1itoriüis Ternatensibus
Breyne).
Clitorius Petiv. ex Dill., Hort. Elth., p. 90. 1732; nom. inval.
(Based on Clitorius Petiv., Flos clitoridls Ternatensibus
Breyne, and Ternatea Tourn.).
Clitoria L., Gen. PI. 1: 216, no. 572. 1737; nom. inval. (Based
on Ternatea Tourn. and Clitori us Dill.).
Ternatea Tourn. ex Mill., Gard. Diet. ed. 4, 3: TERNATEA. 1754.
(Based on Ternatea Tourn.).
Neurocarpum Desv., Journ. Bot. ]_: 119. 1013. (Based on
Crotalaria guyanensis Aubl. and N. el 1ipticum Desv. ).
Vexillaria Eaton, Man. Bot. p. 82. 1817; non Voxi11 aria Raf.
(1818) nec Vexillarija Hoffm. (1824); nom. >uperfl. (Based
on Clitoria L.).
Vexillaria Raf., Amer. Monthly Mag. & Crit. Rev. 2: 268, no. 111.
1818; non Vexillaria Eaton (1817) nec Vexillaria Hoffm.
(1824); nom. i 11 eg. (Type: Clitoria mari ina L.).
Martia Leandr.-Sacr., Denkschr. Akad. Muench. 7: 238. 1821;
non Martia Sprengl. 1818; nom. i11 eg. (Type: M. physodes
Leandr.-Sacr. ).
180

181
Martiusi a Schultes, Marit. 1_: 69. 1822. (Type: Martia physodes
Leandr.-Sacr.).
Rhombolobium Rich, ex H.B.K., Nov. Gen. Sp. 6: 106. 1824; nom.
inval. in synon. (Based on Rhombolobium, jen. nov.,
Richard Herb. ).
Neurocarpon Desv. ex Hamilton, FI. W. Ind. p. SO, no. 101. 1825;
nom. illeg. (orthogr. pro. Neurocarpum Desv. 1313).
Rhombifoliurn Rich, ex DC., Prod. 2: 235. 1825, nonw_ inval. ija
synon. (Based on Rhombifolium canescens, Richard Herb.).
Nauchea Desc., Mem. Soc. Linn. Par. 4: 7. 182C; nom. superfl.
(Based on CTitoria L^).
Clytoj'ia Presl , Rostí. 3: 1 96. 1835; ?nom. i 1 I eg. (?orthogr. pro
C l i torga L. 1 7 53 ). '
Neurocarpips Desv. ex Haaskarl, Cat. Hort. Bog. Alt. p. 276. 1844;
nono i]leg._ (ortho.gr. pro. Neurocarpum De-.v. 1813).
Macrotrul1 ion Klotsch ex Rich. Schomb., Br. Guiana Fauna & Flora
p. 1202. 1848; nom. inval. sin descr. gen. (?Based on
Macrotrul1 ion Klotsch, nov. gen. Mss., Klotzsch herb.).
Trees, shrubs, lianas, subshrubs (woody herbs), herbs or herbaceous
vines, erect or sometimes prostrate, apex sometimes «.limbing by twining,
usually pubescent; miscroscropic trichomes uncinate, macroscopic
trichomes variable; pith typically hollow, occasionally solid, bark
medium to dark, splitting in longitudinal strips, lighter beneath.
1. This author is unable to establish title ot Presl publication
nor obtain a copy. It is presumed that Clytori a Presl was an
orthographic variant of CJitoria L. Reference source for Clytoria
Presl is Index Kewensis, fase. 1: 569.

182
Leaves alternate, odd-pinnately compound, usually -foliate, less
commonly 1-, 5-, 7-, 9-, 11-foliate, petiolate, occasionally subsessile
or rarely digitately compound, terminal leaflet usually largest;
leaflets entire, midrib and nerves impressed above, prominately raised, reticulate below, primary nerve pairs ascending,
usually arcuate towards margin, joining the vein above, secondary veins
conspicuously reticulate below. Stipules and stipe!s usually
persistent, striate, erect, appressed, pubescent aba narrower than stipule, often shorter, terminal stipel larger than
lateral. Petiole and rachis longitudinally striated, sometimes
caniculate to sulcate, often laterally comoressed near lower leaflets,
petiole base swollen; petiolules subquadrate, rugose, dark-colored.
Inflorescence axillary, terminal or cauliferous, racemose or
occasionally paniculate, bearing chasmogamous or inrrequently
cleistogamous flowers, rarely chasmocleistogamous, muítiflowered to
uniflowered; peduncles usually solitary, rarely 2-3 in axils, or
several, short, fascicled, cauliferous; pedicels paired at nodes.
Bracts striated, pubescent abaxially, distributed in three series
below pedicels; inner pair of bracts between the pedicels, caducous,
smaller than middle and outer bracts; middle pair o; bracts concave,
opposite and appressed to pedicels, sometimes becoming reflexed in age,
persistent, the largest and most conspicuous bracts; outer pair of
bracts between the pedicels, diciduous to semipersistent, similar to
and slightly smaller than the middle pair of bracts Bracteoles paired,
persistent, striate, usually appressed to calyx, pubescent abaxially.
CHASMOGAMOUS FLOWERS resupinate, showy, bisexual, 5-merous,
papilionaceous or rarely act inomorphic, blue to violet, pink, or

183
white, sometimes white fading to pale yellow. Caly infundibular,
persistent in fruit, conspicuously multi-nerved to lú-nerved, apex
5-lobed, upper two lobes (opposite vexillum) subconiute, lowermost
narrower, often longer. Vexillum complicate, large and flag-like when
expanded, erect, apex emarginate, short-clawed, spur less, pigmentation
conspicuous on inner face, darker towards periphery, center light-
colored, veins darker-colored, prominent, converging into throat,
pubescent outside. Alae falcate-oblong, spatulate, shorter than
vexillum, extending beyond carina, adherent in middle to carina,
long-clawed. Carina falcate to subfalcate, incurved, acute, shorter
than alae, long-clawed. Stamens 10, diadelphous, vcxillary stamen free
or connate at base to near middle, staminal tube complete or sometimes
split at base, apex incurved, often persistent in fiuit, glabrous, free
filaments short, filiform; anthers uniform, basifix*d. Disc observed
as conspicuously dark-colored, swollen calyx base. Gynoecium enclosed
in staminal tube; ovary stipitate, linear, compre sed laterally,
many-ovuled, densely pubescent; gynophore usually projecting beyond
disc, densely pubescent at apex; style elongate, flattened, dilated
apically, bearded lengthwise, geniculate near tip; tigma terminal,
flattened to capitate, usually pubescent around baso. Actinomorphic
flowers showy, 4-5 merous, deep blue to violet, rarely white. Petals
all banner-like, subequal. Stamens free or subpolyudelphous. Calyx
and gynoecium similar to those of papilionaceous flowers. CLEIST0GAM0US
FLOWERS uncommon, inconspicuous unless with fruit, borne above ground.
Petals usually lacking, when present minute, whitish, subtransparent,
minutely clawed, one to five, similar. Calyx infundibular, small,
persistent in fruit, 5-lobed, similar to chasmogamous calyx. Stamens

184
short, free or diade1phous, to 10, within the calyx tube. Ovary
similar to that of chasmogamous flowers, smaller; sf/le bent abruptly
back toward the base; stigma in contact with anthers. Legume
stipitate, linear, compressed, straight to subfalcaie, thickened on
upper or on both sutures, ecostate or infrequently costate, valves flat
or convex, weak to strongly depressed between seeds, often micro¬
scopically pubescent, beaked by persistent styic base, splitting along
sutures and then spirally twisted; legume from clei: togamous flowers
similar' to those of chasmogamous flowers, shorter. Seeds dark brown
becoming black, glabrous, smooth or viscid, flat or thickened, face
subglobose, or reniform, rarely ovate to oblong.
TYPE SPECIES: Clitori a ternatea L.
NOMENCLATURAL NOTES: The Internation.il Code oi Botanical
Nomenclature established 1 May 1753 as the official starting date for
nomenclatural history. Three genera were described in publications
prior to this date, Ternatea Tourn. in 1706, Clitor ius Petiv. ex Dill,
in 1732, and Clitoria L. in 1737. These generic names are illegitimate
as they were published contrary to the rules of Article 13 of the Code.
The generic name Clitoria L. became a legitimate name when
published in the first edition of Linnaeus' "Specie Plantarum," even
though it was not accompanied by a description of the genus. Article 41
states that those generic names first published by 1innaeus in "Species
Plantarum, ed. 1" are treated as having been validly published on that
date (1753), and Article 13, Note 3 states that these generic names
are associated with their first subsequent description given under
those names in Linnaeus' "Genera Plantarum, ed. 5 (17 54 )." Linnaeus'
description of Clitoria published in 1754 was identical to his

185
description of the genus in 1737. He described Hit iria as:
CAL. Perianthium monophyllum, erectum, tabulaturn,
quinquedentatum, persistens. COR. papilionace.;. Vexi11um
maximum, rectum, emarginatum, margine undulatum, patens.
Alae oblongae, rectae, obtusae, vexillo brevioies. Carina
alis brevior, subrotundo-falcata. SIAM. Fi 1 a.n- uta diadelpha
(simplex & novemsidum). Antherae simp!ices. I 1ST. Germen
oblongum. Stylus adscendens. Stigma obtusum. PER. Legumen
longissimum, lineare, compressum, uniloculare, bivalve,
ápice subulato. SEM. plura, renifornia.
Nearly all references, with the exception of the Gray Card Index,
attribute Ternatea to Humboldt, Bonpland and Kunth in 1823. First
these authors cited the genus as "Ternatea Tourn." second, Stafleu
(1967, p. 226) established the publication date as apt. 1824. Thus,
the full citation would read: "Ternatea Tourn. ex L.B.K., Nov. Gen.
(6) (29): 41 5, 1824." However, the generic name Ternatea Tourn. ex
Mill, became a legitimate name at an earlier date, i /54, when Miller's
fourth edition of the "Gardner's Dictionary" follow. 1 Tournefort's
treatment of the genus Ternatea, with the genus fully described. No
type collection is known, although Tournefort did indicate that the
genus was named after the Mol ucean island of Ternate, and Tournefort's
first species was based on Breyne's Ternate specimen. Thus, it is
presumed that Breyne's specimen (illustrated in Brene, Exot. PI. Cent,
p. 76, tab. 31, 1678; specimen may be with Breyrie herbaria deposited
at Leiden) may be the type.
Desvaux frequently altered the spelling of spe< ¡fie and generic
names in his publications. His genus Neurocarpum (1313) became
Neurocarpon (1825) and Neurocarpus (1844). Desvaux described the genus
briefly in 1813 and more fully in 1814 (Journ. Bot. 2: 75). No type
was designated, although in 1813 he stated: "J'y place la Crotalaria

186
gajanensis, et un espece nouvelle (Neurocarpuin el 11 ¡ ticum Desv.)." In
1814, Desvaux listed two species: (I) Neurocarpum Janensis (syn.
Crotularia guyanensis Aubl.; Crotularia longifol ia Lain.) and (II) Neuro-
carpum el 1 ipticum. Desvaux made orthographic error., in his citations,
which should have been Crotalaria guianensis Aubl. uiid Crotalaria
longifolia Lam. No specimen has been found in Desvaux' herbarium (P)
that bears the name N_. el 1 ipticum. There is a specimen, "hab in
gujanae, Herb. Desvaux (P -59!) that bears the name Crotalaria with
"gujanensis Aubl. guy. 2 p. 76, t 305" and "longifolia Lamrk ency. 2
p. 201" placed after Crotalaria on the label. In a darker ink,
Neurocarpum is written above Crotalaria arid gajanen-.,is is written above
the two specific epithets. This specimen is the only known type
collection for the genus. One final note: Desvaux misidentified this
specimen as Crotalaria guyanensis Aubl. and Crotalaria longifolia Lam.
Desvaux' name, N. gajanensis, is usually included in synonymy with
Clitoria guyanensis (Aubl.) Benth., but should correctly be placed in
synonymy with Clitoria laurifolia Poir.
Vexillaria Eaton and Mauchea Desc. are substitute names for
Cl itoria L. to express the authors' dissatisfaction with the generic
name Cl itoria having an alleged reference to the fen,ale anatomy. These
two names were nomenclaturally superfluous when published, and as
described by Article 63 of the Code, are illegitimate. Vexi11 aria Raf.
was published without a description (contrary to Article 41) and was
a homonym of Vexillaria Eat. (as described in Article 45), and thus is
illegitimate (Article 64). Neither Vexil1 aria Eaton nor Vexillaria
Raf. should be confused with the later published Vexillaria Hoffm.
(1824), a synonym of the genus Centrosema (DC.) Benth.

187
Leandro do Sacromento described Martia without any reference to a
type collection. The description indicated a cleisiogamous specimen.
A Brazilian specimen labeled "Martia physodes Leandij. C. pater
Leandro do Sacromento, prope Rio de Janeiro (M 12423!)" agreed with
the description and is the probable type. The generic name is a
homonym of Martia Sprengl. (1818) and in accordance with Articles 45
and 64 of the Code, is an illegitimate name.
Hutchinson (1964, p. 446) cited "Rhombolobium Rich ex Pfeiffer
(1874)"- as a synonym of Cl itoria L. No reference was given, although
based on Stafleu (1967, p. 357), Hutchinson may be referring to:
Pfeiffer, Nomenclátor Botanicus, vol. 1(2), or 2(1). or 2(2), all
published in 1874." Humboldt, Bonpland and Kunth (1324) published
"Rhombolobium Rich, mss." as a synonym of Neurocarpnn Desv. Thus,
Rhombol obium Rich, ex H.B.K., in accordance with Anide 34 of the
Code, is an invalid name when it is merely cited as a synonym. No
type was designed, although a Paris specimen in Richard's Herbarium
Guyanensi-Anti 11anum (P -52!) bears an attached lábil with the generic
description of Rhombolobium as:
Caul, imapare sublignosus, cylindricus, volubilis, sub...,
rufescenti-pubesc-
Fol. trifolia; petiolo sub 2 uncial i; ..., ad Rasim
...rido, uti caulis piloso; foliolis ovalibus,
emargiumbi, subtus subincauis, supra lucidii,
subbiuncialibus - Stipulae lanceolatus; siipulellae
1ineares.
FI. pedunculi axill. solitaria’ s. gemicei, recti,
rigiduli; ad apicum 3-8 flori; florib. ovicalis,
gemisuit, bracteis 2 lanceolatis singulactive
stipa til.
Cal. uncial is; nervosus; subinfundibu 1is tubulosus,
aequalis 5 fidas; lacinius lanceolati, erecti.
2. This author has been unable to obtain a cop/ of Pfeiffer's
pub!ication.

188
Col. alba; Vexillum calyce duplo longias, maju culum,
lateribus inflexil (ferein forinam connat) uxtrusious
rotundatiam emarginatum- unguiculaturn; ad :entrum
violaceo-1ineatum: Alae i 11 o breviorus, wblongue,
obtusae, carinae ...plicital. Carina parvum
longitudina, 2-..., falcata, acuta.
Stam. 10, diadelpha, inclusa; anth. ovatae.
Pist. ovarium stipitatum, lineare: stylus arcuitus,
...ciferus, ad ult... angustum pubeus: stig.
glabrum, capitatum.
Legumeus: 1-3 uniloc..., oblatera medio carinata
subtrigonuin: ... simina distri. . .atibus.
Sem. 4-6. subglobora, subgrirea, (retículo viscido
obducta).^
This specimen is the probable type for the genus with a Geneva specimen
labeled: "Rhombi 1 obium Richard, genre nouveau qui doit etre place a
coti du Clitoria. II y a un angle que ... du milieu du Legume du
chaqué coti" (G -545!) a probable isotype.
The name Rhombifolium Rich, ex DC. was published by de Candolle as
a synonym of Neurocarpum Desv. and in accordance with Article 34 of the
Code, it is an invalid name. No type was designated, although
de Candolle (1824; p. 236) did publish the name Rhomoifolium canescens
Rich, ex DC. as a synonym of Neurocarpum guianense Desv. A specimen
in the Richard herb, that bears the name R. canescens Rich, is the
probable type for the genus. This author has not seen such a specimen.
The name Macrotrul1 ion Klotsch ex Rich. Schomb. was published
without a description and in accordance with Article 41 of the Code,
it is an invalid name. The type collection was probably a specimen (s)
in Klotsch1s herbarium that bore the names of eithet M. splendens K1.
or PI. elegans Kl. Stafleu (1 967 , p. 240) stated that Klotsch's
herbarium was deposited at Berlin and subsequently destroyed.
3. Portions of this handwriting were difficult to read.
Illegible letters designated by three dots.

189
NOTES ON SUBGENERA: The species of Clitori a full into three
natural groups on the basis of morphological and distributional
criteria. These groups can be distinguished by their fruits and seeds,
and their separate integrities supported by habit, loaf, calyx,
androecium, gynoecium, and ala characteristics. The presence or
absence of cleistogamy supports the recognition of these three groups.
Little work has been undertaken in other areas of research on Clitoria,
but the scant cytological and seed germination data argue for the
preservation of these groups as defined. Each group forms discrete
distributional units and the species are found in similar habitats
within the groups, yet somewhat distinct from the other groups. If
one compares the morphological description of Cl itoi ia_ with other
legume genera in Hutchinson's treatment (1964), Cliiqria is much more
polymorphic than these other genera. The distinction in the fruit and
seeds supported by a number of other differences suggests that these
natural groups represent three genera as opposed to three subgeneric
groups. The Clitoria (syn. Ternatea) group and the Neurocarpum group
have both been treated in the past as distinct genera. The Neurocarpum
group, with more polymorphic differences, lias, in fact, been treated as
more than one genus. However, in attempting to render Clitoria more
natural, this author rejected the treatment of these three groups as
distinct genera because of the structural organization of the flower
which is consistently similar across all three groups plus the number
of similar characteristics of other structures, as indicated in the
generic description. The Clitoria flower is unique among the
papilionaceous legumes with its resupinate position, bearded style,
larger, showy flowers, stalked ovary, and persistent, tubular calyx.
Resupination of the papilionaceous flower is a unique, advanced

190
characteriStic found in only twenty of the four hundred eighty-three
papilionaceous genera. Those genera with resupinati. flowers usually
have small flowers, campanulate cálices, and sessile to subsessile
ovaries. None of these genera possess bearded styKs. The treatment
of the three groups as distinct genera would be, in che opinion of this
author, more artificial than natural. Bentham (185b) was the first to
recognize that these three groups, when treated as a single genus,
rendered a more natural systematic treatment, lie combined them with
the genus Clitoria in his synopsis of the genus. Bentham recognized
these three groups as sections. However, from the greater evidence now
available, and the number of differences that occur, it is better to
treat these three groups at the level of subgenus, and use the major
morphological differences within the groups as the lisie characters for
designation of sections.
The most polymorphic subgenus is Neurocarpum, which has different
types of fruits (turgid throughout vs. turgid only around the seeds;
costate vs. ecostate), seeds (subreniform vs. globoid vs. ovoid),
leaves (3-foliate vs. 3- and 1-foliate or only 1-foliate; stalked vs.
subsessile), inflorescences (paniculate vs. racemose; multi-flowered vs.
biflowered), flowers (chasmogamous vs. cleistogamous and chasmogamous),
and cálices (lobes shorter than tube vs. longer than tube), as well as
a number of size differences. Some of the characteristics are unique
to this subgenus, and absent from other members of the genus. This
author attempted to further subdivide this subgenus into additional
subgenera using the characters mentioned. However, there were always
exceptions to each tentative group formed. Some species were difficult
to place, having affinities with more than one group. Therefore, this

191
concept of splitting this subgenus was rejected as t jo artificial.
Neurocarpum was left intact as a polymorphic subgenus, a more natural
treatment, with these characteristics used for sectional segregation.
KEYS TO THE SUBGENERA: Leaves (Table 3), flowers (Table 4), and
fruits (Table 7) are contrasted among subgenera in the tables noted in
the chapter on Morphology.
1. Fruit flat or occasionally 1omentaceous, weakly convex around
seeds, depressed between seeds, ecostate, long-stipitate (12-33 mm);
seeds smooth, compressed, orbicular, large (7-16 mm dia.); seed
germination epigean; staminal tube (2.5-4 cm); ovary long
(10-20 mm); calyx subcoriaceous, multi-nerved, persistent in fruit;
inflorescence woody, often nodose, few to multi- flowered;
cleistogamy absent; leaves 3-foliate, long-stalled, large
(leaflet 8-20 cm x 3-23 cm); petiolule large (4 10 mm); tree,
tall erect shrub-treelet, with sometimes a climbing apex, or
liana (x = ?) I. Subgenus Bractearia (p. 193 )
1. Fruit flat, ecostate, subsessile (1-4 mm stipe); seeds smooth,
compressed, subreniform, small (4-8 mm dia.); seed germination
epigean; staminal tube short (1-2 cm); ovary short (5-9 mm);
calyx subpellucid, 10-nerved, shrinking in fruit, inflorescence
sublignose, usually 1- or 2-flowered, rarely few-flowered;
cleistogamy absent; leaves 5- to 11-foliate, rarely 3- or 1-foliate,
subsessile, smaller (leaflets 2-8 x 1-4 cm); petiolule small
(1-3 mm); shrub to herb, erect to twining, rarely liana (x = 8)
II. Subgenus Clitoria (p. 486)

192
1. Fruit convex around seeds, deeply depressed between seeds, ecostate,
or convex, turgid, becoming subtetragonous, with a medial nerve on
each valve, short-stipitate (4-14 mm); seeds vi -id, thickened,
subreniform or occasionally orbicular to ovoid, small (3-6 mm dia.);
seed germination hypogean; staminal tube short (1.4-2.3 cm);
ovary short (5-9 mm); calyx subcarti1aginous, lU-nerved, persistent
in fruit, inflorescence sublignose, 2(1)- to several-flowered;
cleistogamy present or absent; leaves 3-, 3- and 1-, or 1-foliate,
subsessile or long-stalked, leaflet size variable, petiolule
usually small (1-5 mm, rarely 6-8 mm); shrub, subshrub (=woody
herb) to herb, erect or twining vine (x = 12).
III. Subgenu: Neurocarpum (p. 628)

193
Subgerms^ Bracteai ia
I. Clitona L. subgenus Bractearia (Mart, ex Benth.) Fantz,
stat. nov.
Cl itoria L. sect. Bractearia Mart, ex Benth., Ann. Wein.
Rus. Natur. 2: 115. 1837.
Clitoria L. sect. Dendrocyamus Benth., Ann. Ant. Hist.
3: 44. 1839; nom. superfl.
Clitoria L. sect. Clitorianthes Benth., Journ. Linn. Soc.
2: 41. 1858; nom. superfl.
Trees, tall erect shrubs-small trees, sometime, with remote branch
apices climbing, or lianas. Leaves 3-foliate, large, leaflet apex
usually acuminate, sometimes obtuse, acumen (0.5) 13 cm long, extreme
apex often mucronate, base cuneate to rotund, midrit impressed to
raised above; leaflets large, (5) 8-20 cm long, 3-10 (23) cm wide.
Petioles elongate, 3-15 (23) cm; rachis much shorter than petiole,
1-4 cm; petiolules large, (3) 4-10 mm. Inflorescence axillary,
terminal, or cauliferous, usually several- to multi-flowered, lignose,
paniculate, subpaniculate or racemose-nodose, axis extremely short
(0.5 cm) to extremely elongated (100 cm); peduncles solitary and
axillary, terminal, or fascicled and cauliferous, ruchis internodes
nearly straight to slightly curved, or occasionally zigzag. Bracts
typically (1) 3-7 (16) mm long, 1-4 (9) mm wide. Bracteoles broad,
coriaceous, usually subequal to or longer than the calyx, protecting
the flower in bud, or narrow and shorter than to sometimes subequalling
the calyx. Flowers chasmogamous, usually medium to large (4-9 cm),
sometimes small (2-4 cm). Calyx subcoriaceous, subimpressed to

194
prominently muí ti-nerved, persistent in fruit, pubescent; tube
usually broad at throat (8-11 mm), elongate (1-3 cm) infrequently
short (0.6-1 cm), somewhat cup-shaped with minute lubes; lobes usually
much shorter than the tube, typically 3-12 mm, occa ionally shorter or
longer. Vexillum long-clawed (typically 6-15 mm). Alae long-clawed
(typically 12-30 mm). Carina broad (typically 4-8 mm), long-clawed
(15-40 mm). Stamens diadelphous, vexillary stamen free to near base;
staminal tube a closed sheath or sometimes split near base, elongate-
(typically 2.5-4 cm); anthers 1-3 mm, connective sometimes apiculate.
Gynophore 1-8 mm; ovary elongated (1-3 cm); style 1 3.5 cm. Legume
long-stipitate, large (typically 6-25 cm x 1-4 cm), Mat, occasionally
weakly raised around seeds and depressed between seeds, ecostate,
lignose, thickened on both sutures, nearly straight, ventral margin
sometimes undulating, base cunéate, position obliquely on stipe, stipe
lignose, often arcuate, 1-5 cm. Seeds smooth, compiessed, orbicular to
lenticular, large (7-16 mm dia.), usually 4-12 (17) seeds per pod.
Seed germination epigean. (x = ?).
The members of the subgenus Bractearia can be characterized by
their woody habit, large 3-foliate leaves with acuminate tips, lignose
inflorescences of many flowers, subcoriaceous and multi-nerved calyx,
elongate staminal tube and ovary, large lignose legumes borne on
elongate stalks, and large, flattened, orbicular seeds.
LECTOTYPIC SPECIES: C. amazonum Mart. ex. Benth.
NOTES: This subgenus has always been treated previously as a
section of Clitoria, always intact, although Bentham twice rejected his
earlier sectional names and substituted new superfluous names for them
(cf. discussion in the chapter on History, pp. 36-40). The name

195
Bractearia was chosen for the subgenus because it v.vs the oldest name
used for this group and the only name validly publi tied. Bentham
(1837) originally fiad placed five species in this gi oup without
designating a type species. Later, Bentham (1858) nad increased the
list to nine species, still lacking a type designation. The name
Bractearia Mart, was obtained by Bentham from a spe^ imen (M 12408!)
in the Martius Herbarium which bore the name Bracte ¡rea amazónica Mart.,
and annoted as Clitoria amazónica Mart, a Benth. (see p. 252). As
only one species, Clitoria amazonum Mart, ex Benth., was involved in
the origin of the name Bractearia, it was selected ns the leptotypic
species for the subgenus.
This subgenus is poorly understood. Over half of the specimens
examined were misidentified. Another fifteen percent were unidentified
to species. Nearly fifteen percent of the remaining specimens were
correctly identified, but bore illegitimate names. Yet the specimens
examined have often been observed and cited in floriStic treatments.
As a result, the descriptions of species in a flora include a composite
of several species placed under one name or a description of one species
placed under an incorrect name. Thus, this author iecommends that one
disregard or correct the treatment of Bractearia spc :ies (usually
included under the superfluous name Clitorianthes) in most floras.
Within the subgenus Bractearia, there are three homogenous groups
and one heterogenous assortment. Bentham (1858) recognized two groups,
without naming them. These groups will be treated as sections.
Section Bractearia is easily recognized by its large bracteoles and
paniculate or subpaniculate inflorescences, and includes the type
species for the subgenus. It is the most primitive section. Section

196
FIexuosa is easily recognized by its zigzag ini lore^ence and the dense
appressed trie homes on the calyx, giving it a somewi it silky appearance.
All species were described since Bentham's treatment in 1858. Section
Brachycalyx is recognized by its small calyx tube with minute lobes
and its sublomentaceous fruits. All but one specie, was described
since Bentham's revision in 1858. Section Cauliflorae is a hetero¬
genous group comprised mostly of lianas, with cauliferous and axillary
inflorescences, and lacking the distinctive characteristics of the
other subgenera. This section includes the larger, unnamed group
designated by Bentham in 1858.
Mixed collections were more prominent in this ; ibgenus,
particularly when liana specimens were collected, often the fruiting
branches came from a different species, or genus, tlmn the flowering
specimen. These mixed collections occurred most frequently in Suriname
col 1ections.
DISTRIBUTION: The members of subgenus Bractearia usually occur
in the moist tropical forests of South America and (anaína.
KEY TO SECTIONS OF SUBGENUS BRACTEAR1A
1. Bracteoles broad (6-16 mm wide), coriaceous, usually long (10-40 mm)
and subequal the calyx (calyx thus hidden from view), rarely shorter
than calyx; inflorescence terminal or axillary, paniculate to
subpaniculate, primary lateral branches which bear pedicels are
subsessile (1-5 mm) to elongate (to 30 mm).
2. Calyx with scattered, short appressed trichunes to glabrate,
or rarely pilose; inflorescence straight to weakly curved,
internodes non-zigzag; trees or occasionally tall shrubs
Sect. IA. hractearia (p. 198)

197
2. Calyx densely pubescent, of short, appressed hairs (more or
less silky appearance); inflorescence tlexuous, at least near
apex, internodes weak to strongly zigzag; limas, occasionally
shrubs or trees Sect. IB. I iexuosa (p. 265)
Bracteoles narrow (1-4 nun, rarely to 6 mm wide), subcoriaceous,
usually short (2-12 mm, rarely very long and th.n always narrow)
and less than calyx length, rarely subequaling I he calyx (then
calyx conspicuously viewed); inflorescence axillary (often
appearing before leaves) and/or cauliferous, racemose, the primary
lateral branches absent or represented by a conspicuous knob to
1 mm (nodose).
3. Calyx tube shortened (0.6-1.3 cm) with minute lobes (1-3 mm,
rarely to 5 mm); legume sublomentaceous, conspicuously
depressed between seeds at maturity; trees »r tall shrubs-
small trees, rarely lianas . . Sect. IC. Biachycalyx (p. 295)
3. Calyx tube elongate (1-3 cm) with conspicuous teeth (4-18 mm);
legume flat, sometimes weakly raised around seeds, but not
conspicuously depressed between seeds; 1 ian. s or rarely
shrub-treelets
Sect. ID. Cauliflorae (p. 349)

198
Section Brae tea na
IA. Clitoria L. subgenus Bractearia (Mart, ex Benth.) Fantz
section Bractearia Mart, ex Benth. ei..end. Fantz
Clitoria L. sect. Bractearia Mart, ex Benin., Ann. Wein.
Mus. 2: 115. 1837.
Cl i toria L. sect. Dendrocyamus Benth., Ann. Nat. Hist.
3: 44. 1839; nom. superfl.
Clitoria L. sect. Clitorianthes Benth., Journ. Linn. Soc.
2: 41. 1858; nom. superfl.
Trees or occasionally tall shrubs. Inflorescence paniculate or
subpaniculate, lignose, the primary lateral branche bearing the
pedicels (1) 4-30 mm; the axis internodes non-zigzag. Calyx pubescence
typically scattered, short appressed trichomes to glabrate, infrequently
pilose. Bracteoles coriaceous, subequal to calyx (i ilyx usually hidden
from view), 6-16 mm wide and (11) 14-25 (28) mm long. Legume flat,
coriaceous to sublignose, typically broad, (1.5) 2-4 cm wide, and often
with sutures thickened; pubescence usually glabrous or microscopically
uncinate, macroscopic trichomes, when present, conspicuous along
sutures.
The members of the section Bractearia are characterized by their
tree-like habit, (sub-) paniculate inflorescences, glabrate calyx,
broad, thick-sutured fruits, and large conspicuous bracteoles which
obscure the calyx.
LECTOTYPIC SPECIES: C. amazonurn Mart, ex Benth.
NOTES: This section is the most primitive section of Cl itoria,
and is probably most typically represented by C. arbórea. Most species
are closely related forming a central core from which there appear to

199
be two distinct evolutionary departures. The first divergence is
represented by C. nervosa in which the calyx becomes pilose and the
flower size and most flower structures become small.-r. The second
divergence is represented by C_. amazonum (the lectot/pic species) in
which the apex shows climbing tendencies and the flower size (and many
component floral parts) become very large.
This section is easily distinguished from the i.losely related
section FIexuosa by the tree-like habit, sparse calyx pubescence, and
non-zigzag inflorescences. Only in C. amazonum are the flowers larger,
as are the members of section FIexuosa, although the conspicuous lateral
branches in the inflorescence would distinguish it trom FIexuosa
members.
DISTRIBUTION (Figure 9): The members of section Bractearia are
located generally in the tributaries of the Amazonum basin, with
endemic species at its fringe.
KEY TO SPECIES:
1.Calyx pubescence glabrate to pubescent with scattered appressed
hairs; calyx tube 13-18 (25) mm; flowers larger, 4-8 cm; bracteoles
(13) 15-28 mm; bracts long, 4-8 mm.
2.Flowers medium, 4-6 cm; vexillum pubescence moderately dense
appressed; stipules (5) 6-12 mm; legume width broad, 2.2-4 cm;
stipe short, 1.3-2.2 cm; bracteole width narrower, 6-12 mm.
3.Inflorescence paniculate, primary branches that bear the
pedicels conspicuous, elongate, 5-30 mm; ovary short, 7-9 mm.
(Amazon Basin; Brazil, Peru) 1. C. arbórea

200
3.Inflorescence subpaniculate, primary br,inches that bear
the pedicels subsessile to very short, i-5 (8) mm; ovary
longer, 11-14 mm.
4.Leaflet pubescence below glabrate to pubescent with
short, appressed hairs confined mainly to nerves;
leaflets acuminate, acumen (1) 1.5-3 cm; style short,
12-16 mm.
5.Inflorescence elongate, 8-42 cm; calyx pubescence
of scattered appressed hairs; stipule narrow,
2-3 mm wide; bracteoles long, 17-23 mm; bracts
short, 4-8 mm; plant of coastal deltas. (Brazil:
Amapa to Maranhao). . . 2. iuirchildiana
5. Inflorescence short, to 5 cm; c,.lyx glabrate;
stipule broad, 4-6 mm wide; bracteoles short,
12-15 mm; bracts long, 8-11 mm; plant of lower
mountains. (Peru) ... 5. iC. moyobambensis
4. Leaflet pubescence below subvelutinous; leaflets obtuse
to short acuminate, acumen absent to 0.5 (1) cm long;
style elongate, 16-21 mm.
6. Bracteoles 22-25 mm long, 9-12 mu wide; flowers
4.5-5.5 cm; vexillurn pubescence of appressed hairs
mainly along nerves; leaflets acuminate, acumen
0.5-1 cm; stipules deciduous; stipels short, 2-5 mm;
tree. (Ecuador) 3. andrei
6. Bracteoles 15-22 mm long, 7-10 i. ,i wide; flowers
4-4.5 cm; vexillurn pubescence or a densely appressed,
tawny-silky appearance; leaflets obtuse to abruptly

minutely acuminate, acumen to 0.5 cm; stipules
persistent, 8-11 mm; stipels long, 5-10 mm; tall
shrub. (Peru) 4. C. juinensis
2. Flowers large, 6-8 cm; vexillum pubescence sparsely appressed
to glabrate; stipules short, 2-5 mm; legume width narrow, 1.3-
1.9 cm; stipe elongate, 2.5-5 cm; bracteole broader, 10-16 mm
wide. (Amazon Basin: Brazil) 6. C. amazonum
1. Calyx pubescence pilose; calyx tube short 10-12 mm; flowers small,
3-4 cm; bracteoles short, 11-13 mm; bracts short, 3-4 mm. (Bolivia
& the state of Acre, Brazil) 7. C. nervosa
1. Clitoria arbórea Hoffm. ex Benth. , Ann. Hein. Mus. Natur.
2: 115. 1837.
Vexillaria arbórea Hoffm. ex Benth., l.c., pro syn.
Clitoria hoffmanseggii Benth., Journ. Linn Soc. 2: 43. 1858;
nom. superf1.
Ternatea arbórea (Benth.) Kuntze, Riv. Gen. PI. 1_: 210. 1891.
Tree, 3-18 (27) m tall, to 46 cm d.b.h.; pubescence typically of
short appressed hairs to glabrate. Buttress roots (vide Schultes &
BJack 8272). Branches hoi 1ow, to 11 mm thick, juvenile branches
uncinate-pubescent with some appressed hairs, becoming glabrous with
age, internodes 1-6 cm; bark brown, prominently longitudinally striated;
buds 6-8 mm long; leaf scar el 1iptical-orbicular, two shallow
longitudinal depressions near margin, bundle scars inconspicuous.
Leaves 3-foliate, coriaceous, leaflets broad elliptic to obovate, apex
generally obtuse and abruptly short acuminate, acumen to 0.7 cm, rarely

202
to 2 cm, occasionally retuse, base rotund, midrib impressed above,
primary nerves of (8-) 10-14 pairs, dark green and glabrous above,
occasionally with minute hairs along midrib, green and glabrate below
with appressed hairs along nerves, rarely spreading S more conspicuous,
5-17.5 cm long, 3-10 cm wide, terminal leaflet largest. Petiole
angular terete, 3-11 cm, with uncinate and short macroscopic trichomes;
rachis compressed, 1.5-3.5 (4) cm. Petiolules subquadrangular, dark-
shaded, uncinate - pubescent, (4) 6-11 cm. Stipules deciduous, elongate,
triangular, acute to acuminate, uncinate and strigo-e, ciliate toward
apex, 5-10 mm long, 2 mm wide; stipels deciduous, linear, acute,
uncinate - pubescence, 5-11 mm long. Inflorescence axillary and
terminal, solitary to few fascicled, paniculate, liqnose, multi-
flowered, rufo-strigose and uncinate, central axis .-30 cm, primary
branches which bear the pedicels conspicuous, striate, arcuate, broader
near apex, 5-30 mm. Pedicels paired, 4-7 (8) mm. bracts at base of
primary branches ovate, acute, concave around branch, persistent,
reflexed in age, 4-6 mm long; bracts at base of pedicels with
persistent middle pair, ovate, acute, concave, 5-8 mm long, 2.5-4 mm
wide, outer bract deciduous, 2 mm x 0.4 mm, erect, inner bracts not
observed. Bracteole large, coriaceous, subequaling the calyx, broad
oblong-elliptic, obtuse or often splitting at apex and appearing
toothed, uncinate and sparsely strigose, 17-26 mm long, 6-11 mm wide.
Rowers purple to violet, medium size, 4-6 cm. Cal;x uncinate with
scattered, short strigose hairs, tube 13-18 mm long. 4-7 mm wide at
base to 8-11 mm wide at throat, lobes deltoid, actué, 4-7 mm long,
3-4 mm wide, ventral lobe 5-8 mm x 2 mm. Vexillum ..bovate, pubescence
tawny appressed, moderate, 2-4 cm wide, claw 6-9 mm Alae extended

203
beyond carina 6-8 mm, blade 16-27 mm long, 4-8 mm wide, claw 13-16 mm,
conspicuously uncinate-pubescent near blade. Carina falcate, 8-13 mm
across, 4-5 mm wide, uncinate pubescent, claw 24-33 mm. Staminal
tube glabrous, 26-33 mm, free filaments 3-5 mm; anthers 1.5-1.75 mm
long, 0.4-0.6 mm wide. Gynophore 4 mm; ovary, 7-9 mm long, 1 mm wide
bearing dense appressed hairs; style 16-21 mm long, geniculate 6-7 mm
from distal end, beard of stiff trichomes; stigma capitate. Legume
stipitate, slightly exerted beyond calyx, green to brown, glabrate
with uncinate and short, erect trichomes along sutures, flat, thick-
coriaceous, to 3-4 mm thick, sutures very thickened, slightly wing-like,
weakly raised around seeds, nearly straight, obliquely positioned on
stipe, 10-24 cm long, 2.3-4 cm wide, narrower at bare, 8-15 mm wide;
beak 4-7 mm; dehiscence causing valves to twist one guarter of a turn.
Seeds with nearly orbicular faces, compressed, 2 mm thick, black to
dark brown, smooth, glabrous, 9-17 mm long, 10-14 mm wide, 7-11 seeds
per pod, hilum oblong, 3 mm x 1.5 mm. Figures 14 and 15.
Hoffmannsegg1s Clitori a is characterized as a tree of the Amazon
Basin with paniculate inflorescences bearing medium sized purple-violet
flowers, large coriaceous bracteoles subequal to caiyx, and large,
broad legumes with thickened wing-like sutures.
PHENOLOGY: The species has been collected in tlower from April to
July and September to January. Fruits were collected in September,
December, and May to July.
TYPE COLLECTION: BRAZIL. Para, Sieber s.n. (I ECT0TYPE: BR-8!
Holotype probably at B, destroyed in war).
Bentham originally published the name C^ arbórea in 1837 based
upon Hoffmannsegg1s name which Bentham cited as "Ve-i11 aria arbórea

Figure 14. C1itoria arbórea - I. Var. arbórea: (a-b) leaflets,
x 1; Ce) juvenile inflorescence, x 1; (d) fie,,ver, x 1;
(e) complicate vexillum, x 1; (f) ala and carina, x 1;
(g-h) calyx with staminal tube, x 1; (i) gynoecium, x 1
(j) calyx, x 1. (Duarte 6922, FLAS 122958: a.
Krukoff 8104, NY: b-jj

205
-

Figure 15. Clitoria arbórea - II. Var. arbórea: (a) fruit, x 1;
(b) x-section of fruit, x 1; (c) three views ..if seed, x 1;
(d) leaf scar, x 2; (e) stipule scar, x 2. Var.
longiramosa: (f) smaller inflorescence, x 1; (g) leaflet,
x 1. Var. pseudoamazona: (h) smaller inflorescence,
x 1; (i) leaflet, x 1. (Dahlgren & Sel 1 a 107, F 601 565:
a-b,d-e. Krukoff 5335, S: c. Krukoff 1689, NY: f-g.
Krukoff 5719, NY: h-i.)

¿oz

208
Hoffm. MSS," in synonymy. The only specimen cited (the type) was
"Para, Sieber." Stafleu (1 967, p. 206) indicated t/at Hoffmannsegg's
types were most probably placed at Berlin. The Benin legume collection
was destroyed by fire during World War II, and presumably the type for
V. arbórea was lost. A Bruxelles specimen believed to be an isotype
was selected as the lectotype. The Bruxelles specimen had the same
data cited by Bentham. In addition, in an iron based ink faded brown,
the name "Vexi 11 aria n." was written on one line with "arbórea H."
written on the next line. The script in brown ink below reads:
"Clitoriae monini huic substituend putavit Hoffmgg-Sed hoc genus a
Clitoria diversum videtur" or a Clitoria name was substituted.
Hof fiuanuseyg had thought that this genus (Vexillariu ) was different
from Clitoria. The word "Clitoria" written in blaci ink beginning to
fade brown was placed in front of the tern, "arbórea H." The Sieber
specimen was originally identified as Vexillaria arbórea Hoffm. and
later annotated to Clitoria. Hoffmannsegg describen the genus
Vexi11 aria (presently in synonymy with Centrosema) in 1824 and
included one species. There is no record of any. additional names
published, except for V. arbórea and V. triqueta published by Bentham
in 1837 in synonymy. The Bruxelles specimen is believed to be an
isotype, and thus selected as the lectotype.
In 1858, Bentham published the superfluous name C_L hoffmanseggii,
citing "C^ arbórea, non C. arborescens Ait.." in synonymy and including
Sieber as one of the three collections cited. Although similar in
spelling, the terms arbórea and arborescens are distinct such that
confusion should not result between them. The name C_. hoffmanseggi i
was invalid since it included the type of a prior n.ined species.

209
The Riedel and Mathew specimens cited by Bentham with the
publication of the name C^ hoffmanseggi i have been cited as types.
However, since Bentham had substituted the name for C. arbórea and
included its type, "Brazil, Para, Sieber," then the only type is the
Sieber specimen, and the Mathew & Riedel specimens are merely
additional cited specimens. Neither the Mathew specimen from
Moyobamba (K!) nor the specimen from Tarapoto (CGE!E!K!) match the type
Sieber specimen. Of several Riedel specimens, Riedel 163 (K!) most
closely matches the type and may be the one Bentham referred to in his
citation. This specimen was placed in a type folder and bears the
words "Hb. Mus. Petrop." cited by Bentham as the location of Riedel
specimen. This specimen also bears the name "C. racemosa Benth., Ann.
Mus. Vind. 2: 114" and has been distributed via photograph1 as the type
of C^_ racemosa, which it is not, and in addition has not been cited for
the species. This name represents an early misidentification.
VERNACULAR NAMES: BRAZIL (ACRE): Ingapia, Krukoff 5335.
COLUMBIA: nai-chee-ke (Tikuna name), Schultes & Black 8450. PERU:
Burro yacuana, Woytkowski 6253.
ECONOMIC IMPORTANCE: In Peru, the leaves were reported as
bluish-brown below and used as a dye for coloring textiles (Woytkowski
6253). The seeds were noted as alkaloid positive (Schultes et al.
24120), although the specimens examined lacked fruits and seeds.
NOTES: Cl itoria arbórea is often cited in early literature as
closely related to C. amazonum or a possible variety of that species.
However, when compared, there are too many characters which segregate
them to consider reducing these two species to varieties of one species.
Cl i tori a arnazonum is easily distinguished by its larger flowers

210
(6-8 cm), short stipules (2-5 mm), and narrow legume (width less 2 cm)
on elongated stipes (2.5-5 cm). These two species have often been
misidentified for each other in the past.
Cl itoria fairchildiana (syn. racemosa Benth.) is also frequently
cited as having possible affinities. Although it is more closely
related to arbórea than C^_ amazonum, C. fairchildiana is a distinct
species, easily segregated by its more elongate, subpaniculate
inflorescence with short primary branches (1-5 mm), narrower oblong-
elliptic leaves, drooping branches which reach the ground, and coastal
habitats.
There are three varieties which can be distinguished by their
inflorescences, leaves, and pubescence of the calyx and vexillum. Some
labels bear annotations "new sp?" but are unnamed. These collections
are best placed at the level of variety because of the large number of
characteristies that agree with C. arbórea.
DISTRIBUTION (Figure 16): The species is found in scattered
locations in forests along the Amazon River and its major tributaries
in Brazil and Peru, and the southeastern apex of Amazonas, Colombia.
The only elevations cited were the Schultes collections in Colombia
as 100mand Woytkowski collections of Peru as 600-800 m for the typical
variety. Elevations of 1200-1600 m were cited for the nontypical
varieties.
KEY TO VARIETIES:
1. Leaflets obtuse to abruptly short acuminate, acumen 0.3-5 mm,
glabrous above, lower surface inconspicuously pubescent of short,
appressed hairs on major nerves (vidi 20-30x); calyx pubescence of

211
scattered appressed hairs; vex i 11um pubescence of moderately
dense, appressed trichomes.
2. Primary branches of inflorescence that bear pedicels short,
5-9 mm; inflorescence short, 2-7 cm (9 cm in fruit) ....
la. var. arbórea
2. Primary branches of inflorescence that bear pedicels
elongate, (7) 10-30 cm lb. var. 1ongiramosa
1. Leaflets long acuminate, acumen 7-15 mm, with short pubescence on
midrib above, lower surface with conspicuous, spreading hairs along
major nerves (vidi 10 x); calyx nearly glabrate, appressed hairs
more densely packed on ventral edge; vexillum pubescence with
appressed hairs concentrated along folded edge, subglabrate
elsewhere or following nerves lc. var. pseudoamazona
la. Clitori a arbórea var. arbórea
Vex ill aria arbórea Hoffm. ex Benth., l.c., pr£ syn.
Cl itoria hoffmanseggii Benth., Journ. Linn. Soc. 2: 43.
1858; nom. superfl.
Ternatea arbórea (Benth.) Kuntze, Riv. Gen. PI. 1_: 210.
1891.
Tree. Leaflets generally obtuse or abruptly short acuminate,
acumen to 5 mm, seemingly glabrous but bearing minute, perpendicularly
appressed trichomes to major nerves below. Inflorescence short, 2-7
(9) cm, primary branches that bear pedicels 5-10 mm. Calyx pubescence
of short, scattered appressed hairs. Vexillum pubescence moderately
dense over surface.

212
DISTRIBUTION (Figure 16):
BRAZIL. AMAPA: Macapa, estrada para Fazendinha, 19 Jul 1951,
Froes & Black 27487(NY); Cachoeira de S. Antonio do Jari em Ilha,
20 Nov 1967, Oliveira 3714 (NY); Coastal region, 0°17'N-51°5'W,
7 Jul 1962, Pires & Cavalcante 51990 (NY,U). PARA: Cacaual Grande
perto da Ilha do Amana, 1 Nov 1950, Black & Ledoux 50-10500 (NY); Boa
Vista on Tapajos River, May-Jun 1929, Dahlgren & Sella 107 (F);
Almeirim, 11 Apr 1903, Ducke 3461 (BM, RB-mixed with C. s i nip 1 i c i f o lia);
Alenquer, 31 Jul 1903, Ducke 3765 (RB); l.c., 27 Dec 1903, Ducke 4925
(BM,RB-2 sh.); chemis de fer d'Alcobaca Station Breu Branco, Rio
Tocantins, 1 Jan 1915, Ducke 15576 (RB-2 sh.); Estrada de Ferro
Tocantins, 25 Sep 1948, Froes 23523 (A,NY). AMAZONAS: Esperanca, ad
ostium fluminis Javari, 26 Oct 1 945, Ducke 1825 (A,F,NY); Carnation,
border of river, 25 Jan 1949, Froes 24006 (NY); ad Rio Madeira, Riedel
163 (K, S-photo of K); Mun. Humayta, near Tres Casas, Basin Rio
Madeira, 14 Sep-11 Oct 1934, Krukoff 6125 (A,BM,F,G-2 sh.,MICH,M0,NY,U);
Mun. Sao Paulo de Olivenca near Palmares Basin Rio Solimoes, 11 Sep-
26 Oct 1936, Krukoff 8104 (A,BM,F,G,MICH,M0,NY,S,U). RONDONIA: Basin
Rio Madeira, 167-9 km, Madeira-Mamore RR., near Mutumparana, 6 Jul 1968,
Prance et al. 5610 (NY,US). ACRE: Rio Branco, 17 Jan 1944, Baldwin
3170 (US) and 3171 (US); Foz do Macauhaw, Rio Purus, (trib. Rio Yaco),
9°20'S-69°W, 9 Aug 1933, Krukoff 5335 (A,BM,F,G-2 sh.,LA,M,MICH,M0,NY,
S,U,UC,US,WIS).
C 0 L 0 M B I A. AMAZONAS: Trapecio amazónico Loretoyacu River,
100 m, Sep-Nov 1944, Schultes 6122 (GH,NY,US); l.c., 20-30 Oct 1945,
Schultes 6611 (GH,US); l.c., Sep 1946, Schultes & Black 8272 (F,US);
l.c., Oct 1946, Schultes & Black 8450 (GH,US); l.c., Leticia, Sep-Nov

Figure 16. South American distribution of three species of section
Bractearia. Clitoris arbórea var. arbórea (•), var.
longiramosa (★}, var. pseudoamazonic a Uáki ; f. andrei
I'A'); C: juninensi s

vn
r
214

215
1944, Schultes 6194B (GH); Rio Amazonas, near mouth of Rio Loretoyacu
& Puerto Narino, 13-15 Sep 1 966, Schultes, RaffojjJ , & Soejarto 24120
(US).
PERU. LORETO: Iquitos, 100 m, 29 Sep 1929., Kill ip & Smith
30000 (A,NY,US); Stromgebiet des Maranon von Iquitos aufwarts his zur
Santiago-Mundung am Pongo de Manseriche, ca 77 30'W, San Isidro, 1 924,
Tessman 4953 (F-frag., G). SAN MARTIN: ca 900 m, 26 Nov 1936,
Woytkowski 25 (F); Rioja, 800 m, 28 Dec 1960, Woytkowski 6062 (GH,UC,
US); l.c. , 10 Feb 1961, Woytkowski 6253 (M0).
Ib. Clitoria arbórea var. 1ongiramosa Fantz, var. nov.
Tree. Leaflets generally obtuse or abruptly short acuminate, acumen
to 5 mm, broad elliptic to obovate, seemingly glabrous below but with
minute appressed hairs along major nerves. Inflorescence elongate, (7)
10-25 (30) cm, primary branches of inflorescence elongate, 12-30 mm.
Calyx pubescence of short, scattered appressed hairs. Vexillum
pubescence moderately dense.
TYPE COLLECTION: BOLIVIA. Guanai , 2000 ft., May 1886, Rusby 2399
(H0L0TYPE: NY-"Columbia College Herbarium." Isotypes: F 164432, GH,
MICH-2 sh.,NY-2 sh.,PH,US 76908).
The Rusby 2399 collection was the only one with both mature fruits
and flowers which indicated its affinities with C. arbórea. The
Columbian specimen now deposited at New York was selected as the
holotype because it had a mature fruit, several inflorescences which
illustrated the elongated primary branches, cálices in place, plus a
packet with flowers. None of the other specimens has this combination.

216
The other New York sheets have less material, as does each Michigan
sheet. The Gray Herbarium specimen lacks inflorescences, whereas the
Field specimen's primary branches have been aborted. Others either
lack fruits or flowers or inflorescences.
DISTRIBUTION (Figure 16): The three collections are widely
scattered within the range of the typical variety, but the only cited
elevation places the variety at 1200-1600 m.
BRAZIL. MATO GROSSO: source Jatuarana River, Machado
River region, Dec 1931, Krukoff 1689 (Paratypes: A,BM,G-2 sh., MICH,
MO,NY,S,U,UC). PERU. SAN MARTIN: Zepelacio, near Moyobamba,
1200-1600 m, Jan 1 934, K1 ug 3505 (Paratypes: A,F,GH,MO,NY,S).
lc. Clitori a arbórea var. pseudoamazonica Fantz, var. nov.
Tree. Leaflets ovate-elliptic, acuminate, acumen 7-15 mm,
conspicuously pilose on nerves below. Inflorescence elongated, 10-30
cm, primary branches which bear pedicels medium length, 8-15 mm.
Calyx glabrate, appressed hairs concentrated alona ventral margin.
Vexillum subglabrate with few appressed hairs along nerves and
concentrated near folded margin.
TYPE COLLECTION: BRAZIL. Amazonas, Mun. Sao Paulo de Olivenca
near Palmares, Basin Rio Solimoes, 11 Sep-26 Oct 1936, Krukoff 8208
(H0L0TYPE: US 2169816. Isotypes: A,F 927714, G 295 & 296, MICH, M0
1175978 & 1250450, S,U 38199A).
Both collections are nearly equivalent in their inflorescences
and flowering material, and lack of fruit. The Smithsonian specimen of
Krukoff 8208 was selected as the type because it had more flowering
material and was a typical representative of the variety.

217
NOTES: The two collections of this variety weie identified as
C. amazonum or near amazonum which they superficially resemble in
leaflet shape and pubescence, inflorescence, and flowers. However, a
closer examination would indicate that these specimens are distinguished
from C. amazonum by: its smaller flowers, bracteolcs, calyx, gynoecium,
and androecium; its different pubescence on the leaves, calyx and
vex ill urn; and its larger stipules, stipels and inflorescence. Other
botanists have reached the same conclusion, as evidenced by their
annotations, but they disagree on its placement. These specimens
have been included in several species, including arbórea, C.
arborescens, C. glaberrima, C. javitensis, and C. racemosa (=C.
fairchildiana). Sandwith (UC 606482) raised the question of a new
species in his annotation. However, there are too many characteristics
which agree with C_._ arbórea to warrant its segregation from the species,
even though the pubescence of some structures varies and the leaves
seem to be quite different. The longer primary branches of the
inflorescence suggest an affinity with var. 1ongiramosa. But it
differs in the characters previously noted, and doe: not agree with
the length of the primary branches of var. 1ongiramosa, being con¬
sistently shorter. Therefore, the better placement is to treat these
collections as a separate variety of arbórea.
DISTRIBUTION (Figure 16): _Z_I L. ACRE: Near mouth of Rio
Macauhan, Basin Rio Purus, 9°20'S-69°W, 27 Aug 1933, Krukoff 5719
(Paratypes: A.BM-2 sh.,F,G-2 sh.,LA,M,MICH,MO,NY-2 sh.,PR-2 sh., PR-2
sh.,S,U,UC).

218
2. Clitoria fa i re hi 1di ana Howard, Baileya 1 5: 16. 1967.
C1 itoria racemosa Benth., Ann. Wein. Mus. Hatur. 2: 115.
1837, non G. Don (1832), nec Sesse & Hocino (1889);
nom. inval^
Terna tea racemosa (Benth.) Kuntze, Riv. Gen. PI. 1: 210.
1891.
Centrosema spicata Glaziou, Bull. Soc. Bot. France 53,
Mem. 3b: 135. 1906; nom. nud.
C1itoria racemosa Benth. f. obovatifolia Rizz., Arq. Jard.
Bot. Rio de Jan. |7: 190. 1963.
Neurocarpum racemosum Pohl, nom. m schedula.
Tree, commonly 3-8 m tall to occasionally reported 26 m tall, to
40 m in cultivation. Branches arching virgate, drooping to touch
ground to form a canopy, pith solid, to 8 mm diameter, juvenile
branches subquadrangular becoming terete, pubescence moderately
uncinate and of minute erect hairs, ca 0.25 mm long, becoming glabrate
with age; bark brownish, peeling in longitudinal strips, a grayish-
white beneath; axillary buds 2-3 mm; leaf scars sulorbicular, bundle
scars inconspicuous, stipule scar falcate. Leaves 3-foliate,
coriaceous, leaflets lanceolate-elliptic to oblong-elliptic, sometimes
the terminal leaflet oblanceolate to lanceolate-obovate, apex narrowing
to an acumen 0.5-2 cm long, or obtuse and abruptly short acuminate in
the obianceolate-obovate leaflets, extreme apex more or less mucronate,
base broadly cuneate, midrib impressed above with oiten minute semierect
and uncinate hairs present, primary nerves of 14-19 pairs, upper surface
bright green, shiny, glabrous, lower surface pale green, with dull,

219
minute appressed hairs, ca 0.25-0.5 mm, more conspicuous and sometimes
longer on major nerves, 6-18 (22) cin long, 3-7 cm wide, the terminal
leaflet larger; juvenile leaves commonly collected, complicate, upper
surface purplish, lower surface with dense white, ai.pressed, ascending
hairs, more persistent along nerves. Petioles compressed, weakly
caniculate, pubescence of uncinate and short erect hairs, 3-7 (9) cm;
rachis more conspicuously compressed, 1.5-3 (4) cam. Petiolules dark
brownish-black, subquadrangular, with dense erect tiichomes, rusose,
5-9 mm.' Stipules deciduous, lanceolate, glabrate, minutely ciliolate,
apex gradually narrowed to stiarp point, 6-12 mm long, 2-3 mm wide;
stipels deciduous, stiff, linear, often 5-nerved, Sparse pubescent,
5-12 mm long, 0.5-1 mm wide, terminal stipels usually shorter than
lateral. Inflorescence axillary or terminal, rarely bearing minute
leaves near base which quickly abort, solitary, mulLi flowered,
subpaniculate, all axes with dense minute hairs, erect and uncinate,
reddish-brown, central axis twisting, subquadrangular-compressed,
becoming terete with age, weakly caniculate, 8-42 cin long, primary
branches which bear pedicels subsessile to 4 mm. Pedicels paired, 3-8
mm long to 4 mm thick in fruit, slightly darker in color, uncinate-
pubescent. Bracts subtending primary branches solitary, semipersistent,
reflexed in age, linear-lanceolate, 6-10 mm long, 2-3 mm wide; bracts
subtending pedicel 2-3, reflexed in age, ovate, acute, short strigose
and ciliate, 4-8 mm long, 2-4 mm wide, middle pair persistent, outer
bract deciduous. Bracteoles persistent, coriaceous oblong to elliptic,
obtuse, sometimes apiculate, reddish-brown, minutely striated, pubescence
of minute appressed hairs and ciliolate, large subequal and enclosing
calyx, sometimes splitting near apex to form 2-4 teeth or lobes, (13)

220
17-23 mm long, 7-11 mm wide, inserted 1 mm below calyx. FIowers
violaceous to rose, medium size, 4-6 cm. Calyx green, pubescence of
scattered, minute, appressed hairs, tube 11-16 (20) cm long, 3-6 mm
wide at base to 7-10 mm wide at throat, lobes ovate deltoid, acute,
4-6 mm long, 3-4 mm wide, ventral lobe 5-8 mm x 2 mm, linear with
abruptly acuminate apex. Vexillum pubescence appressed tawny, denser
towards complicate fold, 3-4.5 mm wide, claw 6-8 mm. Alae oblong-
falcate, extended beyond carina 5-7 mm, blade (15) 18-23 mm long, 5-8
(10) mni wide, claw 12-18 mm. Carina white, falcate, 9-11 mm across,
3-4 mm wide, claw 15-25 mm. Staminal tube glabrous, 25-32 mm long,
free filaments 2-4 mm; anthers 1.25-1.5 mm long, 0.5-0.8 mm wide,
Gynophore 3-4 mm; ovary pubescence dense, white appi essed hairs,
especially vexillarly, 14-17 mm long, 1 mm wide, style beard dense,
12-16 mm long, geniculate last 5-8 mm; stigma capitate, ca 1 x 0.5 mm,
dark-colored, short hairs around base. Legume stipitate, shortly
exerted beyond calyx, green drying brown, flat, glabrate with minute
hairs on sutures, sutures very thick, valve straight to weakly falcate,
12-28 cm long, 2.2-3.3 cm wide, base narrow, often 8-12 mm wide; stipe
13-22 mm; beak to 8 mm; dehiscence causing valves to twist one-quarter
of a turn to one turn. Seeds nearly orbicular, black, smooth,
compressed to 1-2 mm thick, 11-17 mm long, 11-16 mm wide, 5-12 seeds
per pod; hilum elliptic-oblong, ca 3 x 1.5 mm. Seen germination
hypogean. Figures 17 and 18.
Pohl's Clitoria is characterized as a tree of coastal areas with
elongate, racemose inflorescences of numerous medium-sized violaceous
flowers, and shiny-green elongated leaves borne on a roo ping branches.

Figure 17. Clitoria fairchi 1 diana - I. (a-b) leaflets, ;< 1; (c)
inflorescence, x 1; (d) flower, x 1; (e) ala and carina,
x 1; (f) gynoecium, x 1; (g) androecium, x 1. (Constantino
11857, S: a. Huber 9475,‘BM: b. Killip & Smith 30396,
US 1463497: c-d. Pohl 2375, W: e-g.[

222
1 cm

Figure 18. Cl i tori a fairchiidiana - II. (a) stipule, x 1; (b) leaf
scar and stipule scar, x 1.8; (c-d) fruits, / 1; (e) three
views of seed, x 1; (f) calyx, x 1. (Pi res 3110, NY:
a-b. Huber 9475, BM: c. Kill ip & Smith 30396, US
1463497: d,f. Jimenez 5834, US 2615828: e.)

224

225
PHENOLOGY: This species has been collected in (lower mainly from
August to February, with isolated field collections in May & June and
more commonly from cultivated plants. Fruits appear from October to
February with occasional collections in May or June.
TYPE COLLECTION: BRAZIL: Natividade, 1836(?). Pohl 2375
(LECTOTYPE: W. I solectotypes: F 875105, K-not seen, NY,S-photo of K,
and ?M-96).
Bentham described the species and cited only one collection as
"Ad Natividade, Pohl." The Hein specimen bears the name Neurocarpum
racemosum, which has had a line drawn through it, and "Clitoria racemosa
Benth. ipse" written above the name. The New York specimen is a
duplicate from Hein with the pair of names and the rlirase "fide
Bentham" written in a blue ink, much fresher than ti.e Hein specimen,
and presumably labeled at a later date. The ink does not appear to be
the iron-based ink used on the Hein specimen. All other collections
examined bear only Bentham's name and lack any indie ation that Bentham
may have examined them. The Kew specimen is marked "Hb. Benthamianum,
1854," thus probably was observed by Bentham; however, the specimen was
placed in Bentham's herbarium seventeen years after lie had described
the species and is thus an isotype. The Hein specimen was selected as
the lectotype because it is more probably che type specimen from which
Bentham described his species.
Bentham neglected to include the collection number "2375/' which
is present on the Hein label and its duplicates. Bentham had failed to
cite collection numbers in other species of the same publication.
The date 1836 appeared only on the label data of the Kew specimen.

226
Bentham's name was a homonym for G. Don's name published in 1832,
and is thus invalid. Howard (1967) correctly noted this and proposed
the substitution of the name C. fairchild iana (after David Fairchild
whom Howard believed introduced the species into th Montgomery
Foundation Property at the Fairchild Gardens in Miami, Florida, USA).
Howard cited the type specimen of Pohl, but a year later distributed
specimens from the tree on the Montgomery Property as the "type tree"
(Howard 17052), which it is not. The type tree is in Natividade,
Brazil,' and was probably a cultivar since it was located outside of
its coastal range.
Glaziou (1906) had published the nomen nudum, (entrosema spicata,
which has not been synonymized previously. A Few specimen (K-427!)
bears the label "no. 9738, comm. M. Glaziou, Dec. 1. 78. Cultivated at
Rio Janeiro" and a notation in pencil "Centrosema spicata Glaziou,
n. sp., nomen in 'Liste' ex no." The specimen is CIitori a fairchi1 diana,
which is commonly cultivated in Rio de Janeiro, and may possibly be the
type of the nomen nudum, Centrosema spicata Glaziou Therefore, it is
included as questionable in the synonymy of C. fair;hi Idiana.
VERNACULAR NAMES; BRAZIL: Fa cao, Pi res & Cav,, 1 cante 52540
(Amapa); Faveira or Palheteira, Ducke 350 & 614 (Para).
ECONOMIC IMPORTANCE; The species is cultivated as an exotic
ornamental tree, particularly in Rio de Janeiro, less commonly in other
areas.
NOTES; Rizzini (1 963 ) described f. obovatifol la based upon one
specimen (Schwaeke 245, holotype: RB 38376!) with the leaflets in a
packet, some of which were larger and obovate. This author has seen
other collections in which the same plant will have some terminal

227
leaflets obovate, with the other terminal leaflets more typical. With
the variation on a single plant and the narrow range of distribution,
this form is not recognized.
DISTRIBUTION (Figure 19): The species is endemic to the deltas of
northeastern Brazil, although it is cultivated in other parts of Brazil,
the West Indies, Florida (USA), and Philippines.
BRAZIL. WITHOUT LOCALITY: BurcheH 8084 (GH,K) and 8227
(GH,K); Pohl (M). AMAPA: Calcoene, coastal region, 21 Aug 1962,
Pires & Cava!cante 52540 (F,G,GH,MICH,NY,RB-2 sh.,VEN). PARA: Morajo,
in ripa fluv. Laranjeiras, 1877, Schwacke 245 (RB); IIha do Mosqueiro,
27 Jun 1930, Bastas 13 (FLAS, RB-not seen); l.c., 17 Oct 40, Ducke
350 (S) and 6J_4 (F ,M0 ,NY ,UC); l.c., sandy coast, 3-9 Nov 1 929, Ki 11 ip
& Smith 30396 (NY,US); l.c., ad fluv. Para, cult. Jard. Bot. Rio de
Janeiro, 11 Feb 1928, Constantino 11857 (RB,S,U); arredores do Belem,
Oct 1906, Snethlage 7740 (RB); cult. Belem, 27 Jan 1952, Black 52-14147
(P); l.c., grounds Museum Goeldi, 2 Oct 1961, Egler & Jrwiri 46712
(NY-2 sh.); margem do Rio Guama, 11 Jan 1951, Pires 311_0 (NY); l.c.,
Sep-0ct 1961, Pires 51923 (NY,U); margin Rio Tocantins, 14 May 1951,
Froes 27068 (NY); II ha Arapiranqa, Dec 07, Huber 9475 (BM,G-2 sh.);
near Trapiche Hypolito, 24-5 Aug 1934, Krukoff 5877 (A,BM,F,MICH,M0,
NY,S,U,US); Tapana (?), Hoffmannsegg s.n. (HAL). MARAHAHAO: Rio
Itapecuri-Parnahi ba, Dec 1884, Cajaaneura s_._r\. (RB); Rio de Pedreiras,
22 Jun 1909, Anonymous 2286 (BM,G,RB). GUANABARA: cult. Rio de
Janeiro, Glaziou 9738 (K); l.c., Kuhlmann s.n. (RB-2 sh.); l.c.,
2 Dec 1 934, Lage 152 (RB); l.c., 17 Mar 1 936, Lage RB-81425 (FLAS,RB-
not seen); l.c., 8 Nov 1960, Costa RIP-166600 (FLAS,RB-not seen); l.c.,
25 Feb 1932, Pessoal 2414 (RB-2 sh.); Rio de Janeiro pr. Reserva

Figure 19. South American distribution of four species of section
Bractearia. Clitoria amazonum f. amazonum {A , f.
vu 1 gar is (#); fa ire hi! diana (je); C_. moyob nubensis
TO) ’ C- nervosa (☆)'•

229

230
Biológica, 28 Dec 1965, Moreira 85 (US); Rio de Jan iro, Itanhanga
Golf Club, Feb 1962, Silva 24 (FLAS,RB-not seen).
V_£ N El U E L A. cult, in Parque del Este, (iracas, Jun 1 972,
Steyermark s.n. (US).
D 0 M I N I C AJI R E P U B L I C. cult., Estancia Nueva,
Moca, Espaillat prov., 28 Jun 1 970, Jimenez 5834 (NV-2 sh., US-2 sh.).
UNITED STATES. FLORIDA: Dade Co.: Miami, cult.
Mountgomery Foundation, 28 Oct. 1968, Gjl 1 i_s 7030 (S); cult. Fairchild
Trop. Garden, 1 Jul 1 969, Gil 1 is 8311 (US); cult. Jenning Estate,
30 Jun 1968, Howard 1 7052 (HAL-2 sh.). Martin Co.: cult., 14 Jun
1 967 , Campbe11 s.n. (FLAS).
P H I L I P_P I N E_S. cult, some gardens, 28 cpr 1 965, Pancho
6572 (US).
3. Clitoria andrei Fantz, sp. noy.
Tree, 20 in tall. Brandies subquadrangular, becoming terete, to
7 mm diameter, pith hollow, juvenile short appressed to spreading hairs,
becoming glabrate; bark light brown; axillary buds i .5-2 mm, bud scales
ovate, acute, strigose, concave; leaf scar suborbicnlar, bundle scars
inconspicuous, stipule scar falcate. Leaves coriaceous, 3-foliate,
leaflets ovate-lanceolate to elliptic-lanceolate, apex abruptly
acuminate, acumen 1-2 cm, mucronate, base broad cunt ite to rotund,
midrib impressed to subimpressed above, often bearing short erect hairs,
primary nerves 13-17 pair, upper surface dirk green, glabrous, lower
surface pale green, short subvelutinous with semiappressed hairs along
nerves, 7-2Ü cm long, 3-8 cm wide. Petiole subangular-terete to

231
compressed above, 4-7 cm, pubescence of short spreading hairs, rachis
subquadrangular, 1-2.5 cm. Petiolules quadrangular medium-colored,
dense falcate hairs, 6-7 mm. Stipules deciduous, deltoid, acute, 4 mm
long, 3 mm wide; stipels deciduous, linear, pubescence short, appressed,
2-5 mm long. Inflorescence elongate, subpaniculate-racemose, axillary
or terminal, solitary, multiflowered, axes with dense, short rufous-
trichomes on juvenile parts, becoming less dense and whitened with age
towards base, central axis 7-20 cm, primary branches which bear
pedicel's subsessile to 2 mm. Pedicels paired, 6-8 mm. Bracts ovate,
acute, dense pubescence of rufous trichomes, reflexed in age before
dropping, 4-6 mm long, 1.5-2 mm wide; middle pair persistent; outer
bract narrow, semipersistent. Bracteoles large, conaceous, conspicuous,
oblong, obtuse subequaling the calyx, pubescence appressed, 22-25 mm
long, 9-12 mm wide. Flowery "1 ivido-roseis," 4-5 cm. Calyx pubescence
of scattered, short appressed hairs, tube 14-16 mm long, 4-5 mm wide at
base to 9-11 mm wide at throat, lobes deltoid, 5-7 mm long, 2-3 mm
wide, ventral lobe 6-8 mm. Vexillum indumentum of short, tawny,
appressed hairs, 2.5-3 cm wide, claw ca 12 mm. Alae extended beyond
carina 6-8 mm, blade ca 16-18 mm long, 4-6 mm wide, claw ca 15-17 mm.
Carina falcate, ca 9 mm across, 3-4 mm wide, claw ca 24-27 mm.
Staminal tube sparsely uncinate-pubescent near ap< 28-30 mm long,
free filaments sparsely uncinate, 2-3 mm; anthers 1.5-2 mm long, ca
0.3 mm wide, connective acuminate. Gynophore 3 mm; ovary pubescence of
dense, appressed, yellowish-white trichomes, ca 14 mi:i long, 1.25 mm
wide; style 17-21 mm, geniculate 6-7 mm from distal end; stigma flat,
discoid, 0.5 mm diameter, base pubescent, erect. Legume unknown.
Figure 20.

Figure 20. Clitoria andrei. (a-b) leaflets, x 1; (c) inflorescence,
x 1; [d) flower, x 1; (e) bracteoles and caly.. with
beginning fruit, x 1; (f) calyx, x 1; (g) gynnecium, x 1;
(h) androecium, x 1; (i) leaf scar, x 1.2; (j) stipule
scar, x 1.2; (k) vexillum, x 1. (Andre 4675, F 537681:
a ,d-k ; GH: b-c.)

233

234
Andrei's Clitoria is characterized as an Eduadwrean tree with
subvelutinous leaves, elongate inflorescence of man; medium-sized
flowers with large, coriaceous bracteoles, and shori stipules & stipels.
PHENOLOGY: One collection was obtained in flower in July.
TYPE COLLECTION: ECUADOR. Inter Guayaquil arid Santa Rosa, 60 m,
July 1876, Andre 4675 (HOLOTYPE: GH. Isotypes: F 537681, K-438,NY).
All four sheets are very similar, but the Gray Herbarium specimen
was selected as the type because it has a dissected flower plus both
flowers' and leaves are attached to the inflorescence and branches, as
opposed to their being placed in packets.
VERNACULAR NAME: Sapotillo, Andre 4675.
NOTES: This collection v/as identified as C. a¡. azonum from which
it differs in many characteristics. It is easily distinguished by the
inconspicuous primary branches of the inflorescence, smaller flowers,
most floral structures, and leaves. Clitoria andrei has its closest
affinities to C. juinensis, a Peruvian endemic, which is a shrub with
smaller bracteoles, conspicuous primary branches in the inflorescence,
and larger stipules and stipels.
DISTRIBUTION (Figure 16): This species is known only from its type
location in Ecuador.
4. Clitoria juninensis Fantz, sp. nov.
Shrub, erect, 2-4 m tall. Branches elongated (vide Kijlip &
Smith 23380), to 10 mm thick, pith hollow, juvenile branches subangular
with densely spreading,rufous hairs, branches darkening with age, terete,
glabrate with uncinate trichomes becoming more conspicuous; bark brown,

235
splitting longitudinally, lightened beneath; axi 11 ai / buds 1-2 mm,
scales glabrate; leaf scar elliptic to shield-shape.!, bundle scars more
or less inconspicuous, broad lyriform, stipule scar falcate. Leaves
3-foliate, coriaceous, leaflets broad elliptic to o,ate-el1iptic to
obovate, apex generally obtuse, rapidly narrowed to u short acumen
1-7 mm long, more or less mucronate, base broad cunéate to rotund,
midrib impressed above, glabrous or bearing short erect or uncinate
hairs, primary nerves of 10-16 (-18) pairs, upper surface slightly
dark green and glabrous, lower surface pale green and subvelutinous,
8-22 cm long, 4-9 (12.5) cm wide. Petiole subangular-terete, ribbed,
with moderately dense, short spreading, rufous tricliomes, 4-11 cm;
rachis more compressed, (1.5) 2-3.5 cm. Petiolule d irk-colored,
subquadrangular, with dense, spreading rufous trich.mies, 6-10 mm.
Stipules deciduous, densely pubescent, rufous, cilinte, lanceolate-
deltoid, gradually narrowing to acute apex, 8-11 mm long, 2-3 mm wide;
stipels semi deciduous, linear, gradually narrowed to acute tip, 5-10 mm
long, 0.2-0.7 mm wide. Inflorescence subpaniculate-racemose, axillary
and terminal, solitary, multiflowered, pubescence of axes short,
moderately dense to dense, spreading, rufous, central axis 5-23 cm long,
lignose and nodose in fruit, primary branches which bear pedicels
subsessile to 5 (8) mm long. Pedicels paired, 4-7 iui. Bracts ovate,
acute, reflexed in age, pubescence dense, oppressed, rufous, ciliate,
4-7 mm long, 2-3 mm wide, middle pair persistent, outer solitary,
slightly narrower, deciduous; bract at base of primary branch solitary,
semipersistent, narrower. Bracteoles large, coriaceous, oblong-elliptic,
subequal calyx, pubescence short, appressed, 14-18 (20) mm long, 7-10
mm wide. Flowers white to pink vexillum tinged blue- with a blue carina,

236
4-5.5 cm, calyx pubescence appressed, tube 13-17 mm long, 3-5 mm wide
at base to 8-12 mm wide at throat, lobes ovate-deltoid, acuminate,
5-7 mm long, 3-4 mm wide, ventral lobe 7-8 mm, narrow. Vex i 11um
pubescence tawny, short appressed, 3-3.5 cm wide, claw ca 9 mm. Alae
white-pink, extending beyond carina 6-8 mm, blade 18-23 mm long, 5-9
mm wide, claw 13-15 mm. Carina blue, falcate, 9-13 mm across, 4 mm
wide, claw 24-27 mm. Stamina! tube glabrous, 27-32 mm, free filaments
3-5 mm; anthers 1.5-1.75 mm long, 0.5-0.75 mm wide. Gynophore 3 mm;
ovary pubescence dense, appressed, yellowish-white, ca 14 mm long,
1.25 mm wide; style 16-17 mm, geniculate 6-7 mm from the distal end;
stigma flattened, diameter ca 1 mm, base short, ere. t pubescence.
Legume stipitate, slightly exerted beyond calyx, greenish-brown, flat,
glabrous, sutures thick, 21 cm long, 2.5-3.2 cm wid<-; stipe 14-17 mm;
beak to 4 mm; dehiscence causing valves to twist ore -half of a turn.
Seeds nearly orbicular, black, smooth, compressed, 1-2 mm thick, ca
10 mm long, 9 mm wide, 7 seeds per pod; hilum lineai , 3 mm x 1 mm.
Figures 21 and 22.
The Juriin Clitorjia is characterized as a tall Peruvian shrub with
elongate subpaniculate inflorescences of many medium-sized, blue and
pinkish-white flowers which possess the large coriaceous bracteoles of
a smaller size, and subvelutinous leaves, shortly acuminate, with larger
stipules and stipels.
PHENOLOGY: Flowers have been collected from late May to early
June and from mid January to February, with one collection from mid
October. Mature fruits were collected in early June.
TYPE COLLECTION: PERU. Shrub, 8-12 ft, with elongated branches;
standard and wings white, blue-tinged, keel blue, white proximally, La

Figure 21. Clitoria juinensis - I. (a-b) leaflets, x 1; (c) inflores¬
cence, x l;"(d) calyx, x 1; (e) vexillum, x 1, (f) ala and
carina, x 1; (g) androecium, x 1; (h) gynoecium, x 1;
(i) stigma and style apex, x 5. (Isern 2440, F 1575131:
a-b. Ferreyra 1 3555, ílIS: c-i.)

238

Figure 22. Cljtoria juinensis^ - II. (a) pair of flowers at a node,
x 1; (b) bracteoles and calyx with beginning rruit, x 1;
(c) immature fruit, x 1; (d) mature fruit, x 1; (e) three
views of seed, x 1; (f) leaf scar, x 1; (g) stipule scar,
x 1. (Ki 11ip & Smith 23380, NY: a,f. Ferre,ra 13555,
HIS: b,d. McLean s.n., K-445: c. Killip & Smith 25136,
NY: e,g.)


241
Merced, Dept. Junin, 700 in, 29 May-4 Jun 1 929, Kill ip & Smith 23380
(H0L0TYPE: NY. Isotypes: F 616214, US 1358082).
Specimens varied from early to late flowering tages. The Kill ip
& Smith 23380 collection was an intermediate stage in full bloom plus
the collectors had recorded some field observations, thus it is selected
as the type collection. The New York specimen has better flowering
material.
The Killjp & Smith 25136 (NY) specimen lacked an inflorescence and
flowers-, but it has a dehiscent fruit plus the only seed seen.
Therefore, this specimen is selected as a paratype. Another collection
(from Wisconsin) came to this author's attention nine months after the
description of the species had been written. Sheet one of Ferreyra
13555 (WIS) had better material of the inflorescenci , flowers, and a
fruit. Sheets two (WIS) had vegetative material. /1 though this
specimen was nut used in writing the original description, measure¬
ments of a dissected flower placed in the packet plus measurements of
other structures were used to supplement the description of the species.
With the exception of lacking seeds, this pair of sheets has the best
representative material for the species; therefore, it is selected as
a second paratype collection.
ECONOMIC IMPORTANCE: One collection, Ferreyra 13555, reported the
species as a cultivar.
NOTES: Most of the collections were identified as C. nervosa and
cited by Macbride (1943) under this name in his "Flora of Peru."
Clitoria nervosa (absent in Peru) can be easily disi inguished by its
smaller flowers, shorter pilose calyx, and narrower fruit. Cl itoria
juninensis lias also been misidentified as C. racemos i (a synonym of

242
C. fa irchi 1 diana) and C. arbórea. Clitoria fa irchiI diana (absent in
Peru) can be distinguished by its Atlantic coastal habit, elongate
glabrate leaves, and its larger bracteolate, violaceous flowers.
Clitoria arbórea is easily recognized by its elongated primary branches.
The closes affinities are with C^ andrei of Ecuador, which has
elongated acuminate leaflets, shorter stipules and tipel s, slightly
larger flowers with bigger bracteoles.
DISTRIBUTION (Figure 16); This species is a highland shrub,
apparently endemic to the Department of Junin in Peru, and found in
forests and thickets at elevations of 600-2000 m. P E R^U. LOCALITY
UNKNOWN: Payon s.n. (G); Vita, hb. Hooker, McLeeins s.n. (K):
Manchamayo, 16 Oct 1863, Bern 2440 (F) and 2441 (F). JUNIN: La
Merced, 700 in, 29 May-4 Jun 1929, KiJ 1 ija & Smith 23il3 (NY,US); forest
clearing along Rio Perene, near "Hacienda 3," Colon 11 Perene, 600 m,
K i 1 li p_ & Smi th 251 36 (PARATYPE: NY); Pampa Whaley, Colonia Perene,
prov. Tarma, 700-750 m, 14 Feb 1959, Ferreyra 13555 (PARATYPE: WIS-2
sh.); road between Tarma and San Ramon, at Oreja Capelo, 1500-2000 m,
13 Jan 1 945, Hodge 6253 (F); Chanchantayo Valley, lOuO m, Feb 1924-27,
Schunke 307 (F).
5. Clitoria moyobambensis Fantz, sp. nov.
Tree. Branches 4-7 mm diameter, pith hollow, juvenile branches
subquadrate, longitudinally striate and sulcate, grooves shallow to
deep, pubescence uncinate and short falcate to subappressed trichomes,
branches becoming subterete, glabrous with age; bar! medium gray,
splitting into longitudinal strips, whitish-tan beneath; stipule scars

243
compressed, "s-shaped," extending a third to half w.. / around stem.
Leaves 3-fol iate, coriaceous, leaflets ell iptic-lani ■..•oíate, tapering
to acuminate apex, acumen 1.5-3 cm long, to 1.2 cm wide at base, base
broad cuneate, midrib impressed above, uncinate-pubescence, primary
nerves of 13-16 pairs, upper surface dark green, glabrate with
pubescence sometimes of uncinate or short erect trie homes on major
nerves, lower surface light green, glabrous, 12-15.5 cm long, 6-8 cm
wide. Petioles conspicuously twisted at base, longitudinally striate
and can'iculate, pubescence scattered, minute, falcate to appressed,
4-5 cm; rachis similar, 4 cm. Petiolules quadrangular, dark-colored,
rugose, pubescence similar to petiole, 8-9 mm. Stipules large,
conspicuous, concave nearly half way around stem, l.nceolate-ovate,
rapidly tapering above middle to acute apex, 10-12 mm long, 4-6 mm wide;
stipels caducous (?), one terminal stipel seen, 2 im x 0.4 mm.
Inflorescence axillary and terminal, subpaniculate-racemose, solitary,
multiflowered, axes pubescence subappressed, moderately appressed,
central axis twisting, longitudinally striate and coiiiculate, to 4 cm
(juvenile ?), primary branches which bear pedicels subsessile, 1 (2) mm.
Pedicels paired, 8-14 mm. Bracts large, conspicuous, lanceolate, short
acuminate, middle bracts paired, pubescence minute, appressed, 8-11 mm
long, 3-5 mm wide; inner and outer bracts not observed, caducous (?).
Bracteoles large, coriaceous, multi-striated, pubescence appressed,
ciliolate, oblong-elliptic, apex obtuse or abruptly apiculate, subequal
calyx, 12-15 mm long, 9-10 mm wide. Flowers medium size, 4-5 cm, color
unknown (dried state similar to that of C. flexuosa and pozuzoensis
which are lilac-rose to white). Calyx glabrate, very sparse appressed
pubescence, multi nerved and subimpressed, inconspicuous, tube 13-15 mm

244
long, 4-6 nun wide at base to 8-10 mm wide at throat lobes deltoid,
acute, 5-6 ram long, 4 mm wide, ventral lobe 7 mm long, 2 rnm wide.
Vexillum pubescence moderately dense, appressed, tawny, claw ca 11 mm.
Carina and alae hidden and/or torn, thus measurements unknown. Stamina 1
tube glabrous, 27-28 mm, free filaments 1-3 mm; anthers 0.75 mm long,
0.5 mm wide. Gynophore ca 5 mm; ovary pubescence dense, of uncinate
plus 1 mm ascending trichomes along dorsal and ventral edges, ca 13 mm
long, 1 mm wide; style ca 14 mm long, geniculate 6 .n.n; stigma dark-
colored, subflattened, ca 0.7 mm diameter, base with ring of dense,
short reflexed trichomes. Legume unknown. Figure 33.
The Moyobamba Cl itona is characterized as a ti ee with conspicuously
large bracts and stipules, medium-sized flowers with an uncinate
pubescent ovary, and large bracteoles that subequal the calyx.
PHENOLOGY: Flowering dates are unknown.
TYPE COLLECTION: PERU. prov. Moyobamba, 1838, Mathews s.n.
(H0L0TYPE: K-439, hb. Hookerianum of 1867).
This specimen was identified as C. hoffmannseg m i (a synonym of
C. arbórea) which is easily recognized as differing by the elongated
primary branches, larger bracteoles, and smaller bracts and narrower
stipules. This specimen is unique in the size of its bracts and
stipules and the pubescence of the ovary. Its clos. st affinities appear
to be with C^ andrei and C. juninensis based upon the size of several
floral structures and the similar stigma.
DISTRIbUTION (Figure 19): This species is known only from the type
collection of prov. Moyobamba, Peru.

Figure 23. Cl itoria moyobambensis. (a) leaflet, x 1; (I.) portion of
stem with stipules and stipular scar, x 1; (o) inflores¬
cence, x 1; (d) flower, x 1; (e) calyx, x 1; (f) gynoecium,
x 1; (g) stigma, x 1. (Mathews s.n., K-439: a-g.)

246

247
6. Cl i tori a amazonuin Mart, ex Benth., Ann. We in. Mus. Natur.
2: 115. 1837.
Clitorja acuminata Benth., Ann. Wein. Mus. Natur. 2:
1837, non Graham (1828); nom. i 1 leg.
Clitoria muc£onulata Benth., Ann. Nat. Hist. 3: 345. 1839.
Ternatea amazonum (Mart, ex Bentli.) Kuntze, Riv. Gen. PI.
210. 1891.
BracteaHa amazónica Mart., nom. in schedula.
Clitorja ajiiazonica Mart. a. Benth., noun in schedula.
Erect shrub to small tree, 1-4 m tall, occasionally to 8 m, apex
rarely reported as climbing. Branches elongate, la, apically compressed
with longitudinal caniculi, becoming subquadrangulai then terete,
juvenile branches with pubescence short, spreading, becoming glabrous,
pith solid to sometimes minutely hollow, diameter 28 mm; axillary buds
3-5 mm; stipule scars elliptic, bundle scars iriconspicuous, stipule
scars falcate; bark brownish-black, peeling in longitudinal strips,
light-colored beneath. Leaves 3-foliate, coriaceou , leaflets variable,
generally ovate to ovate-elliptic to ovate-1anceolale with lateral
leaflets sometimes round, nearly orbicular or slightly longer than broad,
apex acuminate, acumen 0.5-2.5 cm, mucronate, base broad cuneate to
rotund, rarely weakly cordate, midrib impressed abo.e with numerous
short spreading hairs, primary nerves (6-9 orbiculai leaflets) 8-13
pair, upper surface glabrous, lower surface with mouerately dense, short
pilose hairs from nerves, 4.5-15 cm long, 3-8.5 cm wide. Petiole
slightly compressed, longitudinally grooved, pubescence of few, short
appressed hairs, 2-8 cm; rachis 1.5-3.5 cm. Petiolules dark-colored,

248
subquadrangular-terete, bearing short, erect, white hairs, especially
adaxially, 5-8 mm. Stipules ovate-deltoid, acute, deciduous, pubescence
sparse, short-ci1iate near apex, 2-5 mm long, 2-3 min wide; stipels
linear to awl-shaped, pubescence sparse, 2-5 mm Ion.,, less than 0.5 mm
wide, terminal stipel always shorter. Inflorescence axillary or
terminal, paniculate, solitary, multiflowered, axes with dense reddish-
brown, appressed pubescence, central axis rarely bearing small leaves
near base, deciduous, longitudinally grooved and striated, 2-8 cm long,
primary' branches that bear pedicels conspicuous, 4-16 mm long, swollen
at apex. Pedicels paired, 6-12 mm. Bracts variable; bracts subtending
primary branch paired, ovate, acute, deciduous, 2-3 ¡am long, 1-2 mm
wide; bracts subtending pedicel, middle pair persistent, ovate, acute,
pubescence dense, appressed, concave, reflexed in aâ– ,, 4-8 mm long,
3-6 mm wide, inner pair caducous, outer bract narrower, deciduous.
Bracteoles large, coriaceous, elliptical, obtuse to short apiculate,
minutely striate, pubescence concentrated near base, nearly glabrous
towards apex, scattered, minute, appressed, subequaling calyx, 20-28 mm
long, 9-16 mm wide. FIowers large sized, 6-8 cm, pale violet or rose
to white, medial lines dark violet. Calyx glabrate. with sparse
appressed hairs, tube 16-25 mm long, 5-7 mm wide at base to 9-13 mm
wide at throat, lobes deltoid, 5-7 mm long, 4-5 mm v.ide, acute, ventral
lobe slightly pubescent, 6-8 mm long, 2-3 mm wide, /exilium pubescence
glabrate, of minute appressed hairs concentrated near complicate fold,
4-6 cm wide, claw 12-16 mm. Alae oblong and falcate near middle
extending beyond arina 8-11 mm, blade 28-34 mm long, 4-8 mm wide, claw
19-28 (34) mm long. Carina falcate, 13-17 mm across, 5-6 mm wide, claw
28-44 mm. Staminal tube glabrous, nearly straight to subfalcate,

249
35-43 (48) min long, free filaments 3-5 mm; anthers -3 mm long, 0.5-
0.8 mm wide. Gynophore 6-7 mm; ovary pubescence de;,3e, appressed,
white, less dense in center than at margins, elongated, 22-28 mm long,
1.5 mm wide; style 14-27 mm, geniculate 6-9 mm from distal end; stigma
capitate. Legume nearly straight, stipitate, shortly exerted beyond
calyx, flat, coriaceous, glabrous except for slightly thickened
sutures which bear short, spreading to erect hairs, 10-27 cm long, 1.3-
1.9 (2.1) cm wide; stipe elongate, 24-51 mm long, 3 mm wide expanding
to 6 mm at apex; beak 1-7 mm; dehiscence 1-1.25 turns. Seeds black,
nearly orbicular, compressed, 1-2 mm thick, smooth, 3-10 mm long,
9-10 mm wide, 10-17 seeds per pod; hilum minute, linear, 2 mm x 0.5 mm.
Seed germination hypogean. Figure 24.
Martius1 Amazon Clitoria is characterized as a tall shrub to small
tree with large bracteoles subequaling the calyx, with conspicuous
elongated primary branches in the inflorescence, and with large flowers.
PHENOLOGY: Flowers have been collected in two periods, May to
July and September to December. Fruits have been collected in June and
July, and occasionally in November.
TYPE COLLECTION: BRAZIL. Para, ad. fluv. Ama, on urn prope Ponte
de Mattary, Sept., Martius 2740 (LECTOTYPE: M 1241a! Isolectotypes:
K-hb. Hooker, not seen; M 12408!, 12409!, 12411!, 12412!, 12413!;
A! and S!, each a photo of K).
Bentham originally described C. amazonum as having suborbicular
lateral leaflets. He cited two syntype collections as: (1) "In sylvis
ad fluv. Amazonum, prope Conta de Mattury, provinciue Para, Martius"
and (2) "Ega Amazonum, Poppig." Only the Martius specimens have
suborbicular lateral leaflets. Six sheets of tfie Martius collection at

Figure 24. Clitoria amazonum. f. amazonum: (a-b) lateral leaflets,
x 1; (c) inflorescence, x 1; (d) flowers, x 1; (e) compli¬
cate ve/illum,x 1; (f) ala and carina, x 1; (g) calyx,
x 1; (h) androecium, x 1; (i) gynoecium, x 1; (j) fruits,
x 1; (n) leaf scar, x 1.8; (o) stipule scar, x 1.8.
f. vulgaris: (k) seed within fruit, x 1; (1) three views
of seed, x 1; (m) lateral leaflet, x 0.5. (Ducke 8650,
RB 11843: a. Prance et al. 14138, M -7: b-i. Prance
et a 1. 6597, S: j,n-o. Prance et a 1^ 6194, S: k-1.
Prance et aj_. 8178, NY: m.)

251

252
München are numbered consecutively from 12408 to 12.13, each stamped
"vid. Bentham." Only the first sheet (M 12408) bears both the names
Bractearia amazónica Mart, and Clitoria amazónica, cuereas the remaining
five sheets bear only the last name. The Marti us collection matches the
original description plus it bears the unpublished basionym of Martius
upon which Bentham selected the specific epithet and the original names
for the section. Therefore, the Martius collection was selected over
the Poppig collection as the lectotype for the species.
Of the eight sheets of the Martius collection, the Muchen specimens
were known to have been seen by Bentham and they were the probable
source for the species epithet. With their consecutive numbers and
enclosure in one folder, the six sheets are treated as equivalent. Two
specimens are mostly vegetative (M 12408 R 12411), the other four
specimens nearly the same. Since the International Code of. Botanical
Nomenclature recommends the selection of a single sheet for the type,
the specimen numbered "M_J2410" was selected as the lectotype because
it has inflorescences, flowers, and a dissected flower inside the
mounted packet.
Bentham neglected to cite the collection number of the Martius
specimens, although the number "2740" was present. This was typical of
Bentham in his citations for other species in his publication. In
addition, he altered the spelling of "Ponte de Mattary" (label data)
to "Conta de Mattury" in his publication. This raises the question of
another change in spelling, the alteration of the ending of the specific
epithet. The type specimens bear the name "amazónica" whereas Bentham
(1837 ) published the original name as amazoriunn. The ending "-urn"
usually indicates the masculine or neuter gender, not feminine. In

253
1858 and 1862, Bentham again published the species as C. amazonum. In
comparing triese three treatments, it is apparent that Benthain rewrote
the descriptions rather than recopy his data from the earlier publica¬
tion. Since Bentham twice failed to change the naim from C. amazonum
to C. amazónica, it seems that he intended the name to be C. amazonum.
Based upon two Latin works, Stearn (1966) and Cassel (1959), "Amazonum"
is the feminine genitive plural noun of the term "Amazon." It is
believed that Bentham meant the basionym to be a possessive noun in the
plural, to include all the clitorias of the Amazon, which would include
the Rio Amazonas and its major tributaries, such as the Rio Solimóes
and Rio Negro, as he cited in his distribution in 1.;58. Therefore, in
this particular case, the term "Amazonum" is feminine and the "-urn" is
a proper ending which is in agreement with the generic name Clitoria.
The name C. acuminata Benth. was a homonym of i.raham's name
published in 1828, therefore an invalidly published name. Bentham
(1858) reduced the name as a synonym of C. amazonum. Bentham cited one
collection as: "Ad flum. Amazonum ripas, locis sylvaticis provinciae
Rio Negro, Marti us." Two Muchen sheets of this collection have been
examined. They match the type except in lateral leaflets, which agree
with Poeppig specimens. One sheet (M 12414) has vegetative material
whereas the other sheet (M 12415) has mostly flowering material. Since
Bentham examined these specimens and Muchen is the main depository for
the Marti us collection, these two sheets probably represent the holotype
for the name C. acuminata.
The name C. mucronulata disappeared from botanical literature after
its original publication in 1839. Bentham did not include the name in
synonymy in either his 1858 or 1862 treatments. Bemham cited one

254
collection as: "Ad Rio Madeira et Borba in Brasilia" with a prior
notation that the type was in the Petersburgh llerbai iurn (=LE). The
type collection was not seen, but an Arnold Arboretum Herbarium specimen
was examined which was collected by Riedel. It agreed in label data
and was stamped "Herb. Acad. Petrop." This specimen was identified as
C. mucronulata and is a probable isotype. It agreed with the type
except for the slightly smaller flowers and smaller leaves. Although
the flower size is 5.5-6 cm, the androecium and gyriuecium agree with
the size of the type, larger than any other species in this section.
The leaves are narrow (2-4 cm), but are crowded towards the tip of the
branches along with the inflorescences. It is believed that this
specimen plus two other's identified as C_. mucronulai a (Riedel 1 336 and
Spruce 1870) are C. amazonum collected in an immatuie state. Thus the
name C. mucronulata is placed as a synonym of C. amazonum.
NOTES: This widespread species shows a great ,.mount of vegetation
in a number of structures. However, there appear to be two distinct
forms based upon the minor variation in lateral leaflet shape.
Intermediates have not been seen. Those specimens labeled C. mucronulata
are distinct in size from these two forms, but are believed to be
immatured leaves of the form more commonly collected
The name C. amazonum frequently is placed on specimens belonging
to most other species in the section Bractearia. Clitoria amazonum
can easily be distinguished from the other species by its large flowers
and calyx, and its much narrower fruit. It appears to be an evolutionary
shoot off the evolutionary line of C. arbórea. Cl i I oria amazonum and
C. arbórea are the only two species which have the woody, paniculate
inflorescences with conspicuous, elongated, primary branches. However,

255
in many characteristics, C. amazonum is more advanc i than the previous
species listed.
DISTRIBUTION (Figure 19): This species is found scattered in the
riverine forests of the Amazon Basin in Brazil, gem rally on varzea
land subjected to periodic flooding, infrequently 01. terra firma.
KEY TO FORMS:
1. Lateral leaflets suborbicular to short ovate, luigth/width ratio
is less than 1.8 (2):1, apex broad, generally obtuse with short
acumen, base rotund to weakly cordate ... 6a. f. amazonum
1. Lateral leaflets elongated, ovate-elliptic to ovate-lanceolate,
length/width ratio is greater than 2:1, apex gradually narrowed
to acuminate tip, base broad cuneate to weakly iotund
6L. f. vulgaris
6a. Clitori a amazonum f. amazonum
Clitoria amazonum Mart, ex Benth. f. rotundifolia Rizz.,
Arq. Jard. Bot. Rio de Jan. V7: 189. 1963.
Lateral leaflets short ovate to suborbicular, generally obtuse
with a short acumen to 5 min, base rotund to weakly cordate, length/
width ratio is 1.1-1.8 (2): 1 .
NOTES: Rizzini described f. rotundifolia as "yariat foliolis
latiore ambitu, fere orbicularibus, 6-9 cm x 5-7 cm, basi leviter
cordatis, ápice acumine 1-4 mm longo praeditis1' bas¡ d upon one
collection (Holotype: Ducke 8_650, RB 1 1843!). Since the type of the
species belongs to the typical form, Rizzini's name could not be
adopted.

256
DISTRIBUTION: B R A Z I L. AMAZONAS: Environs de Manoas, 1906,
Labroy s.n. (P); Rio Purus, Lago Préto, 2 km north uf Lábrea, 25 Jun
1971, Prance et al_. [3690 (NY,S,U); Rio Ituxi, vie. Boca do Curuquete,
11 Jul 1971, Prance et al. 14138 (M,MICH,NY,U). Fortaleza, Rio Jurua,
Nov 1901, Ule 5907 (G-2 sli.). PARA: Martius s.n^ (W); Parana do
Adanaca, Faro, 7 Sep 1 907, Ducke 8650 (RB). RONDÓN IA: Rio Riberao
near confluence with Rio Madeira, 27 Jul 1968, Prance et al. 6579
(M,MICH,NY,S,US).
6b. Clitoria amazonum Mart, ex Benth. f. vulgaris Fantz, f. nov.
Clitori a acuminata Benth., Ann. Wein. Mus Natur. 2: 115.
1837, non Graham (1828); noun i 11 eg.
Clitoria mucronulata Benth., Ann. Nat. Hi t. 3: 345. 1839.
Lateral leaflets similar to terminal leaflets, smaller,
ovate-elliptic to ovate-lanceolate, apex gradually narrowed to acumen
(0.5) 1-2.5 cm, base broad cuneate to rotund, 1 eng11./width ratio
greater than 2:1.
TYPE COLLECTION: BRAZIL. Ad littora lacus Egmsis, Septbr. 1831,
Ega Amazon, Poeppig 2500 (H0L0TYPE: W. Isotypes: G-7, HAL 37532, NY).
This collection was selected as the type for the more common form
because it is a syntype for the species C. amazonuni, but was excluded
from the typical form because of its elongated lateral leaflets. The
Wein specimen has better material, including more matured leaflets, plus
flowers and juvenile fruit. Complete data was found on only the New
York specimen, a duplicate of Herbarium Vindobensis (W).

257
n
DISTRIBUTION: BRAZIL. LOCALITY UNKNOWN: Martius s.n. (G).
AMAZONAS: 1 924, Mel in 204 (S); Rio Acari, Mar 1 945, Cooper III s.n.
(NY); Muratuba Feuchte campas, Aug 1927, Luetzelburg 21205 (M-2 sh.);
Ad f 1 uni. Amazonum ripas, Rio Negro, Martius s.n. (M-2 sh., type of
synonym C. acuminata); Rio Negro Basin, between 11 ha Jacaré and Airao,
11 Oct 1971, Prance et al. 15057 (M,NY,S,U); Rio Negro, inter Barra et
Barcellos, Nov. 1851, Spruce 1870 (BM,CGE,G,NY,W-2 sh.); Fazenda Santo
Antonio, Parana do Xiborema, em frente de Manaus, May 1953, Fróes
29628 (NY); Manaos, Jun-Jul 1888, von Bayern s.n. (leafless, M); l.c.,
30 Apr 1882, Schwacke 3871 (RB); Rio Japura, Jubara, 15 Sep 1904,
Ducke 6770 (BM,G); Rio Jurua Basin, Teffe, 11 Jun 1906, Ducke 7331 (RB);
Paranagua, mun. Teffe, 22 May 1933, Krukoff 4529 (G-2 sh.,NY,S,U,US);
Rio Solimoes, Iauara, 18 Sep 1971, Prance & Prance 14751 (NY,U); Rio
Ituxi-Purus, Igarape Caietitu, 2 km S of Labrea, 1 Nov 1968, Prance,
Ramos & Farias 8J_7<3 (NY); Rio Ituxi, vie Boca do Curuquete, 8 Jul 1971,
Prance et al. 13998 (NY,U); Rio Madeira, Borba, 23 Apr 1937, Ducke 479
(NY,US); Rio Madeira, ad Casaraesan, May 1828, Riedel 1336 (NY); Ad
Rio Madeira et Borba, Riedel (A-isotype of synonym C. mucronulata).
R0ND0NIA: Basin Rio Madeira, 4 km above Jaciparana on Rio Jaciparana,
28 Jun 1968, Prance 5288 (NY,S,U); Basin Rio Madeira, Rio Pacaas Novas,
7 Aug 1968, Prance et al. 6854 (NY,S,US); Rio Guapore, Vila France,
10 Jun 1952, Black & Cordeiro 52-14813 (leafless, NY); N bank of Rio
Abuna, between Cachoeira Tres S and Fortaleza, 4-16 km above mouth,
18 Jul 1968, Prance et al. 6194 (GH,NY,S,U,US). PARA: Rio Tapajos, Boa
Vista, May-Jun 1929, Dahlgren & Sella 68 (F,S); l.c., 1931, Costa 32 (F)
Rio Tapajos, Cachoeira Maranhaozinho, 8 Dec 1915, Ducke 15870 (RB-2

258
sh.); Rio Xingu, 4 Jul 1909, Snethlage 10421 (G-2 sii.); Belem, hort.
bot., Jun 1908, Huber 9352 (BM-2 sh. , G-2 sh. ,RB ,U ,!J).
7. Clitori a nervosa Herzog, Fedde Repert. Sp. Nov. Reg. Veg.
7: 56. 1909.
Tree, 6-18 m tall, to 20 cm diameter; bark grayish-brown (black
teste Herzog) narrowly longitudinally striated, breaking off in square
to rectangular pieces, brown beneath. Branches to 7 mm diameter, pith
hollow, leaves tufted near apex, juvenile branches lightly compressed,
angular, pubescence dense, yellowish-brown, spreading, becoming terete,
short puberulent, white pubescence; bark dark brown, splitting
longitudinally; axillary buds 2 mm, scales ovate, concave, pubescence
minute, 0.1-0.2 mm; leaf scars suborbicular, bundle scars inconspicuous;
stipule scar nearly straight to weakly falcate. Leaves 3-foliate,
coriaceous, leaflets elliptic to ovate-elliptic, unequal, terminal
leaflet largest, apex obtuse to abruptly short acuminate, acumen to
0.5 cm, sometimes splitting alongside midrib appearing emarginate, base
rotund, midrib impressed above, primary nerves of 14-21 pair, upper
surface dark green, glabrous, lower surface subvelutinous, densely
pubescent, trichomes more dense along nerves, semiappressed, 5.5-15 cm
long (25-28 cm teste Herzog), 3-8 cm wide. Petiole compressed-terete,
base dark-colored, swollen 7-9 mm long, pilose, 3.5-3 cm; rachis
similar to peiole, 1.5-3 cm. Petiolules subquadrangular to terete,
reddish-brown densely velutinous, 7-10 mm. Stipule; deciduous, narrow,
deltoid-lanceolate, acute, sometimes weakly arcuate, sparsely pubescent,
sometimes ciliolate, striations inconspicuous, 6-7 mm long, 2 min wide;

259
stipels linear, gradually tapering to bluntly acute apex, 6-8 mm long,
1 mm wide at base. Inflorescence subpaniculate, axillary, 1-3 per
axil with usually 1 long axis and 1-2 short axes, multi flowered, all
axes densely pubescence, velutinous, reddish-brown, central axis 4-11
cm, usually from denuded nodes, primary branches which bear pedicels
subsessile to 4 mm. Peduncle 10-13 mm, internodes axis 2-10 mm,
shorter towards apex; pedicels 4-6 mm, reddish-brown velutinous. Bracts
pubescence reddish-brown, dense, velutinous; bract base primary branch
solitary, deciduous, ovate-lanceolate, 5-7 mm; middle bracts at base
pedicel paired, concave, ovate, acute, semipersistent, somewhat
reflexed in age, 5-7 mm long, 2 mm wide; inner and outer bracts
caducous. Bracteoles large, coriaceous, narrow oblong-elliptic,
subequaling calyx, obtuse, pubescence dense, short- elutinous, 11-14 mm
long, 6-7 mm wide, inserted 1 mm below calyx. Flowers small size,
3-4 cm, white-lilac with rose-violet lines on inner vexillum. Calyx
pubescence dense, short pilose, tube 9-12 mm long oiten splitting
between dorsal and one of the lateral teeth, 3-4 mm wide at base to
4-6 mm wide at throat, lobes deltoid, acute, 3-4 mm long (appear to be
7-8 mm because of tube slit), ventral lobe 5 mm. Vexillum pubescence
uncinate, covered by tawny, moderately dense, appre sed trichomes,
margins somewhat ciliolate. Alae with uncinate pubescent, extended
beyond keel 4-5 mm, blade 17-22 mm long, 4-7 mm wide, claw 10-16 mm.
Carina weakly falcate, uncinate-pubescent, 7-8 mm across, 4-5 mm wide,
claw 18-20 mm. Staminal tube uncinate-pubescent .t apex, 17-22 mm,
free filaments uncinate - pubescent, 2-3 mm; anthers 2 mm long, 0.7-1.0 mm
wide. Gynophore 4 mm; ovary pubescence dense, white, appressed,
dark-colored beneath and more conspicuous towards margins, 11 mm long,

260
1.5 mm wide; style 14-15 mm long, geniculate 4-5 nun from distal end,
beard of short trichomes, 0.6-0.9 mm long, stigma capitate, 0.5-0.8 mm
diameter. Legume nearly linear, flat, coriaceous, Jabrate with short
hairs along both sutures, green, sutures thickened, stipitate, slightly
exerted beyond calyx, 19-25 cm long, 2.1-2.3 cm wide, ventral margin
undulate; stipe 12-21 mm, 3-4 mm thick to 6-9 mm wide at apex, beak
7-15 mm; dehiscence not observed. Seeds not seen, ca 6-8 seeds per pod.
Figure 25.
Herzog's Clitoria is characterized as a tree with subvelutinous
leaves, small flowers, and with a pilose calyx hidden by the conspicuous,
large, coriaceous bracteoles.
PHENOLOGY: Flowering specimens have been col lee ted from September
to December with the only matured fruits collected in mid-October.
TYPE COLLECTION: BOLIVIA. In den Savannenwalclchen der Hugel von
Buenavista, 400 in, Oct 1907, Herzog 552 (Type believed to be at Berlin
and destroyed; location of isotypes unknown). NEOTiPE: BOLIVIA.
"Chancho," árbol hasta 18 m, sépalos verde-amarillos, petalos blancos
lilas con venas rosa-violaceas, sobre la parte céntrica del pistalo
mayor androceo, filamentos y pistilos blancos, estambres amarillos.
Madera de un solo color amarillento, bastante resistente. Cascara sin
grietas y casi sin fibra, Bosque del Rio Quimori, prov. Sara, Depto.
Santa Cruz, 400 m, 20 Dec 1924, Steinbach 6781 (NEOTYPE: M0 929318.
Isoneotypes: A,BM,F 563846, G-275,K-446,S,UC 285029,W 25642).
A neotype collection was selected because the type at the Berlin
Herbarium was destroyed by fire during World War II and the location of
any isotypes, if they exist, is unknown. The Stejinhac^p 6781 collection
most nearly fits Herzog's original description except for its smaller

Figure 25. Clitoria nervosa. (a-b) leaflets, x 1; (c)
x 1; (d) flower, x 1; (e) calyx, x 1; (f) al
x 1; (g) gynoecium, x 1; (h) androecium, x 1
x 1; (j) stipular scar, x 1.5; (k) leaf scar
vexillum, x 1. (Steinbach 6781, UC 285029:
G-275: b-c; MO 929318: j-k. Prance et al.
i.)
i nflorescence,
u and carina,
; (i) fruit,
, x 1.5; (1)
a ,d-f,1;
7961, NY:

262

263
leaves. The Missouri botanical Garden specimen was hosen as the
neotype because it had both attached leaves and inflorescences with
flowers and juvenile fruits. Many of the other spec ¡mens have much of
the material in packets. This collection also has small sample of
the wood.
Herzog does not indicate the calyx pubescence, hut he does mention
the tube size of 1 cm, which agrees with the specimens examined, and a
lobe length of 7 mm, which is longer than the typical length. Although
in apparent disagreement, the calyx tube often spliis between the
dorsal and lateral teeth, thus the tube apex is not always distinct,
and a measurement of tfie lateral teeth will produce lengths of 7-8 mm.
Other characters are in agreement, including the neives, which are
more tightly spaced in the smaller leaflets than in other closely
related species. All bolivian specimens examined with the large
bracteoles that subequal the calyx, hiding it, belonged to this species.
Even with the absence of a type for comparison, the Steinbach 6781 and
other collections cited are C. nervosa.
VERNACULAR NAMES: BOLIVIA: Chancho, Steinbach 6781; Poroto
(árbol), Steinbach 6399.
NOTES: Macbride (1943) reported C. nervosa from Peru. All the
specimens that he cited belong to the species C. jui.inensis, which like
other species closely related, can be distinguished from C. nervosa by
their larger flowers and calyx pubescence of scattered appressed hairs.
DISTRIBUTION (Figure 19); The few collections of this species were
obtained from forests in Santa Cruz, Bolivia, and from southeastern Acre,
Brazil, at elevations reported as 400 m. BOL I__V I A. SANTA CRUZ:

264
Bañado del Rio Guenda, prov. Sara, 400 m, 8 Sep 192:, Steinbach 6399
(A,G). B R A Z I L. ACRE: 1-3 km road Sena Nadun. ira to Rio Branco,
10 Oct 1 968, Prance et ah 7961 (NY,US).

265
Section Flexuosa
IB. Clitoria L. subgenus Bractearia (Mart, ex Benth.) Fantz
section Flexuosa Fantz, sect, nov.
Lianas, rarely erect shrubs or trees. Inflorescence weak to
strongly zigzag towards apex, internode segments abruptly bent
alternatively in opposite directions, subpaniculate, primary branches
which bear pedicels inconspicuous, subsessile to 6 mm. Calyx
pubescence dense, short, appressed, somewhat silky in appearance.
Bracteoles coriaceous, broad, (6) 9-15 mm wide, subequaling the calyx
or slightly shorter, or infrequently much shorter than calyx, 1-4 cm
long. Flowers large, 6-8 (9.5) cm, white to lilac-rose. Staminal
tube elongate, 3.5-5 cm; anthers large, 2-3 mm. Legume long-
stipitate (2.5-4 cm), flat, coriaceous, pubescence appressed, typically
1.5-2.5 cm wide. Seeds brown, oblong, length greater than width,
thickened.
The members of the section FIexuosa are characterized by their
subpaniculate, zigzag inflorescences, large flowers, and somewhat silky
calyx hidden usually by large, conspicuous bracteoles. Most members
exhibit a climbing habit.
HOLOTYPIC SPECIES: C. flexuosa Fantz
NOTES: This section is the more advanced section of the two whose
members have primitive, large bracteoles which subequal the calyx and
obscure it from view. In this section there is a trend towards
shortening of the bracteoles, slightly so in a variety of C. flexuosa
and most notably in C. pozuzoensis. However, the bracteoles remain
broad (6 mm or more wide) even with a decrease in length, and are wider
than the bracteoles of sections Brachycalyx and Cauliflorae. A trend

266
also exists towards the climbing habit, with must members being woody
vines, and toward large flowers. Large flowers are typically 6-8 cm,
although in this section, flowers of pozuzoensis sometimes attain a
size of 9.5 cm.
This section is unique in the production of zigzag inflorescences.
The flexuous rachis may be weakly developed in sone individuals, but it
can usually be observed near the apex of the inflorescence. Strongly
flexuous inflorescences suggest sympodial growth in which the lateral
branches exhibited dominant growth over the main axis. This type of
growth might better be called pseudosympodial since the Clitoria
paniculate inflorescences have only one lateral branch and the pedicel
origin would shift from the apex of lateral branches to the main axes.
Section FIexuosa probably has its evolutionary affinities with the
ancestors of C. amazonum which also exhibits the larger cálices,
flower-size, and androecium, plus a tendency towards climbing. Although
the seeds exhibit a different shape, the legume has decreased in width
as in amazonum.
DISTRIBUTION (Figure 9): The members of section Flexuosa are
reported from Peru and Ecuador occurring in forests at elevations of
250-600 m. Infrequently collected, most members are known from one to
a few collections.
KEY TO THE SPECIES:
1. Trees; inflorescence short, 1-2.5 cm; bracts medium-sized, 7-10 mm
long, 4-7 mm wide; petiolules elongate, 8-11 mm; leaves
obovate-el1iptic or elliptic, length/width ratio less than 2:1;
petiole of larger leaves 10-13 cm ....... .8. C. woytkowskii

267
1. Lianas, rarely erect shrubs; inflorescence elongate, 2-10 (20)
cm; bracts small, 3-6 mm long x 2-4 mm, or large, 8-15 mm long x
6-9 mm; petiolules shorter, 5-9 mm; leaves elongate, elliptic-
lanceolate, oblong-lanceolate, to ovate-lanceolate, length/width
ratio greater than 2:1; petioles of larger leaves 6-10 cm.
2. Bracteoles large, subequal to calyx to slightly shorter (half
calyx length), 18-41 mm long, (8) 10-17 mm wide; bracts large,
8-15 mm long, 6-9 mm wide; alae claw short, 23-26 mm; carina
claw short, 33-38 mm; legume long-stipitate, stipe 38-40 mm;
staminal tube short, 39-45 mm; leaves subpilose below
9. f1exuosa
2. Bracteoles small, much shorter than calyx (one-quarter to
one-third of the calyx length) (6) 8-13 (15) mm long, 6-10 mm
wide; bracts small, 3-6 mm long, 2-4 mm wide; alae long clawed,
26-33 mm, carina long-clawed, 39-42 mm; legume short-stipitate,
stipe 25-34 mm; staminal tube long, 43-51 mm; leaves
inconspicuously pubescent below, short, appressed
10. pozuzoensis
8. Clitoria woytkowskii Fantz, sp. nov.
Tree, 12 m tall. Branches 6-7 mm thick, pith hollow, pubescence
strigulose, trichomes 0.1-0.4 mm; bark dark brown, splitting in
longitudinal slits, light tan beneath; axillary buds 3 mm, scales ovate,
concave, glabrate, striate with impressed nerves, leaf scar transversely
oblong, rounded on lateral margins, bundle scars conspicuous, broad
horseshoe-shaped, absent below, stipule scar horseshoe-shaped on nodose
nodes, 4 mm long. Leaves 3-foliate, coriai.eou1 , concolorless,

268
leaflets broad, elliptic to obovate-ellipt ic, apex broadly obtuse,
abruptly acuminate, acumen 1-2 (2.5) cm long, 2-3 mm wide with expanded
base of 6-10 mm wide, acumen apex mucronate, base broad cuneate, midrib
weakly raised above with shallow groove on each side, primary nerves of
8-11 pair, upper surface glabrous, lower surface conspicuously strigose
along nerves, (6) 10-18 cm long, 5-10 cm wide. Petiole quadrangular-
terete, strigose, with one to three shallow longitudinal grooves,
swollen base 7-11 mm, (5) 9-13 cm long, rachis (1.5) 2.5-3 cm.
Petiolules slightly darkened, sparsely pubescent, 8-11 mm. Stipules
deciduous, deltoid, broadly acute, sparsely pubescent, 4-5 mm long,
1-2 mm wide, stipels narrowly deltoid, tapering from base to apex,
acute, 3-4 mm long, 0.6-0.9 mm wide, terminal stipels not observed,
caducous (?). Inflorescence subpaniculate, axillary, crowded
terminally, several-flowered, pubescence of juvenile axes silky,
densely strigose, becoming sparsely strigose to glabrate, penducle
3-5 mm, central axis 1-2.5 cm long, 2-3 mm thick, nodose, primary
branches that bear pedicels subsessile, 1-3 mm; rachis internodes
1-2 mm, weakly zigzag. Pedicels not observed. Bracts conspicuously
broad, reflexed in age, multi-nerved, moderately densely strigulose,
somewhat silky, ciliolate, ovate, acute; middle pair 7-10 mm long,
4-7 mm wide, persistent; outer bract 4-6 mm long, 4-5 mm wide, semi-
persistent; inner bract narrowly triangular, acute, deciduous, 7-8 mm
long, to 1 mm wide. Bracteoles very large, coriaceous, broad and
elongate-oblong, apex rapidly tapering, acute, pubescence somewhat
silky, densely strigulose, 41 mm long, 13 mm wide. Flowers white,
flushed with red veins (teste Hoytkowski), banner darkened, grayish in
dired state, mature flower not seen (Immature vexillum 5.5 cm,

269
protruding slightly beyond calyx. Comparison with other species of
equivalent state indicates mature flower probably 7-3 cm). Calyx
pubescence somewhat silky, densely strigulose, multi-nerved, tube 26 mm
long, 9 mm wide at base, becoming slightly narrowed towards middle
(7 mm) and expanding to 10 mm at throat, lobes ovate, short acuminate,
6 mm long, 4 mm wide, ventral lobe 8 mm x 1.5 mm, long-deltoid.
Vexillum pubescence densely strigulose, silky, claw broad, cuneate,
7 mm. Alae extended byond carina ca 5 mm, blade 22 mm long, 6-11 mm
wide, subfalcate, broadly expanded beyond carina, uncinate pubescent,
claw 14 mm. Carina flacate, acute, uncinate pubescent, blade 16 mm
long, 5-6 mm wide, claw 22 mm. Disc exerted ca 5 mm above calyx base.
Stamens diadelphous, free stamen coherent to near the middle; tube
split near base into two groups, 35 mm long, fiee filaments 2-5 mm;
anthers 3 mm long, 1 mm wide, connective apiculate. Gynophore 11 mm,
dark-colored, nearly glabrate near base, npidly sericeous toward apex;
ovary densely sericeous, trichomes white with yellowish tinge, 17 mm
long, 1.8 mm wide; style conspicuously dilated near point of geniculation,
similar to a "cobra hood," nearly 1.5 mm wide, 18 mm long, geniculate
8 mm; stigma terminal, capitate, 1 mm diameter. Legunie unknown.
Figure 26.
Woytkowski's Clitoria is characterized as a tree with short,
densely broad-bracteated inflorescences with zigzag internodes, and
large flowers with silky cálices and exceptionally large bracteoles
subequaling the calyx.
PHENOLOGY: The only collection was made in the immature flowering
state in mid-April.

Figure 26. Clitoria woytkowskii. (a-b) leaflets, x 1; (c) leaf scar
and bud, x 3; (cf)-flower bud with bractcoles, x 1; (e)
calyx, x 1; (f) inflorescence, x 1; (g) ala and carina,
x 1; (it) gynoecium, x 1. (Woytkowski 7231, GH: a,d-e,
g-i; US 2575475: b-c,f.)

1
271

272
TYPE COLLECTION: PERU. Tree 12 in high, flower white, rose
flushed with red veins, in the forest, alt.. 50u m, naposoa, dept. San
Martin, 12 Apr 1962, Woytkowski 7281 (HOLOTYPE: GH. Isotype: US
2575475).
The Gray Herbarium specimen was selected as the holotype because it
has both a flower, which was dissected, and mature leaves. However, no
inflorescence was present. The Smithsonian specimen has smaller leaves
plus a few short inflorescences, but no flowers. Ihus both specimens
were deficient in some of the taxonómica!1y critical structures. The
short inflorescences and conspicuous bracts suggested segregation from
the liana species, but the floral structures provided more evidence
since the comparison of it with other species showed differences.
NOTES: In the description of the floral structures of C.
woytkowskii, some structures do not represent the mature state.
Comparison with flowers of similar stages in tie other two species
(C. flexuosa and pozuzoensis) of section Fltxuosa, indicate that the
calyx, the bracteoles, the androecium, and the gynorcium represent
matured structures. The anthers examined were wel1-developed. Thus,
these measurements should be somewhat reliible. The petal sizes,
especially the alae and vexillum, are smaller than those of the mature
flower, thus, these measurements are unreliable.
This species has close affinities with flexuosa from which it is
easily segregated by its arboreal habit, shorter inflorescences, larger
bracts and petioles, and slightly larger bracteoles, and by its broader
leaflets obovate in shape, and with inconspicuous pubescence below.
DISTRIBUTION (Figure 27): This species is known only from the type
collection in forests near Saposoa, San Ma. tin, Peru, at an altitude of
500 m.

Figure 27. South American distribution of three species of section
FI exuosa. Cl i tor ia f 1 exuosa var. f 1. xuosa (★.), var.
brevibracteoía (Q); C. pozuzoensis var. pozuz.jensis
f. pozuzoensis (♦), var. pozuzoensi■ f. subpaImata (A.),
var. schunkei ; C. woytkowskii .

274

275
9. CJitoria flexuosa Fantz, sp. nov.
Liana, to 30 ft long (teste Belshaw 3328). B, anches to 5 mm
diameter, pith hollow, pubescence puberulciit and uni mate, denser on
juvenile branches, subterete, weakly longitudinally striated, internodes
307 cm; bark dark brown; axillary buds 6-7 mm, scalts ovate, acute,
with minutely appressed pubescence. Leaves 3-foliale, coriaceous,
leaflets elongate, el 1iptic-laceolate to oblong-lani.eolate, occasionally
ovate-lanceolate or oblanceolate, nearly concolorous, apex acuminate,
acumen 2-2.5 cm, more or less mucronate, base rotund, midrib deeply
impressed above, seemingly in a groove, often with short, suberect
hairs, primary nerves of 8-11 pairs, uppet surface glabrous or with
suberect hairs on major nerves, lower surface subpilose with spreading
hairs on veins, or hairs stiff, appressed, becoming sparse to
subglabrate, 8-22 (27) cm long, 3-9 (12) cm wide, latióles
subquadrangular-terete, base darkened, swollen 6-8 mm long, pubescence
moderate to dense, spreading, 4-10 cin; rachis 1-2 (3) cm. Petiolules
subquadrangular, dark-colored, uncinate pubescence with overlying
spreading trichomes, 6-8 mm. Stipules deltoid-lanceolate, acute,
pubescence short, subappressed, moderately dense, 4-7 mm long, 2-3 mm
wide; stipels linear, pubescence short, stiff, appressed, 3-6 mm long,
0.2-0.7 mm wide. Inflorescence axillary and terminal, subpaniculate,
solitary, multiflowered, pubescence on all axes dense, appressed,
reddish-tinged, central axis 2-10 cm, inte, nodes conspicuously zigzag,
bent alternatively in opposite directions, 4-9 mm long, primary branches
which bear pedicels subsessile to 3 mm. Pedicels 4-9 mm. Bracts
conspicuous, middle pair large, ovate, more or less apiculate, concave,
(6) 8-15 mm long, 6-9 nun wide, reflexed in age; inner bract caducous;

276
outer bract deciduous, 6-7 mm long, 4-5 nun wide; b»\. :t at base primary
branch deciduous, 4-8 nun long, 2-5 nun wide, ovate-Lmceolate.
Bracteoles very large, conspicuous, broad-oblong, ol tuse to apiculate,
pubescence minute, appressed, subequaling to somewhat longer than calyx,
occasionally shorter to nearly half the calyx length, (17-25 mm in
short bracteole variety) 23-41 mm long, 9-15 nun widt . Flowers large,
6-8 cm, white to lilac-rose, veins of vexillum pink to dark red. Calyx
pubescence silky, dense, appressed, tube 21-29 (32) mm long, 4-6 mm
wide at base to 8-11 mm wide at throat, lobes del to id-ovate, nearly
glabrous toward margins and apex, acute, 4-6 mm long, 3-4 mm wide,
ventral lobe 6-7 mm. Vexillum pubescence silky, dense appressed,
argenteous in bud becoming reddish-brown, 4-5 cm wide, claw-12-16 mm.
Alae extended beyond keel 7-8 mm, blade 23-30 mm long, 5-12 mm wide,
broadened beyond keel, claw 21-26 mm. Carina falcate, 13-19 mm across,
4-7 mm wide, claw 33-38 nun. Stamens diadelphous, vceillary stamen free,
tube glabrous, split near base into two groups, 39-45 mm, free
filaments 2-6 nun; anthers large, 2.5-3 mm long, 0.7-0.9 mm wide.
Gynophore elongate, 7-10 nun; ovary pubescence white slightly yellowish
tinged, very dense, appressed, 17-18 mm long, 1.5-1.8 mm wide; style
18-26 mm long, geniculate last 7-9 mm; stigma subcapitate, 0.8 mm
diameter. Legume flat, coriaceous, long-stipitate, extending beyond
calyx, pubescence appressed, ca 16-17 cm long, 14-19 mm wide; stipe
33-39 mm long; beak not observed; dehiscence not observed. Seeds brown,
suborbicular, slightly longer than wide, thickened lo 4 mm, surface
smooth, bearing multiple ridges, ca 12 nun long, ca 10 mm wide, 6-7
seeds per pod; hilum oblong, 2 mm x 1 mm. Figures 23 and 29.

Figure 28.
Clj toria fl exuosa - I. Var. fl exuosa: (a-b)
x 1; (c) inflorescence, x 1; (d) flower, x 1;
x 1; (f) bracteole, x 1; (g) androecium, x 1;
x 1. (Spruce 452^, K-441 :a,d. Klug 3820, Gil
BM: e-h).
1eaflets,
(e) calyx,
(h) gynoecium,
b-c;

278

Figure 29. Cl i tori a flexuosa - II. Var. flexuosa: (a) . alyx with
bracteoles and staminal tube, x 1; (b) vexillmu, x 1;
(c) alae and carina, x 1; (d) inflorescence with persistent
calyx and fruit, x 1. Var. brevibracteola: (e) flower,
x 1; (f) calyx with immature fruit, x 1; (g) ihree views
of seed, x 1; (h) node, x 1. (Klj^g 3820, F 7b634: a-c.
Spruce 4527, K-441: d. Kluq 3105, MO 1056939: e-f,h;
G-191T gT7~

280

281
The Flexuose Clitoria is characterized as a woo-/ vine with
inflorescences conspicuously zigzag, large braeteati 1, and bearing
large flowers with conspicuous bracteoles slightly â–  iiorter to longer
than the calyx, or bearing long-stipitate fruits.
PHENOLOGY: The typical variety flowers from August to early March
with var. brevibracteola flowering and bearing fruits in June. Fruits
are rarely collected for the typical variety, the only collection having
been made in February.
TYPE COLLECTION: PEPJJ. Frutex volubili robustos, flores rosei,
Tarapoto, secus r..ilos [sic], Febr. 1856, Spruce 41-27 (HOLOTYPE:
K-441, Herb. Bentham. Isotypes: K-442, Herb. Hooker, W 18669).
Spruce 4527 was chosen as the type because, in addition to the
flowers and zigzag inflorescence typical of most specimens, Spruce
also collected fruit. The specimen of the Kew collection (Herb.
Bentham) has the only fruit and was selected as the ho 1otype. The
leaves tend to be narrower than in other collection , but the flowers
are typical. The Kew specimen (Herb. Hooker) has a typical,
conspicuously zigzag inflorescence for which the specimen was named.
Klug 3820 lias the typical larger leaves, zigzag inflorescences with
many flowers, and a dissected flower (Gray Herbarium). Thus it was
chosen as a paratype. Spruce s.n. from the type locality, lacking
other data, was chosen as a second paratype.
VERNACULAR NAMES: PERU. Raca-sisa (i.e. flos < I itoridis),
Spruce s.n. (K-443).
NOTES: This species has very close affinities with the liana
pozuzoensis from which it can be easily distinguished by the longer
bracteoles and bracts, long-stipitate fruits, shorter staminal tube,

282
short-clawed petals, and foveolate, suborbicular seeds. Cálices are
slightly longer than those of pozuzoens i s and leaf pubescence is
typically pilose.
Several of these collections were cited by Macbride (1943) under the
name C. amazonum, a species which is easily distinguished by its
elongate primary branches in the inflorescence and lack of zigzag
internodes of the main axis, plus the general glabrate appearance.
Woytowski 35227 is a mixed collection with some sheets bearing
vegetative shoots with evenly-pinnate (4-foliate) leaves, with a gland
present on the leaf-stalk between the petiolules (a Cassia). The
Cl itoria flowering shoots of this collection may lack leaves and be
mounted with the non-Clitoria vegetative shoots.
DISTRIBUTION (Figure 27): This species is an endemic in forests
near Tarapoto (San Martin) and Balsapuerto (Loreto), Peru, at elevations
of 300-400 m.
KEY TO VARIETIES:
1. Bracteoles from slightly shorter, to subequal, or up to 1.25
x calyx length, width commonly 10-17 mm, length 23-40 mm;
leaflets short-pilose below, becoming glabrate, trichomes spreading
on nerves (San Martin) 9a. var. flexuosa
1. Bracteoles shorter than calyx, 0.5-0.75 calyx tube length, width
8-10 mm, length 17-25 mm; leaflets pubescence below, Trichomes
minute to short, appressed (Loreto) . . . 9b. var. brevibracteola
9a. Clitoria flexuosa var. flexuosa
Leaves pilose below, becoming glabrate, trichomes on major nerves
ascending, spreading. Bracteoles elongate, subequal to longer than

283
c a 1 y: , occasional i / slightly shorter (0.8 . cai/x), 23-40 mi.i long,
10-1/ mm wide. Flowering commonly from Ai just larch.
111STRI BUT ION (1 igure 27): The typic, i vai iety is an endemic near
Tarajjto, Peru, at elevations of 300-900 i, . P l R U. SAN MARTIN:
Tarapoto, Spruce ..u. (Paratype: K-443); l.c., Mathews 1 584 (CGE-2
sh., E,K); 4 mi E or Tarapoto, 9 Mar 1947, Woytowski 35227 (Mixed,
F-2 :h.,G-3 sh., M0,UC-2 sh.); near Tarapi ro on trail to Juan Guerra
when it crosses Rio Hahuashiyacu, ca 100i ft, 25-27 Aug 1937,
Bel si m 3328 (f,Gil,MO,NY ,UC,US): Juan Jui, Alto Rio Huallaga, ca 400 m,
Dec 1935, K]ucj 41 y9 (A,BM,F,MO,NY,S,U,UC); lc.. Sep 1 934, Klug 3820
(Paratype: BM,F 766344,RH,MO 1105520 & 11)5521, NY,S); Alto Rio
Huai luga, Tarapoto, 360-900 m, Will jams 6, ill (l ).
9b. Clitoria 11 exuosa Fantz var. brr. /ibr. cteola Fantz, var. nov.
Leaves minutely pubescent below (vidi 20x;, appressed. Bracteoles
shortened, one-half to three-quarters of c ilyx length, 17-25 mm long,
8-10 mm wide. Flowering period in June.
TYPE COLLECTION: PERU. Liana, fls. white, red-striped, forest
dealing, Balsipuerto, Dept. Loreto, 220 i, , Jui. 1 933 , Klugj 3105
(H0L0FYPE: G-190 & 191, Herb. Delessert. Isol/pes: A,BM,F 684787,
GH,M0 1065939,NY,S).
This collection has the leaf pubescei e o pozuzoensis and the
larger bracteoles of C. f1exuosa, althougl intermediate in size between
C. f 1 exuosa and C. pozuzoensis. It would ippe,,r to be a hybrid except
that the fruits and the flower structures aid bracts compare favorably
with C. f 1 exuo .a, and not with C. pozuzoei.ds. The one seed differs
from those of C. pozuzoensi s_ by its suborl icuK.r shape and the

284
numerous, minute depressions on the surface of the seed coat. There
are no seeds in typical C. f1exuosa specimens for comparison, but the
fruits agree. Therefore, it seems that this collection belongs to the
species flexuosa, with the differences in biacteole length, leaf
pubescence, and flowering period indicating varietal status, and
supported by apparent geographic separation.
The Geneva specimen, mounted on two sheets within a common cover
folder, was selected as the holotype because it was the only specimen
that had a mature seed and fruit, in addition to the vegetative and
flowering structures. The first sheet has a branch, inflorescence
(juvenile), and leaves. The second sheet has a fruit with leaf, two
flowers, one seed, and a piece of a fruit in a packet.
DISTRIBUTION (figure 27); This variety is known only from the type
locality of Balsapuerto, Loreto, Peru.
10. Clitoria pozuzoensis Macbride, Pub. Field Mus. Nat. Hist.
Bot. 8: 103. 1930.
Liana or rarely an erect shrub, climbing 6-9 in high. Branches
to 5 mm thick, pith hollow, nearly terete, pubescence uncinate and
short-appressed, soon glabrate, seldom branching, internodes typically
1.5-3.5 cm; axillary buds 2-3 mm, scale pubescence appressed; bark
dark brown, splitting longitudinally in strips, lightened beneath.
Leaves 3-foliate, chartaceous to thin-coriaceous, leaflets elliptic-
lanceolate to ovate-laceolate, occasionally obianceolate, apex
acuminate, acum 1-2.5 cm, Liase cunéate to rotund, midrib impressed
above, primary nerves of 9-13 pair, upper surface brilliant green,
glabrous, lower surface green, pubescence minute (vidi 20-30x), stiff,
appressed, 7-16 (19) cm long, 3-9 cm wide. Petioles weak quadrangular

285
to terete, longitudinally striate-caniculate, scattered minute,
appressed pubescence, 3-8 (11.5) cm; rachis slightly laterally
compressed, 1.5-3.5 cm, rarely nearly lacking. Petiolules subquad-
rangular, brown, pubescence rufous, appressed, 5-9 mm. Stipules
deciduous, deltoid, pubescence uncinate and rulous, appressed, 3-6 mm
long, 1-2 mm wide; stipels caducous, linear, 2-5 mm long, 0.5-1 mm
wide. Inflorescence in upper part of branchlets, appearing before
leaves, axillary and terminal, subpaniculate, solitary, multi flowered,
pubescence on all axes dense, rofous, appressed, central axis 1.5-6
(20) cm, internodes short, weakly zigzag, primary branches which bear
pedicels 2-6 mm. Pedicels 5-12 mm. Bracts with rufous, appressed
pubescence, acute; middle bracts paired, ovate, persistent, 3-6 mm
long, 2-4 mm wide; outer bract oval, 3-4 mm long, 2-3 mm wide.
Bracteoles broad, coriaceous, conspicuously shorter than calyx,
oblong to ovate to nearly orbicular, obtuse, pubescence rufous,
appressed, (6) 8-13 (15) mm long, 6-10 mm wide. Flowers large, 6-8
(9.5) cm, white to pale lilac, vexillum streaked pink to red. Calyx
pubescence silky, tawny, appressed, dense, tube 19-25 (27) mm long,
5-8 mm wide at base to 8-12 mm wide at throat, lobes broadly deltoid,
nearly equal, 4-6 mm long, 4-5 mm wide, ventral lobe 2-3 mm wide.
Vexillum 4-6 cm wide, pubescence dense, tawny, appressed, claw 13-15mm.
Alae extended beyond carina 7-9 mm, blade 27-33 mm long, 6-10 mm wide,
claw 26-33 mm. Carina strongly falcate, blade 13-18 mm across, 6-7 mm
wide, claw 39-42 mm. Stamens diadelphous, staminal tube glabrous,
split at the base into two parts, 43-51 mm long, free filaments 2-6 mm;
anthers large, 2-2.75 mm long, 0.5-0.8 mm wide Gynophore 7-8 mm,
with moderately dense, appressed pubescence near apex; ovary pubescence

286
white, dense, appressed, 15 mm long, 2 mm wide; style beard of dense,
uncinate pubescence dorsally, 30-32 mm long, geniculate 9-10 mm from
distal end, broadened below geniculate point, I mm wide; stigma
subcapitate, 1 mm diameter, base pubescence of short trichomes.
Legume long-stipitate, exerted beyond calyx, green, flat, pubescence
inconspicuous (vidi 20x), appressed, ca 14-16 cm long, 18-25 mm wide;
stipe 25-34 mm; beak to 4 mm; dehiscence causing valves to twist
one-half of a turn. Seeds brown, smooth, oblong, conspicuously longer
than wide, slightly thickened 3-4 mm, ca 16 mm long, 10 mm wide.
Figures 30 and 31.
Macbride's Clitoria is characterized as a woody climber with large
flowers bearing the large, primitive bracteoles, much shorter than the
calyx.
PHENOLOGY: This species flowers from April to September. Fruiting
specimens are rarely collected. Juvenile fruits were collected in
July and a mature fruit was collected in September.
TYPE COLLECTION: PERU. Shrub, fls. purplish, Pozuzo, ca 3000 ft,
20-22 Jun 1 923, Mac bride 4652 (H0L0TYPE: F 535727. Isotype: G-292,
Herb. Delessert. Cotype: G-291 , Herb. Delessert, flower and bud only).
Only the specimen at the Field Museum bears the word "shrub" on the
label, the habit description Macbride used in his original description
of the species in 1930 and again in his treatment (1943) of the Peru
flora. All other specimens matching the type bear the word "liana"
or "woody vine." The type specimen is poor, with two leaflet fragments
and a juvenile inflorescence with buds. One mature flower is present
in a packet. A better representation of the species is obtained from
other collections, such as Mexia 7242.

Figure 30. Clitoria pozuzoensis - I. Var. poznzoen is f. pozuzoensis
(a-b) terminal leaflets, x 1; (cl í i¡ flor uscence, x 1;
(d) flover, x 1; (e) calyx and andrneciuu, petals removed,
x 1; (f) vexillum, x 1; (g) alae an ¡ can na, x 1;
(h) gynoecium, x 1. (Mexia 7242, F 8619.7: c-h;
ÃœC 743483: a-b.)

288

Figure 31. Clitoria pozuzoensis. Var. pozuzoensis f. pozuzoensis:
t"aT calyx, x TTTd] juvenile fruit, x 1. Var. pozuzoensis
f. subpalmata: (g) subpalmate leaf, x 1. Var. schunkei:
(b) inflorescence, x 1; (c) flower, vexillum removed, x 1;
(e) fruit, x 1; (f) three views of seed, x 1. (Mexia
7242, F 861987: a. Taylor 3583, Mathias & Taylor 3583
US 2639622: d. Grubb et al. 174: g. Schunke 2773,
F 1688611: b-c,e-f.)"

290

291
VERNACULAR NAMES: PERU: Soga de Machete, Mathias & Taylor 3533;
Vaina huasca, Mathias & Taylor 3583.
ECONOMIC IMPORTANCE: This species niuy be used locally near
Anaconda Island, Rio Napo, Ecuador, as a medicinal remedy administered
to couples to increase fértilization (see p. 15).
NOTES: This species lias close affinities with C. flexuosa which
is easily recognized by its bracteoles subequaling the calyx, large
bracts, conspicuously zigzag inflorescences, and pilose leaflets.
Macbride (1943) indicated that C. pozuzoensis ’seems doubtfully distinct
from C. grandifolia or C. javitensis." These two species are
superficially similar in appearance but different in the cauliflorous
infle.rescences with non-zigzag internodes, minute bracteoles, subsessile
ovaries, smaller anthers, densely pubescent fruit with smaller seeds,
and sparse calyx pubescence.
The leaves are typically pinnately 3-foliate, with a long rachis.
There is a tendency for the rachis to neatly disappear, giving the
leaves a subpalmately compound appearance, a minor variation recognized
as a form. A greater variation occurs in one collection with the
more primitive erect growth and longer inflorescences and pedicels,
and is treated as a variety.
DISTRIBUTION (Figure 27): This species is found in the riverine
forests of Peru and Ecuador at elevations repotted as 135-600 m.
KEY TO VARIETIES AND FORMS:
1. Inflorescence short, 1.5-6 cm; pedicels 5-7 mm; woody vine
10 a. var. pozuzoensis
2. Rachis present, 1.5-3.5 cm; petiole 3-b (11.5 cm)
1Ooa. f. pozuzoensis

292
2. Rachis nearly absent, 0.1-0.4 cm; petiole shorter, 1-4 cm
lOab. f. subpalmata
1. Inflorescence elongated, (5) 8-19 cm; pedicels 10-12 mm; shrub
with climbing apex 10b. var. schunkei
lOaa. Clitoria pozuzoensis var. pozuzoensis f. pozuzoensis
Woody vine. Inflorescence short, weakly flexuosus, 1.5-6 cm.
Pedicels short, 5-7 mm. Leaves pinnately compound, long-stalked,
petiole 3-8 (11.5) cm; rachis conspicuous, 1.5-3.5 cm.
DISTRIBUTION (Figure 27): PERU. HUANUCO: Bosque Nacional de
Imparia, a lo largo del Rio Pachitea cerca del campamento Miel de Abeja,
1 km arriba del pueblo de Tournavista o unos 20 km arriba de la
confluencia con el Rio Ucayali, Dtto. Honoria, Prov. Pachitea, 300-
400 m, 26 Apr 1967, Schunke 1866 (G,NY,US). LORETO: Bosque Nacional
de Iparia del Rio Ucayali, en la trocha a Tabacoa, al oeste de Iparia,
80 km arriba confluencia con el Rio Pachitea, Dtto, Iparia, Prov.
Coronel Portillo, 250-300 m, 22 Aug 1968, Schunke 2655 (F,US); west
bank above Rio Aquaytia, 8-10 km below Aguaytia, 4 Jul 1959, Mathias &
Taylor 3583 (US-2 sh.).
ECUADOR. MANABI: Rocafuerte, Flacienda Arcadia, 200 m,
22 Jul 1933, Heinrichs 51J_ (G,M,NY). ORIENTE: Canyon Ñapo, Rio Tena
near Tena, 11 Apr 1935, Mexia 7242 (F,UC).
lOab. Clitoria pozuzoensis Macbride var. pozuzoensj_s_ f. subpalma ta
Fantz, forma nov.
Woody vine. Inflorescence short, weakly flexuous, 1.5-6 cm.
Pedicels short, 5-7 mm. Leaves seemingly palmately compound, terminal

293
leaflet stalk subsessile, petiole short, 1-4 cm; rachis nearly absent,
inconspicuous, 0.1-0.4 cm.
TYPE COLLECTION: ECUADOR. Liane, fls. white, veins pink,
disturbed ground, forest edge, Puerto Napo, 1500 ft, 14 Jul 1960,
Grubb, LIoyd, Pennington, & Whitmore 174 (H0L0TYPE: K-451. Isotypes:
NY).
This form is easily recognized by the 3-foliate, subpalmately
compound leaves. The Kew specimen had a flowering inflorescence as
well as the vegetative structures, thus it was selected as the holotype.
The label indicated that a duplicate specimen was sent to the Quito
Herbarium (Q). It was not seen by this author.
DISTRIBUTION (Figure 27): This form is known from two widely
separated areas, the type locality in Ecuador and Loreto, Peru.
PERU. LORETO: wooded banks of the lower Rio Huallaga, 135 m,
5-11 Sep 1 929, Kill ip & Smith 29005 (Paratype: NY).
10b. Clitoria pozuzoensis Macbride var. schunkei Fantz, var. nov.
Climbing shrub. Inflorescence elongated, weakly flexuous towards
apex, (5) 8-19 cm long. Pedicels elongated, 10-12 mm.
TYPE COLLECTION: PERU. Loreto: Arbusto trepador de 7-8 m, flores
rosadas, fructors inmaturos de color verdoso pardo; en el interior del
estandarte con lineas longitudinales de color rojo violeta, sépalos
verdes amarillentos; las laminas de las hojas son ondulades, en la cara
superior brillosas de color verde amarillento, en la cara inferior con el
nervio central de color amarillo; en bosque bajo, terreno de aluvión a
1 km abajo de Iparia al noroeste del Rio Ucayali, 250 m, 5 Sep 1968,
Schunke 2773 (HOLOTYPE: F 1688611. Isotypes: 0-187 & 188, NY).

294
This variety is easily distinguished from the typical variety by
its shrubbier habit and elongated inflorescences. The Field Museum
specimen was selected as the holotype because it had fruits, flowers,
and inflorescences. This variety is known only from its type locality.

295
Sect_ic n_ BrachycdJy x
iC. Clitoria L. subgenus Bractearia (Mart ex Benth.) Fantz
section Braehycalyx Fantz, rect.nov.
Irees or tall erect shrub-treelets, rarely lianas. Leaves
deciduous, typically about flowering time, upper surface pubescence
minute hirsute, infrequently glabrous, lower surface conspicuously
pubescent, infrequently inconspicuously pubescent to glabrous. Stipules
deciduous to caducous, lanceolate, acute, narrow, 1-2 mm wide.
Inflorescence nodose-racemose, axillary, usually appearing at denuded
nodes. Calyx tube short, sometimes cup-shaped (campanulate-tubular),
7-15 ii,m long, lobes broadly deltoid, minute nearly lacking, width
usually subequalling length to broader than long, 1-4 mm long, 3-4 mm
wide. Bracteoles minute, 2-5 mm long, 1.5-3 mm wide. Legume pubescent,
weakly to strongly depressed between the seeds at maturity. Seeds
suborbicular to slightly longer than wide, compressed, typically 7-13 mm
diameter, 2-3 mm thick.
The members of the section Brachycalyx are easily recognized by
their short tubular to cup-shaped cálices with minute lobes, inflores¬
cences appearing from denuded nodes, sublomentaceous fruits, and usually
erect habit. With the exception of C^_ glaberrima and C^_ canescens, the
members are also easily distinguished by the minute hirsute pubescence
on the upper leaf surface, which has the texture of sandpaper as one
moves his hands over the surface from leaf apex to its base. Clitoria
sel 1 oi has a short calyx but large calyx lobes and longer bracteoles.
Some individuals of C. javitensis have short lobes of 4-5 mm, but

296
elongated tubes with other individuals having lunger lobes. Both
species are lianas which bear cauliflorous inflorescences and other
structures which agree more with section Cauliflorae. Fruits have not
been seen in C. selloi, out those of C^. javitensis bear seeds similar
in shape and thickness to section Cauliflorae. These two species,
C. sellout and C. javitensis, which bear some superficial resemblance
to section Brachycalyx, are more naturally affiliated with members of
section Cauliflorae.
IlOLOTYPIC SPECIES: C. brachycalyx Harms.
NOTES: This section has closer affinities to section CajjJ ijfl o_rae
than to the other two sections within subgenus Bractearia. Both these
sections lack the large, coriaceous, primitive bracteoles that subequal
and ooscure the calyx. They also possess the racemose, nodose
inflorescences which lack the primary lateral branches. The pedicels
are borne directly from a conspicuous lignose node on the central axis.
In addition, the inflorescence often appears separate from the leaves,
developing from denuded nodes before the appearance of leaves in section
Brachycalyx and being cauliflorous and developing below the leaves in
section Cauliflorae. The members of section Cauliflorae can be
distinguished from section Brachycalyx by their longer cálices with
prominent lobes, by their legumes which are not conspicuously depressed
between the seeds, by the thicker seeds, and usually by the liana habit
and glabrous upper surface of the leaves.
DISTRIBUTION (Figure 10): The members of section Brachycalyx are
found in dry tropical forest or sometimes in open grassy areas or
savannas, usually at lower altitudes to 900 m in elevation, in
northwestern South America and adjacent Panama. Members have been

297
collected from western Guyana to Ecuador, and Panama, with isolated
collections of C L glaberrima recorded in a few countries of Central
America north to Chipas, Mexico.
KEY TO THE SPECIES:
1.Flowers medium-sized, 4-5.5 (6) cm; ovary elongate, 14-18 mm;
legume narrow, 1-1.5 cm wide; stipules minute (caducous), 3-4 mm;
vexillum claw long, 8-10 mm; carina claw long, 20-27 mm, ala
extended well beyond carina, 6-10 mm.
2.Inflorescence elongate, (3) 5-15 (22) cm; leaflets acuminate
(acumen 0.5-1 cm), upper surface glabrous (juvenile leaflets
pubescent along midrib); stamina 1 tube long, 27-33 mm;
anthers 1.5-2 mm; flowering period August to October (SW
Ecuador) 'll. C. brachystegia
2. Inflorescence short, to 2 cm long; leaflets obtuse or abruptly
short-acuminate (acumen to 0.5 cm), upper surface pubescence
minute, subappressed to hirsute; staminal tube short, 24-26
mm; anthers 2-2.5 mm; flowering period January (SE Colombia)
12. C. herniannii
1. Flowers small, 2.5-4 cm; ovary short, 10-13 rum; legume broad,
1.6-2.6 (3) cm wide; stipules conspicuous (caducous), 4-7 (10) mm;
vexillum claw short, 4-6 (8) mm; carina cl aw short, 12-20 mm; alae
shortly extended beyond carina, 4-7 mm.
3. Leaves glabrous on both surfaces; calyx pubescence microscopic,
uncinate; inflorescence very short, to I cm long; bracts large,
3-5 mm; stipe short, 10-14 mm; vexillum claw 7-8 mm; anthers
1-1.5 mm (Central America , northern Venezuela)
13. C. glaiberrima

298
Leaves conspicuously pubescent below, minutes pubescent above
(except in C. canescens); calyx pubescence macroscopic,
moderately dense, short-appressed; inflorescence 1-7 cm; bracts
minute, 1-3 mm, stipe elongate, 24-45 mm, vexillum claw 4-6
(7 ) mm; anthers 1. 5-3 nun.
4.Leaves glabrous above, with appressed pubescence below;
calyx tube 14-15 mm long; stamina 1 tube glabrous; style
18 mm; flowers appearing in late march (S. Venezuela) . .
14. C. canescens
4. Leaves minutely pubescent above, velutinous or pilose
pubescent below; aclyx tube 7-13 mm; staminal tube more
or less uncinate-pubescent towards apex; style 10-16 mm;
flowers appearing in October to February.
5.Trees or tall shrubs; leaf pubescence below moderate to
densely velutinous, petiolules 7-10 mm; calyx tube
shorter, 7-11 cm long; anthers small, 1.5-2 mm; ala
claw 10-17 mm; leaflet apex obtuse or abruptly
acuminate, acumen to 0.5 cm.
6.Calyx cup-shaped, broad at throat, 8-12 mm wide;
flowers white; staminal tube 23-31 mm; pedicels
5-8 mm; carina 10-14 mm x 3-5 mm; bracts minute,
1-2 mm; stipules 4-6 mm; inflorescence longer,
2-7 cm (W. Guyana and adjacent Brazil)
15. brachycalyx
6. Calyx narrowly tubular, throat 5-7 mm wide;
flowers purple; staminal tube 16-22 mm; pedicels
3-6 mm; carina 7-9 mm x 2-3.5 mm; stipules 6-8 mm;

299
inflorescence shorter, 1-3 (5) cm (Venezuela and
Colombia) .16. C_. dundrina
5. Liana; leaf pubescence bel.iw pilose, more dense on
nerves; petiolules 5-7 mm; calyx tube long, 12-13 mm;
anthers elongate, 2.3-3 mm; ala claw 7-9 mm; leaflet
apex obtuse to acuminate, acumen to 1.5 cm (W. Brazil).
17. froesii
II. Cl i tor' i a brachystegia Benth., Bot. Voy. Sulph. 84. 1844.
Terna tea brachystegia (Benth.) Kuntze, Riv. Gen. PI. 1: 210.
1891.
Erect shrub to small tree, 2-8 mm tall, with spreading branches,
d.b.h. = 4 ft ¡Little 6685); bark light gray, smoothish, but with many
cracks (Little 6685). Branches to 8 min diameter, pith dark-col ored,
solid to occasionally hollow, juvenile branches blunt-angular,
longitudinally striated, pubescence moderately dense, short, appressed,
branches becoming subterete, non-striated, glabrous with age, axillary
buds 3 mm. Leaves 3-foliate, subcoriaceous, deciduous, leaflets
ovate-elliptic to elliptic to ovate, rarely obovate, apex acuminate,
acumen obtuse, 0.5-1 cm long, base rotund to rarely weakly retuse,
midrib impressed above, primary nerves 9-12 pair, upper surface glabrous
(juvenile leaflets minutely hirsute, denser towards midrib), lower
surface lighter green, usually pilose with 0.5-0.8 mm hairs, moderate
to dense, occasionally nearly glabrate with trichomes confined mainly
on larger nerves, lamina 7-15 cm long, 4.5-8 (11) cm wide. Petiole
subterete, subpilose, 6-11 (19) cm; rachis 2-4 (6) cm. Petiolules dark,

300
subqundrangular, weakly rugose, moderately densely pilose, 4-6 mm.
Stipules caducous, usually lacking on specimens, linear-lanceolate,
acute, pubescence appressed, 3-4 mm long, I (1.5) mm wide; stipels
caducous, linear, acute, (1) 2-3 mm long, 1 mm wide. Inflorescence
usually appearing before leaves, at denuded nodes, axillary and
terminal, multiflowered, racemose, nodose; axes blunt-angular,
longitudinally striated, pubescence moderately dense, rufus, spreading
to erect, with a few trichomes appressed, axes elongated, 3-15(22) cm
long, occasionally with small-sized, deciduous leaves appearing at the
lower inflorescence nodes during late flowering stages. Pedicels
4-7 mm. Bracts small, deciduous, lanceolate-ovate, acute, pubescence
appressed, dense, rufescent, ca 1 mm long trichomes, usually 2 (3) mm
long, 1-1.5 nun wide. Bracteoles small, oval to ovate, acute deciduous,
at first appressed to calyx, becoming spreading, pubescence appressed,
rufus, dense, 2-3 mm long, 1.2-1.6 mm wide, inserted 1 (2) mm below
calyx. Flowers basically white with creamy to rose to lilac to pale
lavender tints occurring from the medial nerves of the inner face of
the vexillum, medium-sized, 4-5.5 (6) cm long. Calyx short-tubular and
narrow to cup-shaped, pubescence short, appressed, sparse to moderate
with inconspicuous uncinate trichomes beneath, more conspicuous on the
lobes, tube 9-13 mm long, 5-7 mm wide near base to 8-10 mm wide at
throat, lobes minute to nearly lacking, broad deltoid, ciliolate,
1-2 ii.::i long, 3-4 mm broad at base, ventral lobe acute to apiculate,
generally 2 mm long, 2 mm wide at base. Vexillum densely sericeous
outside, blade 3.5-4 cm wide, claw 8-10 mm. Alae well extended beyond
carina, 6-10 mm, blade 17-24 mm long, 6-10 mm wide, claw 14-23 mm.
Carina falcate, 9-13 mm long, 4-5 mm wide, claw 21-27 mm. Stamina 1

301
tube 27-33 mili, tree filaments 2-4 mm, gl ibron .; anthers lanceolate,
1.5-1 nun long, 0.25-0.5 mu wide. Gynophoi â–  ca w mm; ovary 14-18 mm
long. 1 mm wide, pubescence appressed, den e, i site; style 15-18 mm,
geniculate the 6-:; mm from the distal end; stiVma dilated, 0.5-1 mm
diameter, pube cence dense at base. Legun â–  stipitate, exerted v;e 11
beyor. 1 the calyx, pubesc nee moderate to d .-nsel / uncinate with scattered,
short, suberect, macroscopic trichomes, valves flat, depressed between
the seeds, coi laccous, 15-19 (24.5) cm long, 12-17 mm wide; stipe
24-21 rum; beat when present, to 12 mm long; dehiscence causing valve to
twist (1.5) 3-4 turns. Seeds black, smootn, compressed-lenticular,
face .uborbicular to oblong, slightly longrr than wide, generally
9-13 nm long, ,-12 nun wide, 2 mm thick, 10-12 reeds per pod, hilum
obi or, ], 2 mm x 1 mm. F i jure 32.
oentham's Clitori a is characterized a, a curub-treelet with
elongate inflorescences bearing numerous 1 lower s that have a short,
cup-soaped calx, and legumes which are ccispir lously depressed between
the seeds.
TIEN0L0GY: This species was collects I in rlower from July to
Octol r, with mature fruits appearing in Gtober. One collection of
late lune and small buds. This species is unigue in the section
Brack /calyx with its phenology as an earl i r bloomer, as all other
species blossom from October through February, with one poorly
colls ted species blooming in late March.
TYPE C0LUCTI0N: ECUADOR. Guyaquil, Dr. bind air s.n. (LECT0TYPE:
K-32, hb. HooI cri-muin, 1 T67; S - photo of K sp< imen). Bentham cited
only ihe Sinclair s pec in n but did not inoicati where it was deposited.
The oily duplicate seen was selected as th let r.otype, although it is

Figure 32. Cl i tori a brachystegia. (a) leaflet, x 1; (b) juvenile
inflorescence, x 1; (c) inflorescence, x 1; (d) calyx,
x 1; (e) vexillum, x 1; (f) ala and carina, x 1;
(g) androecium, x 1; (h) gynoecium, x 1; (i) flower with
immature fruit, x 1; (j) legume, x 1; (k) three views of
seed, x 1. (Holway & Holway 1000, GH: a,c-i. Hitchcock
21229, NY: b. Wiggins 10895" US 2170652: j-k. 1

303

304
atypical in that the leaves are nearly gla orate. However, other
colle tions have both the typical pubescent leaves as well as those that
are nearly glabrate. This specimen is pronably a duplicate of the type
specimen that Bentham saw because it was added to Hooker's herbarium
over two decades after Bentham described the species. The location of
the specimen that Bentham saw is unknown.
VERNACULAR NAME: ECUADOR: Flor de la reina, Me_xia_ 6557.
NOTES: The origin of Bentham's specific epithet "brachystegia"
(brachy - short; steyium = a roof or covering) is unclear. Based upon
the specimens of C1itoria species that Bentham saw, this species is
distinctive with its short, cup-shaped calyx. One can speculate that
Bentham's name was based upon the short calyx which covered the floral
and reproductive structures. The only other species with a cup-shaped
calyx was described by Harms in 1921, well after Bentham's death.
Clitoria brachycalyx Harms is easily distinguished by its broadly oval
to suDorbicular leaves, densely velutinous below and minutely hirsute
above at maturity, its shorter inflorescences bearing smaller flowers,
and the much longer stalked fruits.
DISTRIBUTION (Figure 33): This species is an endemic in the
Ecuadorian states which surround the Gulf of Guayaquil. Scant data are
available on ttie habitat, recorded once from "open woods" and once from
"alongside stream." Individuals have been recorded at elevations from
50-1150 m.
F C U A D 0__R. LOS RIOS: Hacienda Santa Lucia, Canton Vinces, ca
50 m, 18-28 Oct 1 934 , Mexia 6557 (NY,US). GUAVAS: secus fluvium
Daule, ca 400 m, Jul 1876, Andre 2571 (NY): Huayaquil, Ririz & Pavón s.n.
(BM); Sta. Ana, pr. Guayaquil, Sodiro s.n. (P); prope Guayaquil,

Figure 33. South American distribution of six species of section
Brachycalyx. Clitoria brachycalyx (A); C. brachystegia
TWT; C. canescens TO) ; C, froesi i (dm¿¡; dendrina
( •); C. hermanni i (■&).

306

307
Oct 1 927 , Mil le s.n. (F). EL ORO: between Hacienda Ingenio to Piedras
over Cordillera Marcabeli, 666', 21 Jun 1943, Little Jr. 6685 (F,UC,US);
between Curtincapa & Portovelo, 640-1555m, 18 Aug 1943, Steyermark
53971 (F,M0); vie Portovelo, 6-15 Oct 1918, Rose & Rose 23372 (NY,US);
near junction Rio Luis & Rio Amobocas, due S of Portovelo, 2200-2500',
6 Oct 1 944, Higgins 10895 (MO,NY,US); Rios Amarillo y Calera, ca de
Portovelo 680-710 m, 14 Aug 47, Espinosa 1772 (NY); Portovelo, gold mine
near Zaruma, 600-1000 m, 30 Aug-1 Sep 1923, Hitchcock 21229 (GH.NY);
1•c., 22 Sep 1920, Hoi way & Hoi way 1000 (GH); Zaruma, 1150 m, 12 Aug 47,
Espinosa 1687 (NY).
UNITED STATES. FLORIDA: Highlands Co., vie. Avon
Park (cult.), 10 Aug 51, McFarlin 6453 (MICH).
12. Clitoria hermannii Fantz, sp. nov.
Erect shrub, to 2 m tall. Branches to 4 mm diameter, pith solid,
pubescence moderately dense, short, appressed to suberect, internodes
2.5-4 cm; axillary bud 3-3.5 mm. Leaves 3-foliate, subcoriaceous,
leaflets suborbicular becoming broadly oval as enlarging, apex obtuse,
abruptly minutely acuminate, base weakly cordate to rotund, midrib
impressed above, primary nerves of 8-10 pair, secondary nerves
prominently parallel between the primary nerves, upper surface
pubescence moderately soft, short (ca 0.5 mm), appressed, lower surface
velutinous, juvenile leaflets 4-7 cm diameter, lamina becoming 10-12 cm
long by 7-9 cm wide. Petiole elongate, 6.5-15 cm, pubescence appressed,
dense; rachis (1.5) 2-2.5 cm. Petiolules quadrangular, tomentose,
5-8 mm. Stipules caducous, not seen; stipels caducous, linear, acute,

308
spars: pubescent, 3-nervad, 2-2.5 mm long, 0.5 nu wide. Jng lore scene e
axilliry, at denuded nod:s, subpaniculate axes crowded toward brancti
apex, pubescem e dense, ippressed, central axi short, 1.5-2 cm, primary
later tl brandies which bear the pedicels short, subsessile to 4 mm.
Pedicels 5-7 mm. Bracts persistent, middle pair, ovate, obtuse,
pubescence dense, appressed, 1.5-2 mm long, 1 mm wide. Bracteoles
minute, ovate to suborbieular, obtuse, pubescen e dense, appressed,
2-3 n i long, 2 mm wide, inserted at the base or the calyx. FIowers
white to faintly orchid, medium-sized, 4.8-5.2 cm. Calyx short-tubular,
pubescence dense, appressed, tube 11-13 mm long, widtii 5-6 mm wide at
base to 8-9 mm wide at tnroate, lobes minute, broad ovate-deltoid, 3-4
mm long, 2-2.5 mm wide, ventral lobe 4-5 mm long. Vexillum pubescence
dense, appressed, tawny, blade ca 3 cm wid , cl iw broadly cuneate,
8-9 in i. Alae • xtonded well beyond carina, 6-9 min, blade 19-23 mm long,
6-10 mm wide, claw 13-15 mm. Carina blade fair ite, 10-12 mm long,
4 mm wide, claw 21-24 mm. Stamens diadelphous, vexillary stamen
coherent near middle with tube, free above and below; tube 24-26 mm
long, pubescence uncinate near apex, free filaments 2-5 mm; anthers
lanceolate, 2-2.5 mm long, 0.5 mm wide, connective long apiculate.
Gynophore 5 mm, with a tow spreading white trichomes near apex; ovary
14-15 mm long, 1.25 mm wide, pubescence white slightly tinged yellow;
style 16-17 mm long, geniculate 7-8 mm from distal end; stigma
subcapitate, 0.5 mm diameter. Legume unknown. Figure 34.
Hermann's Clitoria is characterized as a shrub with short,
subpaniculate inflorescences bearing medium-sized white flowers that
have j shortened calyx with minute lobes, and tears leaves which are
soft ibove and velutinous below.

Figure 34. Clitorja hermanni i. (a-b) leaflets, x 1; (c-d)
inflorescences, x 1; (e) flower, x 1; (f) vexillum, x 1;
(q) calyx, x 1; (h) ala and carina, x 1; (i) androecium,
x 1; (j) anther with apiculate connective, x 5; (k)
gynoeciuin, x 1; (I) leaf scar and axillary bud, x 1.
(Hermann 11019, NY: b. Hermann 1J3994, GH: a; NY: c-1 . )

310

311
PHENOLOGY: Collections of this species were made with flowers in
mid-January.
ÍYPE COLLECTION: COLOMBIA. Vichada: shrub, 18 in., almost
leafless, fl. white, grassy llanos, along Rio Vichada, San Jose de
Ocune, 100 m, 19 Jan 1944, Hermann 10994 (MOLOTYPE: NY. Isotype: GH).
The specimen selected as the holotype had the best flowering
material, including a dissected flower in the packet, which this author
placed in a cellophane envelope.
NOTES: Tiiis species appears similar to C. dendrina and C.
brachycalyx in appearanee of its leaves, but has larger flowers and
smaller stipules, with a number of other structures having close
affinities to C._ brachystegia. Cl i toria brachystegia is easily
distinguished by its elongated inflorescences and acuminate leaflets
which are pilose below.
DISTRIBUTION (Figure 33): This species is known only from the
type locality near San Jose de Ocune, Colombia, at an altitude of 100 m,
in grassy llanos.
COLO M B I A. VICHADA: 6 km NE of Masaguara on llanos along
Rio Vichada, ca 36 km NE of San Jose de Ocune, 100 m, 21 Jan 1944,
Hermann 11019 (Paratypes: NY, US 1879110).
13. Clitoria glaberrima Pittier, Bol. Soc. Venez. 8: 264. 1943.
Tree, 3-12 m tall, 3.5-38 cm dbh. Branches 3-7 mm thick, pith
hollow, terete, longitudinally striate to sulcate, pubescence of
juvenile branches uncinate and short-appressed, soon becoming glabrous.
Leaves 3-foliat.e, light green concolorous, thick-membranous, leaflets

312
ovate to ovate-elliptic, apex abruptly narrowed to an acumen 1-2 cm
long, 2-8 nun wide, extreme apex acute to obtuse, mucronate, base rotund,
midrib impressed above, primary nerves of 7-10 pair, upper surface
glabrous, lower surface glabrous or occasionally with sparse trichomes
subappressed on midrib, lamina 6-16 cm long, 3-9 cm wide, terminal
leaflet conspicuously larger than lateral leaflets. Petioles striate
longitudinally, somewhat twisted, glabrous to sparsely pubescent,
3-9 cm; rachis 1-4 cm. Petiolules subquadrangular, rugose, glabrous,
5-7 (9) mm. Stipules caducous, lanceolate, acute, pubescence uncinate
and sparse, short, appressed, apex ciliolate, 5-7 mm long, 1.5-2 mm
wide; stipels caducous, linear to subulate, 4-7 mm long, 0.4-0.8 mm
wide. Inf1orescence axillary and at denuded nodes, contracted, several-
flowered, racemose, axis pubescence sparse, appressed, to 1 cm long.
Pedicels with uncinate pubescences 3-5 mm long, becoming 7-9 mm long
and 1ignose in fruit, 2-3 mm thick. Bracts ovate to ovate-lanceolate,
acute, pubescence uncinate, middle bracts persistent, 3-5 mm long,
1-2 mm wide. Bracteoles deciduous, lanceolate, acute, pubescence
uncinate, 3-5 mm long, 1.5 mm wide, inserted 0.5-2 mm below the calyx.
Flowers white with deep blue to purple marginally on the vexillum and
darker purple veins within, alae and carina white to white tinted
purplish, flowers small, 2.5-4 cm. Calyx with uncinate pubescence to
glabrate, tube short, 7-12 mm long, width 3-5 nun near the base to
5-7 mm wide at the throat, lobes broadly deltoid, acute to rarely short
acuminate to subulate, cioiolate, 2-3 mm long, 2-3 mm wide at base,
rarely to 5 mm long with the subulate tip present, ventral typically
ca 1 mm longer. Vexillum with uncinate pubescence, blade 2-2.5 cm
wide, claw 7-8 mm. Alae extended beyond carina 4-6 mm, blade

313
oblong-spatula te, 12-18 :iun long, 5-7 mm wile, ciaw 8-12 mm. Carnia
falcate, blade 8-10 mm long, 4-5 mm wide, claw 14-21 mm long. Staminal
tube nearly straight, glabrous, 18-21 mm long, free filaments 2-3 mm;
anthers 1-1.4 mm long, 0.6 mm wide. Gynophore 1-2 mm, uncinate-
pubescent; ovary 10-13 mm long, 1 mm wide; pubescence sericeous, white,
style 11-13 mm, geniculate 5-7 mm from distal end; stigma capitate, ca
0.5-0.6 mm diameter, pubescent around base. Legume short-stipitate,
more or less included within the calyx, yellow-green, coriaceous, flat,
depressed between the seeds, pubescence uncinate and sparsely pilose,
(8) 11-20 cm long, 16-20 mm wide; stipe with uncinate-pubescence, 10-12
(14) mm long, 2-3 mm thick expanding to 5-7 mm thick apex, weakly
twisted; beak 5-9 mm; dehiscence 1-1.5 turns. Seeds brown to very dark
brown, smooth, compressed lenticular, face suborbicular to slightly
longer than wide, base truncate, 8-10 mm long, 7-10 mm wide, 2-3 mm
thick, (4) 7-12 seeds per pod. Figure 35.
The Glabrate Tree CIitoria is characterized as a tree with glabrous
leaves, short inflorescences bearing small whitish-purple flowers and
fruits which are sublomentaceous, and possessing a short tubular calyx
with minute lobes and inconspicuous uncinate trichomes.
PHENOLOGY: Flowers are borne from August to December (rarely
March) with juvenile fruits appearing in October and maturing through
January. Two collections from June and July bore old fruits or
fragments of the legume valves.
TYPE COLLECTION: VENEZUELA. Miranda: Tree, ±4 m, thickets near
town, Santa Lucia, 160 m, 6 Mar 1 943, Kill ip & Tamayo 37021 (LECT0TYPE:
VEN 4009! Isotypes: F 1158369! GH! NY! K-not seen).

Figure 35. Clitoria glaberrima. (a-b). Leaflets, x 1; (c) inflorescence,
x 1; (dj flower, x 1; (e) calyx, x 1; (f) vexillum, x 1;
(g) ala and carina, x 1; (h) ardroecium, x 1; (i) gynoecium,
x 1; (j) legume, x 1; (k) three views of seed, x 1.
(d'Heguert 837, NY 2486: b, e-i . Pit tier 37021 , A: a,c-d.
Lewis, Dwyer, & Elias 8, US 2572166: j. Pittier 6647,
US 716707: k.)

315

316
Pittier cited three collections, indicating the Ki11ip & Tamayo
37021 as the type, but failed to eastablish at which institution the
specimen he had examined was located. The Caracas specimen (VEN) is
treated as the holotype because this author is convinced that it was the
specimen examined by Pittier and labeled by him as the type. This
conclusion is based on the following evidence: (1) the VEN specimen is
marked "tipo" whereas all other specimens are marked "type coll.";
(2) the Caracas Institution is the only herbarium known to have one
sheet of each of the three cited collections; (3) the Caracas specimen
was the only one to have fragments of the legume, which was described
by Pittier; (4) the Caracas specimen was also the only one to possess
a dissected flower; its measurements agree with the published
description; (5) the Caracas specimen included attached sheets which
illustrated the flower structures, some of which were published with
the description; (6) the description was published in a technical
bulletin associated with the Caracas Institution; and (7) Pittier was
associated with the Caracas Institution, thus, more likely to have
examined a VEN specimen than one from another institution. From the
collective evidence, the Caracas specimen (VEN 4009) is the holotype.
VERNACULAR NAMES: VENEZUELA (Miranda); Carauto, Pittier 7074;
(Carabobo) Anima, Saer 837, PANAMA (Panama): Peronil, Pittier 6799.
EL SAiVADOR: Chapelno, Calderon 1228 & 1423.
NOTES: Although Cl i tori a glaberrima has been frequently collected
in Panama, it remains a relatively unknown species to botanist. As a
result, these specimens were usually identified as C. arborescens, a
name historically applied to an array of arborescent-like Clitorias.
Part of the confusion may lie in R. Brown's description of C. arborescens,

317
"C. foliis ternatis, pedunculis multiflori., germine tomentoso, stylo
villoso, caule arborescenti," a description so vague that it could fit
almost all of the members of the subgenus Bractearia. A second
Panamanian species, C. j-ivitensis, was usually included with
glaberrima under the name arborescens. Each name applies to a
distinct species. Clitori a arborescens is easily distinguished from
C. gldjDerrmma by having leaves tomentose below, elongated inflorescences,
conspicuously long bracteoles (8-15 mm), and medium-sized flowers
(4-6 cm). Clitoria javitensis is easily distinguished by its large
flowers (5.5-8 cm), elongate calyx tube (17-24 mm), ovary with rufus
trichomes, and fruit with dense rufus pubescence. Croat (1974) was
the first to associate correctly the name CL glaberrima with the
Panamanian members of this species.
The Venezuelan material usually differs from the Central American
specimens by usually having stipels and calyx teeth with an aristate
apex. However, occasionally one can observe a flower with aristate
calyx lobes on a Central American specimen which also has other flowers
whose calyx lobes are acute, lacking the subulate tips. Although the
minute stipels were not seen in Central American specimens, stipel scars
were evident. This distinction does not warrant nomenclatura!
recognition.
DISTRIBUTION (Figure 36): This species is found in dry tropical
forests from Venezuela to Southern Mexico, but it is poorly represented
in collections outside of Panama, with four collections from Venezuela,
and one collection each from Colombia, Costa Rica, El Salvador,
Guatemala, and Mexico. It has been collected at elevations from
150-1/00 m.

Figure 36. Central and South American distribution of Cl i tori a
glaberrima, section Brachycalyx.

319
^ \

320
VENEZUELA. MIRANDA: bosques de Galipan arriba de Caracas,
1600-1700 m, 4 Nov 1917, Pittier 7576 (Paratype: VEN 4011); Cardenas,
Guinard Estate, Squire Valley, 400-800 m, Pittier 7074, 25 Mar 1917
(G-2 sh.) and 5-7 Apr 1917 (G-2 sh.,US). CARAB0B0: orillas del Rio
Cabrialeo, Carabobo, Nov 1941, Saer 837 (Paratype: VEN 4008.
Isoparatype: NY).
C 0 L 0 M B I A. Carthagena, Miller s.n. (BM).
PANAMA. 6 Jun 1914, Pittier 6647 (US); Toboga Isl., 30 Oct
1917, Ki11ip 3180 (US); l.c., on the Morro, 1 Oct 1862, Hayes 687
(K,M); Al hojuela, Bro. Paul 496 (MICH,US). LOS SANTOS: Las Tablas,
28 Jul 1962, Dwyer 1180 (MO). PANAMA: Ag. Exp. Station at Matias
Hernandez, 8 Oct 1914, Pittier 6799 (GH,K,M,P,US); Isla Tabaguilla,
20 Oct 1962, Duke 5882 (MO). CANAL ZONE: Farfan Beach Area, 23 Oct
1 965, Tyson 1833 (MO.US) and 1839 (MO); l.c., 19 Dec 1 965, Tyson & Blutn
2606 (M0,US) and 2618 (MO.US) and 2619 (MO) and 2620 (MO); l.c.,
26 Jan 1966, Tyson, Dwyer, & Blum 3172 (GH,M0); hill above Thatcher
Ferry Bridge, 22 Aug 1971, Croat 17012 (MO); Madden Dam, 50 ft,
3 Dec 1966, Lewis,Dwyer, & Elias 8 (M0,UC,US); l.c., 1938, Woodson,
Allen, & Seibert 1553 (GH,M0,US); l.c., Boy Scout Camp Road, 22 Jan 1968,
Dwyer 8387 (MO); l.c., 27 Aug 1969, Dwyer 9166 (MO); 1 ini beyond Madden
Dam Bridge, 12 Oct 1967, Correa & Dressier 353 (MO).
COSTA RICA. El Coyolar, Jan 1931, Lankaster 1283 (F,S).
GUATEMALA. Intro, en Envoy Java et la Tortuga, Nov 1877,
Bernoul1 i & Cajro 1177 (K).
L_L SALVADOR. La Cebadilla, 1 922, Calderon 1 223 (GH,NY)
and 1428 (MO-2 sh.).

321
M E X I C 0. CHIPAS: Pijijiapan, 200 m, 24 Dec 1949, Matura
18734 (NY).
14. Clitoria canescens Pittier ex Fantz, sp^ nov.
Cl 1 toria canescens Pittier, nom. in sched.
Shrub. Branches 4-6 mm diameter, subterete, pith solid, internodes
weakly zigzag, juvenile branches with pubescence uncinate plus scattered,
short, appressed trichomes, becoming glabrous; bark dark brown, smooth,
splitting in longitudinal strips, dark tan beneath; axillary buds 3 min,
scales ovate, short strigose; leaf scar suborbicular, vascular bundle
scars inconspicuous. Leaves 3-foliate, coriaceous, long-petiolate,
leaflets lanceolate-elliptic, apex acuminate (acumen damaged in
specimen, mostly lacking), base broad cuneate, midrib impressed above
appearing as a shallow groove, primary nerves of 9-10 pair, upper
surface faintly glossy, dark green, glabrous, lower surface dull, green,
pubescence short, appressed, inconspicuous on nerves, lamina 15-17 cm
long, 6-7 cm wide. Petioles subquadrangular-terete, 9 cm long;
pubescence scattered, uncinate with some appressed trichomes; rachis
3.5 cm. Petiolules subquadrangular, rugose, dark-colored, glabrate,
8-10 mm. Inflorescence terminal, racemose-nodose to subpaniculate (?),
pedicels borne on central axis at lignose knobs or occasionally on a
secondary lateral branch, 6-7 mm long; axes dark-colored, compressed
laterally, subquadrangular, twisting, pubescence moderately dense, short,
appressed, central axis 1.5-3.5 cm long or more (??, apex broken).
Pedicels strigose, dark-colored, 6 mm. Bracts minute, ovate, acute,
strigose; middle pair semipersistent, 1-2 mm long, 0.5-1 mm wide, outer

322
bract deciduous, narrow, 1 mm long, 0.5 mm wide. Bracteoles minute,
ovate, inserted at calyx base to 1 mm, strigose, acute to short
acuminate, 2-3 mm long, 1-1.5 mm wide. Flowers small, 3.5 cm long,
color unknown (purple? - dry state appearance similar to dendrina
which has dull to dark purple flowers). Calyx dark-colored, pubescence
moderate, short, appressed (buds densely pubescent, subcanescent),
tube short, 13-15 mm long, width at base 3-4 nun broad to 7 mm wide at
throat, lobes minute, broad deltoid, short acuminate, 3-4 mm long,
2.5-3 min wide at base, ventral lobe linear, 6 mm long, 0.5-0.7 mm wide.
Vexillum dark-colored, pubescence dense canescent dorsally, trichomes
appressed, minute, blade obovate, 1.5 cm wide, claw 6-7 mm long, 2-3 mm
wide, cuneate. Alae extended beyond the carina 4-5 mm, blade 17 mm
long, 3-5 mm wide, claw 16 mm. Carina flacate, acute, blade 16 mm
long, 4-5 mm wide, claw 20 mm. Androecium diadelphous, stain i na 1 tube
glabrous, 28 mm long, free filaments 3-5 mm, alternately long and short;
anthers lanceolate, 3 mm long, 0.5 mm wide. Gynophore 3 mm long
brownish-black, moderately densely sericeous, trichomes white; ovary
13 mm long, 1 mm wide, blackened, pubescence very dense, sericeous,
white; style blackened toward base, white sericeous pubescence towards
base, beard moderately dense, 18 mm long, geniculate 3-4 mm from apex;
stigma dark-colored, compressed, 0.5 mm diameter. Legume unknown.
Figure 37.
Pittier's Canescent Clitoria is characterized as a shrub with short
racemose-subpaniculate inflorescences bearing small canescent flowers
that have a long brachycalyx with minute lobes.
PHENOLOGY: The only collection of the flowers was made in late
March.

Figure 37. Cl i tori a canescens. (a) leaflet, x 1; (b) two inflores¬
cences, x 1; (c) flower, x 1; (d) calyx, x 1; (e) vexillum,
x 1; (f) ala and carina, x 1; (g) androecium, x 1;
(h) gynoecium, x 1; (i) node with axillary bud, stipule
scar, and base of petiole, x 1. (Williams 14893, VEN
4038: a-i.)

324
J

325
TYPE COLLECTION: VENEZUELA. Amazonas: itios arenoso de la boca
del Gahama, Cano S. Miqnel , Guaima, alto Rio Negro, 26 Mar 1942,
Willliams 14093 (H0L0TYPE: VEN 4038).
This specimen bore the name "Clitori a canescens Pittier, sp. nov."
on the label and was placed in a separate folder bearing this name. No
record of the publication of this species has been found. Pittier
(1944) did not include this species in his key to the Clitorias of
Venezuela. Since the specimen was collected in 1942, it is presumed
that Pittier saw this specimen after his publication, at which time he
recognized it as a new species and gave it a name.
NOTES: The dried flowers of this species suggest C. dendrina more
than any other species, although C. canescens differs by the canescent
vexillum, longer style, calyx tube, and some of the petal structures.
In addition, the leaves are glabrous above, the inflorescence is less
crowded with flowers and bears secondary lateral branches, and the
ovary pubescence lacks any yellowish pigmentation. The leaves appear
similar to brachystegia and C. glaberrima. Clitori a brachystegia is
easily distinguished by its larger flowers, its elongated inflorescences,
and many differences in reproductive structures. C1itori a glaberrima
is easily recognized by its shorter inflorescences, calyx with uncinate
pubescence, glabrous leaves, and differences in several flower structures.
There is also a superficial resemblance to some specimens of C.
javitensis which have immature flowers present. This species can be
easily distinguished from C. canescens by the longer calyx tube (16-24
mm) and lobes (4-6 mm), the long-clawed vexillums (16-19 mm), and
ovaries with rufus pubescence. The minute calyx lobes and smaller
flower size, plus a shorter calyx, suggest that C. canescens is better

326
placed in section Brachycalyx where it has close affinities with
C. dendrina and some of the characteristies of C. brachystegia &
C. glaberrima, than placed in section Cauliflorae next to C. javitensis
which it superficially resembles.
DISTRIBUTION (Figure 33): This species is known only from the type
col lection.
15. Clitoria brachycalyx Harms, in Fedde Report. Spec. Nov. Regni
Veg. 1_7: 444. 1 921 .
Erect shrub to small tree, 5-12 m tall. Branches to 8 mm diameter,
pith solid to occasionally hollow on larger sized branches, juvenile
branches blunt-angled, somewhat subquadrangular, soon becoming terete,
longitudinally striated, pubescence dense, rufus, erect to spreading
with a few appressed trichomes, becoming glabrate with scattered
macro-trichomes and uncinate trichomes now conspicuous, then glabrous;
axillary buds 2 turn long, scales ovate, acute, pubescence strigose.
Leaves 3-foliate, thick membranous, leaflets broad ovate to elliptic-
ovate or ovate, apex generally broadly obtuse, occasionally obtuse and
abruptly acuminate, acumen to 0.5 cm, infrequently slightly longer, more
or less mucronate, base broad rotund, midrib impressed above, primary
nerves of 7-9 pair, upper surface pubescence of moderately dense,
minute, erect stiff hairs (ca. 0.1-0.2 mm), texture similar to fine
sandpaper, lower surface slightly lighter green, pubescence short
velutinous, conspicuously dense along the nerves, soft to the touch,
lamina generally 8-16 cm 6-12.5 cm wide. Petioles subterete, (5)
7-15 (18) cm; pubescence of uncinate and sparse to dense, appressed,

327
short macro-tr ichomes; rachis 3-4.5 cm. Petiolules subquadrangular,
pubescence dense, rufus, spreading, 7-10 nun. Stipules deciduous,
lanceolate, weakly striated, often weakly arcuate, 4-6 mm long, 1-2 mm
wide; pubescence moderate, appressed, ciliate; stipels linear to
lanceolate, 3-7 mm long, 0.3-1 mm wide, uncinate, oubescent with a few
appressed trichomes, terminal stipel narrower and shorter.
Inflorescence appearing before the leaves, from denuded nodes and
axillary, racemose, nodose; axes angular, slightly twisting, pubescence
dense, minute, erect to appressed, 1.5-7 cm long, mu1tiflowered, lignose
and similar to old branches in fruit. Pedicels 5-8 mm long, lignose
and elongated to 7-12 mm in fruit. Bracts minute, inconspicuous, ovate,
broadly acute, pubescence appressed, middle pair 1-2 mm long, 1-1.5 mm
wide. Bracteoles minute, inconspicuous, ovate, broadly acute, weakly
striated, 2-3 mm long, 1.5-2 mm wide, inserted 1-2 mm below the calyx
base in flower to occasionally located 2-4 mm below the calyx in fruit,
pubescence appressed. Flowers small, 2.5-4 cm, white with cream to
pink to lilac veins on the ventral face of the vexillum. Calyx short,
cup-shaped, pubescence dense, minute, appressed to suberect, tube 7-11
mm long, 5-7 nun broad at base expanding to 8-12 mm at throat, width of
tube at throat subequaling its length, lobes minute, broadly deltoid,
apex broadly acute, width slightly more than length, 2-3 mm long,
3-4 nun wide at base, ventral lobe subequaling other lobes, slightly
narrower. Vexillum pubescence dense, tawny, appressed, blade 2.5-3.5 cm
wide, claw 4-5 mm. Alae extended beyond carina 4-7 mm, blade 15-22 mm
long, 5-10 mm wide, claw 10-17 mm, with densely uncinate pubescent.
Carina falcate, with uncinate pubescence, blade 10-14 mm across, 3-5 mm
wide, claw 16-1 mm. Stum i nal tube 23-31 mm long, more or less

328
uncinate pubescent at apex, free filaments glabrous, 2-4 mm; anthers
lanceolate, apiculate, 1.7-2.3 mm long, 0.5-0.8 mm wide. Gynophore
6-7 mm, moderately dense sericeous; ovary 11-13 mm long, 1.2-1.5 mm
wide, pubescence sericeous, white; style 11-16 mm long, geniculate
6-8 mili from distal end; stigma capitate, 0.5-1 mm diameter, densely
pubescent at base. Legume 1ong-stipitate, exerted well beyond the
calyx, flat, becoming depressed between the seeds, base attenuate,
12-19.5 cm long, 16-26 (30) mm wide, pubescence dense, short, erect,
reddish-brown, soft to the touch; stipe short, erect, 28-45 mm long,
2-3 mm thick expanding to 4-5 mm at apex, pubescence moderately dense;
beak 6-15 mm when present; dehiscence causing values to twist one-
quarter to three-quarters of a turn. Seeds lenticular, smooth, face
orbicular to slightly wider than long or occasionally longer than wide,
brown soon turning black beginning near the placenta, generally 10-13 mm
long, 10-13 mm wide, 2-3 mm thick, (4) 7-10 seeds per pod; hilum
2 mm x 1 mm. Figure 38.
Harm's Cl itoria is characterized as a tree with leaves conspicuously
pubescent on both surfaces and bearing small flowers with a conspicuously
short, cup-shaped calyx, and developing rufo-pubescent, sublomentaceous
fruits which are well exerted beyond the calyx.
PHENOLOGY: This species blooms from October through February with
mature fruits collected in March and April. Leaves are deciduous during
the flowering period, usually appearing after the flowers have appeared
on the plant. One collection, Irwin 647, was collected with flowers in
mid April .
TYPE COLLECTION: BRAZIL. Rio Branco, Surumu, camp de Serra de
Mairary, Feb 1909, Ule 8398 (LECTOTYPE: K-31).

Figure 38.
Clitoria brachycalyx. (a) leaflet, x 1; (b) inflorescence,
x 1; (cl calyx, x 1; (d) vex illum, x 1; (e) ala and carina,
x 1; (f) androecium, x 1; (g) gynoecium, x 1; (h) legume,
x 1; (i) three views of seed, x 1. (Smith 3393, S: a,h-i.
Smith 3128, S: b g.)


331
Harms cited only the U_l_e collection, but did not designate the
location at which the specimen was deposited. It was probably deposited
at Berlin, and therefore, now destroyed by fire during World War II.
The Kew specimen is the only type duplicate seen, and therefore was
selected as the lectotype. However, this specimen does not bear any
leaves, which must have been present on the type collection since Harms
described the leaves.
NOTES: Harms noted that brachycalyx had weak affinities with
C. grandifolia Ducke (=C. javitensis var. grandifolia). Both species
have similar fruits with rufus trichomes, although the fruits of
C. javitensis are slightly swollen around the seeds, nearly flat to
weakly depressed between the. seeds, unlike C. brachycalyx. Both
species have long inflorescences. But javitensis is conspicuously
different with a long calyx tube (17-24 mm) and distinct lobes (4-8 mm),
large flowers (5.5-8 cm), and with leaves glabrous above and the
pubescence below moderate to sparse or lacking.
Clitoria brachycalyx has close affinities with dendrita from
which it is easily recognized by its cup-shaped calyx, white to pink
flowers, longer stamina! column, and shorter bracts and stipules. The
unique cup-shaped calyx of brachycalyx is found only in one other
species, C. brachystegia of Ecuador. Despite tlie similarity in cálices,
C. brachystegia is a more distant relative of C. brachycalyx and
C. dendrina. Clitoria brachystegia has larger flowers, narrower legumes,
longer ovaries, and shorter stipules.
DISTRIBUTION (Figure 33): Clitoria brachycalyx is an endemic of
westcentral Essequibo, Guyana, and adjacent Brazil. The species has
been collected predominately in the Kanuku Mountains of Guyana, with

332
the type collection representing the only Brazilian locality known.
The specimens collected were taken from dense forest to scrub, on sandy
to rocky ground or exposed rocky ledges, at elevations of 60-400 m.
0 U V AN A. ESSEQJIBO: Schomburgk 106.5 (K); Wabuwak Kanuku
Mts., Oct 43, Wilson & Bronne, Forest Dept. of( Br. Guiana 5774 (K,NY);
W extremity of Kanuku Mts. in drainage of Takuta River, 300 in, 4-22 Mar
1938, Smith 3128 (A,F,G,K,M0,NY,S,U,US,W); NW slopes Kanuku Mts,
drainage of Moku-Moku creek, Takutu territory, 150-400 m, 31 Mar-
16 Apr 1938, Smith 3393 (A,F,G,M0,NY,S,U,US,W); 5 mi S of Lethem at
mokomoko creek, 16 Apr 1956, Irwin 647 (US); Watu Ticaba, rec. 20 Jun 67,
Appur 2294 (K); Turuk Wav, 5 Dec 1957, Cook 243 (K); Annai, 200',
27 Dec 60, Graham 485 (k).
16. Clitoria dendrina Pittier, Cont. Nat. Herb. 20(3_)p 126. 1918.
Yfleurocarpuin niacrophyllum FI.B.K., Nov. Gen. Sp. 6: 410. 1824.
Clitori a tend!eri Rusby, nom. in sched.
Erect shrub to tree, usually 2-11 m tall, rarely to 25 m tall (fide
Espina A18) generally to 45 cm diameter, rarely to 100 cm diameter.
Branches to 8 mm diameter, pith hollow, juvenile branches angular, with
pubescence dense, tawny, appressed, branches becoming subterete,
glabrate with age, trichomes whitening, losing pigmentation; axillary
buds 2-3 mm, scales ovate, acute, pubescence appressed. Leaves
deciduous, 3-foliate, thick membranous, leaflets very large, generally
broad ovate to rhomboid-ovate, oval to orbicular, apex obtuse to
abruptly short acuminate, acumen to 0.5 cm or occasionally to 1 cm,
base rotund to retuse, subcordate on some large orbicular leaflets,

333
midrib weakly raised above, primary nerves of 8 10 pair, upper surface
pubescence minute, moderately dense, erect, similar to fine sandpaper
to the touch from apex to base, with age becoming confined to major
nerves, velutinous-tomentose, soft to the touch, becoming less dense,
pilose-like with age, lamina generally 10-20 (29) cm long, 8-18 (23) cm
wide. Petioles elongate, weakly angular to subterete, 8-16 (23) cm
long; pubescence minute velutinous; rachis 3-4.5 (5) cm. Petiolules
subquadrangular, 7-10 mm, densely velutinous pubescent. Stipules
caducous, rarely observed, lanceolate, acute, 6-8 mm long, 2 mm wide,
pubescence appressed; stipels caducous, linear to narrow lanceolate,
acute, 3-6 mm long, 0.2-0.8 mm wide, pubescence appressed. Inf1orescence
from denuded nodes, fascicled, few to several, crowded, short, racemose,
nodose, multiflowered, occasionally branched, pubescence of axes dense,
appressed, rufus, central axis 1-3 cm long. Pedicels 3-6 mm. Bracts
deciduous, minute, ovate, obtuse, pubescence appressed, rufus, middle
bracts 2-3 mm long, 1-1.5 nun wide. Bracteoles minute ovate, occasion¬
ally suborbicular, obtuse to broadly acute, 1.5-3 mm long, 1.5-3 mm
wide, inserted at calyx base, pubescence appressed. FIowers small,
3-4 cm, dull to dark purpie. Calyx pubescence dense, minute, appressed,
tube short, narrow, scarcely flaring at throat, 7-11 mm long, 3-5 mm
wide at base to 5-7 mm wide at throat, lobes minute, nearly equal,
broadly deltoid, acute, 2-3 (4) mm long, 2-3 mm wide, ventral lobe
3-4 mm x 1 mm. Vex i 11um pubescence dense, short, appressed, blade 1.8-
2.8 cm wide, claw 4-6 mm. Alae extended beyond the carina by 4-6 mm,
uncinate pubescent, blade 14-18 mm long, 3-6 mm wide, claw 10-13 mm.
Carina falcate, blade 7-9 mm long, 2-3.5 mm wide, claw 14-18 mm.
Stamens diadelphous, vexillary filament coherent below the middle, free

334
above and below, tube glabrous or occasionally more or less uncinate
pubescent toward apex, 16-22 mm, free filaments glabrous, 2-5 mm;
anthers lanceolate, apiculate, 1.5-2 mm long, 0.5 mm wide. Gynophore
short, 1-3 mm; ovary 9-10 mm long, 1.25 mm wide; pubescence white with
yellowish tinge; style dark-colored, 10-16 mm long, geniculate 5-6 mm
from distal end; stigma capitate, 0.5 mm diameter. Legume long-
stipitate, exerted well beyond calyx, flat, weakly depressed between
seeds, 16-20 cm long, 17-24 mm wide, pubescence minute, erect, dense,
rufus, of soft velvety texture; stipe pubescence dense, minute, erect,
arcuate to geniculate near apex, (18) 24-34 mm long, 2-4 mm wide
expanding to 5-7 mm at apex; beak to 7 mm; dehiscence causing valves to
twist one-half of a turn. Seeds lenticular, face suborbicular to
slightly longer than wide, 9-11 mm long, 8-11 mm wide, 2 mm thick,
7-8 seeds per pod, hilum elliptic, obtuse at both ends, 2 mm x 1 mm.
Figures 39 and 40.
Pittier's Arboreal Clitoria is characterized as a tree with large
pubescent leaves and very short, fascicled inflorescences from denuded
nodes bearing small purple flowers which have a short, narrow calyx
with minute lobes, or bearing rufo-pubescent fruits.
PHENOLOGY: This species blooms from (November) December through
February with fruits collected from December through March. Leaves are
deciduous at this time with a higher percentage of collections lacking
leaves in February and March.
TYPE COLLECTION: VENEZUELA. Aragua: Vegas de Rio Lemon cerca de
Maracay, 440 m, 27 Jan 1913, Pittier 5773 (H0L0TYPE: US 601479, not
seen. Isotypes: G-99! NY! VEN 4029!).

Figure 39. Cl i tori a deiidrina. (a-b) leaflets, x 1; (c-d) inflores¬
cences, x 1. (Wiii dack & Monachino 41006, US 2167572:
a,c. Wurdack & Monachino 40979, NY: b,d.)

336

Figure 40. Cl i tori a dendrina: (a) flower, x 1; (b)
IcJ vexillum, x 1; (d) ala and carina, x
x 1; (f) gynoecium, x 1; (g) branch apex
inflorescences, x 1; (h) fruit, x 1; (i)
seed, x 1. (Wurdack & Monachino 40979, [
1 1638, US 1801717 : g, i; F 1002908:“ h.)
calyx, x 1;
1; (e) androecium,
with persistent
three views of
!Y: a-f. Williams

1 cm

339
The New York and Geneva collections have the label data "La
Trinidad de Maracay, 440 m." These specimens may represent a collection
from a different site than the type locality of "Vegas de Rio Lemon
cerca de Maracay," although placed by Pittier under the same collection
number of 5773 as the type. If that is correct, then the New York and
Geneva specimens would not be isotypes because they would not be
duplicates of the type collection. Both collections are good
representatives of the species, although the New York collection lacks
fruit.
Stockholm has two photos of Kew specimens (Fendler 305), each
speciment collected from a different locality, but labeled as types of
C. dendrina Pitt. Pittier never included these Fendler specimens in
his original publication in 1918 nor in a later article in 1944.
Fendler 305 is not a type for the species, but may be the source for
the name selected by Rusby.
Specimens of Smith 925 were distributed from the New York
Botanical Garden and labeled "Cljtqria fendleri Rusby, sp. nov.," yet
the two New York specimens of this collection were labeled C. dendrina.
Cl itori a fendleri Rusby is a nomina in schedula as no publication of
this name has been found.
The name Neurocarpum niacrophyl 1 um H.B.K. has been placed in
synonymy on a questionable basis. These authors cited no type collec¬
tion, but they did cite the collection data as "Savaneta incolarum,
cresit prope Turbaco Novo-Granatensium, alt. 180 hex. floret Martio."
Stafleu (1967, p. 223) placed thetypes for newly described species by
these authors at Paris. A Paris specimen, Bonpland 141_2 (P-46!),
labeled isotype, appears to be a representative of the type collection.

340
It bt irs the libel data Mars 1801, Cartha ¡ene no. 1412 Turbaco (Melle
Grana la), Savei.ata, Bonj land" which is in type lata. Ties specimen consists of onl^ one counted leaf, lacking
any infloresce nce or flowers. A cotype s| cimen at the Field Museum
(F 93/281!) consists of i leaflet fragment (2 cm x 1 cm) taken from the
terminal leafl t mounted at the bottom of the I iris sheet. These leaf
specimens agre-- with C. dendrina.
U the to,- of the Imris sheet, there are two separately mounted
leaflets which are non-t iitori a! They are asymmetrical, different in
vein pattern, pubescence, and shape. They appo.ir similar to some
obser/ed in th genus Ed thrina. These leives lo not match the
published description of the leaves.
I he description of d. macrophyllum agrees . / it h C. dendrina rather
well xcept fo, two phas ,-s, (1) "red flowers" (Mowers are dull to dark
purple in C. d. odrina) and (2) "calyx camp, mul a i e-tubular" (calyx
narrowly infund ibular-timular in C_^ dendrina). However, some flowers
of C. dendrina have been described as maroon or other similar shades
indie 11ing a reddish comwonent. The term campa.iulate-tubular has been
used .o describe other Species of Clitorja with a similar short calyx.
Much .if the de.cription, however, could match m.ny species of Cl i tori a
because of the generalized structures described, but the closest fit
appears to be c. dendrj/u.
fhe histo./ of the name N. macrophyllum H.n.K. is short in the
literature and there is no indication that the name was transferred to
the g mus Cl it. ria . Aft -r the name was pui.lish d in 1824, de Candolle
(182b i treated t.h< species N. macrophyllum H.B. . as a doubtful species
under the genu Neurocar -urn. Don (1832) follow d de Candolle's

341
treatment by stating that perhaps this shrub should be removed from the
genus Neurocarpum. Bentnam (1837) listed N. macrophyl!urn H.B.K. as a
dubious species of Neurt arpum. Steudel (1841) listed N. macrophullum
H.B.K. under the genus lieurocarpum. Then the name disappeared from
literature. When Bentham (1858) combined the genus Neurocarpum with
the genus Clitoria, the name N. macrophyl1 urn H.B.K. was lacking in his
treatment.
If the name N. macrophyl1 urn is accepted as synonymous with
C. dondrina, then the Rule of Priority of the International Code of
Nomenclature would indicate that this species should bear the name
Criteria macrophylla (H.B.K.) Fantz. However, any transfer of the name
would create a homonym with Clitoria macrophyl1 a Wall (1828); thus the
name C. macrophylla (H.B.K.) Fantz for this species would be
illegitimate.
If additional collections of Bon pi and 1412 are located that include
the flowers and inflorescence, the questionable status of the synonymy
can be ascertained from their examination. Fortunately in either case,
whether this author was correct or in error in placing the name
N. macrophyllum H.B.K. in synonymy, the name of the species will remain
C^ dendrina.
VERNACULAR NAMES: VENEZUELA (ARAGUA): Sinvergüenza, Tamayo 1649.
(BOLIVER): Generala, Tamayo 2192; Oreja de tigre, Tamayo 3456,
Peonio de cerro, Marcano-Berti 2553. (CARABOBO): Ánima-flor de
anima, Hequert 838; Flor de ruima, Pittier 8669. COLOMBIA (MAGDALENA):
Guayacan Polvillo, Espina & Giacometto A18.
NOTES: CI i tor i a dendrina has close affinities with C^ brachycalyx
from which it is easily recognized by the purple flowers, narrow tubular

342
calyx, short, fasicled inflorescences, shorter staminal tube, and
longer bracts and stipules. Clitoria dendrina also has affinities with
C. froesii, a liana with a longer calyx tube, larger flowers, longer
ovaries and staminal tube, and larger anthers.
DISTRIBUTION (Figure 33): This species has a broad distribution
in Venezuela and northern Colombia, with one collection from central
Colombia. It inhabits dry forest in open areas, along forest borders,
and savannas at elevations of 50-900 m. Steyermark & Espinaza 108815
noted that the plants were located below the cloud forest belt. Cowan
& Wurdack 31538 noted the trees frequent on an igneous knob.
VENEZUELA. AMAZONAS: ca 3 mi down Cano Asisa from Paru
Savanna camp, 17 Feb 1951, Cowan & Wurdack 31 538 (NY,US,VEN). BOLIVAR:
sabanas de Las Majadas, Puruey, 18 Jan 1948, Tamayo 3456 (VEN-2 sh.);
cerro Negro Perón, on EW crystalline range on right bank of river, just
below El Carmen, ca 50 river km from mouth, Rio Paraguaza, 120-350 m,
27 Dec 1 955, Wurdack & Monachino 40979 (GH, NY,U, VEN); just below Raudal
Maraca, ca 110 river km from mouth, Rio Paraguaza, 115 m, 29 Dec 1955,
Wurdack & Monachino 41006 (F,NY,RB,US,VEN); Las Hi cotas, Feb 1 942,
Tamayo 2192 (VEN); 9 km E de Turiba, Ho. Cedeno, 6-11 Dec 1970,
Marcano-Berti 2553 (VEN); de la Prisión, Medico Caura, 120 m, 29 Mar
1 939, Williams 1 1638 (F,US,VEN). ARAGUA: Maracay, 1928, Vogel 620
(M-2 sh.); N of Maraquay, 25 Feb 1941, Tamayo lj54_9 (US); Parque Nacional,
10 Feb 1946, Pittier 1 5222 (VEN); Choroni , 800 m, Smith 3270 (VEN);
El Castaino, 400-600 m, 15 Mar 1940, Pittier 14278 (VEN). CARABOBO:
fe [sic] mi S of Quique, 3000', Fendler 305 (S!=photo of K); betv/een
Valencia & Yuma, 500 m, 24 Dec 38, Als_ton 5660 (BM,NY); vic Valencia,
400-800 m, 8 Dec 1919, Pittier 8669 (GH,NY); Entrada, N de Valencia,

343
Nov 1941, Huguert 833 ([¡0; l.c., 500 m, 20 Dec 1938, Alston 345 &
Willi inis 10972 (Note: distributed under either number or both; BM,NY,
S, VEN); l.c., 23 Dec 13)1, Warmings 106 (US); vie. Las Trincheras,
200-400 m, 30 Dec 1917, Pittier 7632 (GH.US.VEN); l.c., 1891-2,
Warmings 103 (US). COJEDES: along road from camp de Carabobo to San
Carlo, Feb 1931, Aristeguieta 1556 (VEN). LARA: between Paso de
Angostura & Parque Nacional Yacambu, ca 7.5 km from Paso de Angstura,
900 m, 29 Dec 1973, Steyermark & Espinaza 108815 (FLAS-2 sh.!,VEN);
El Aliar, 4 Dec 1967, Smith VI288 (VEN). PORTUGUESE: vis Aqua Blanca,
5 Feb 48, Curran 267M (NY); road Acariqua-Ospino, Feb 1 969, Aristeguieta
6 Zabila 6924 (VEN). BAKINAS: road Guanore-Barinas, Feb 1 953,
Aristeguieta 1561 (VEN; Karina, 12 Feb 1957, Curran 28 (NY). MERIDA:
Tovar, 1856-7, Fend!er 305 (S! = photo of K).
C 0 L 0 M BJ A. CESAR: Mirrea road, 1200', 7 Jan 1899, Smith
925 (type of synonym C_. fend!eri Rushy; flower: A,BM,E,F,G-2 sh.,GH,
MICH,NY,PH,S,U); Masiga Viega, 800', March 1899, Smith 925 (type of
synonym C. fend!eri Rusby; fruit: NY). MAGDALENA: Poponte, Magdalena
Valley, 150', 3 Jan 25, Allen 916 (K,M0); Minea region, 500-600 m,
Espina & Giacometto A18 (F,NY). CORDOBA: Sahagun, 150-200 m, 27 Jan
1918, Pennell 40816 (NY). META: La Macarena, Rio Guayabero, sabannas
de Arenisca, 215-700 m, i’eb-Mar 1 959, Barr inga & Me.jia 17034 (NY) and
17141 (NY).
W E S T i N p I £ I. 1786-91, von Rohr 62 (BM: locality unknown,
cultivar?).

344
17. Clituria froes i i Fantz, sp. nov.
Mana. Branches 3-) mm diameter, pit.li solid becoming hollow on
large diameter branches, juvenile branches longitudinally striated,
angular-terete, pubescence densely uncinate underneath erect to
appressed macro-trichomes, branches becoming uncinate pubescent only,
then glabrate; axillary buds 5 mm. ieaves 3-foliate, thick membranous,
conspicuously pubescent, leaflets slightly asymmetrical, broad-ovate to
broadly.oval or broadly oblong, apex acuminate, acumen to 1.5 cm, base
rotund, midrib and some of the primary nerves near the base weakly
raised above, midrib densely pubescent, primary nerves of 9-11 pairs,
upper surface dark green, pubescence minute, hirsute, texture similar
to fine sandpaper, lower surface light green, pilose pubescent, nerves
densely pilose-hirsute, lamina 10-25 5 cm long, 7-16.5 cm wide.
Petioles elongated, 13-21 cm, subterote, pubescence dense, erect;
rachis 4-6.5 cm. Petiolules 5-7 mm, quadrangular, pubescence dense,
erect, tawny. Stipules caducous, lanceolate, acute, 4-7 mm long x 2 mm
wide on juvenile branches and densely pubescent, to 10 mm x 3 mm when
persistent, and glabrate; stipels caducous, linear, acute, weakly
3-nerved, 3-6 mm long, 0.5-1 mm wide, pubescence uncinate, ciliolate.
Inflorescence axillary and terminal, racemose-nodose, axes solitary,
occasionally branched ne.tr base, few-flowered, pubescence dense, minute,
erect, 1.5-3 cm (observed mostly with buds). Pedicels 6-7 mm. Bracts
minute, ovate, acute, densely pubescent, 2-3 mm long, 2 mm wide.
Bracteoles minute, ovate, acute, den ely pubescent, 3-4 mm long, 2 mm
wide, inserted at calyx base. Flowers small, 3.5-4 cm, color unknown.
Calyx pubescence dense, ippressed, tube short, 12-13 mm long, 4-6 mm

345
wide at base to 7-9 mm at throat, i ibes minute, deltoid, acute, (3)
4 mm long, 2 mm wide, ventral lobe s jbequaI, 11.5 mm wide. Vex i 11 uni
pubescence short, sericeous, blade c i 2.5 cm wide, claw 4-6 mm. Alae
extended beyond carina ly 4-6 mm, blade 18-21 mm long, 4-8 min wide,
claw short, 7-9 mm. Caí ¡na falcate, 9-13 mm long, 4 mm wide, claw
12-16 mm. Stamina 1 tul e diadelphous, vexillary stamen coherent below
to near its middle, column clabrous. 24-28 mm long, free filaments
2-4 mm; anthers lanceolate, 1 ong-apiculate, 2.4-3 mm long, 0.5-0.8 mm
wide. Gynophore 2-3 mm; ovary 10-11 mm long, 1.2-1.3 mm wide,
pubescence dense, white ; style 14-15 mm, geniculate 6 mm from distal
end; stigma capitate, 0.5 mm diameter. Legume unknown. Figure 41.
Froes' Cl itoria is characterize I as a liaría with small flowers that
have short calyx tubes and minute lobes.
PHENOLOGY: The only known collection of flowers was made in
January.
TYPE COLLECTION: BRAZIL. Amazonas: terra firma, Macubeta on
Rio Maria, basin Rio Negro, 31 Jan 1942, Froes 12441/185 (H0L0TYPE: NY,
Krukoff Herb, placed in Britton's Herb. Isotype: A).
Of the two known collections, the type collection has flowers
whereas the paratype is vegetative. The New York collection was
selected as the holotype because it had inflorescences and a dissected
flower in a paper towel packet (placed in cellophane envelope by this
author). The Arnold Arboretum specimen had fewer flowers (one
dissected) and lacked inflorescences. The New York specimen has
immature leaves whereas the isotype lias the larger leaves.
VERNACULAR NAMES: BRAZIL (AMAZONAS): Darume, Froes 12441/185 &
12431/175 .

Figure 41. C Vitoria froes/i. (a) mature leaflet, x 1; (b) juvenile
leaflet appearing with inflorescence, x I; (c) juvenile
inflorescence, x I; (d) flowei , x 1; (e) calyx with
persistent stamina 1 tube, x 1 (f) calyx and bracteole,
x 1; (g) vexillum x 1; (h) ala and carina, x 1;
(i) androecium, x 1; (j) anther, x 4; (k) gynoecium,
x 1. (Froes 1244 I/]85, A: a,d,f,h,j; NY: b-c ,e ,g, i, k.)

347
1 cm

348
ECONOMIC IMPORTANCE: Darume vine is considered the most toxic
plant in the region of the Rio lean. It is planted by the Indians and
found in fair quantities, according o Frees 12441/185. Froes does not
indicate for what purpose the toxin is used by the Indians, but this
author assumes that the toxin is use! as a fish poison, similar to
C. arborescens.
NOTES: The Darume vine does not: fit in the known species. Its
nearest relative is probably C. dendrina which differs by its arboreal
habit, smaller calyx, pubescence on che lower leaf surface, smaller
anthers, smaller pedicels, more crowded inflorescences, and general
leaf shape. Although the Darume vine is the only liana in the section
Bractiycalyx, which is characteriStic .1ly comprised of tall erect shrubs
to trees, the short calyx and minute lobes plus the pubescence on the
upper leaf surface suggest that this species belongs more naturally in
section Brachycalyx than in section haul if1orae, which is character¬
istically comprised of lianas.
DISTRIBUTION Figure 33): This .pecies has been poorly collected
and is known only from the Rio Negro basin of Brazil, where it is
cultivated by the Indians.
B R A Z I L. AMAZONAS: Pirapu. a on Rio Negro, 1941, Froes 12431/
175 (Paratype: S, Erik Wall Herb.).

349
Section Cauliflorae
ID. Clitoria L. subgenus Bractearia (Mart, ex Benth.) Fantz
section Cauliflorae Fantz, sect. nov.
Lianas or occasionally an erect shrub. Inflorescence typically
cauliflorous and axillary, racemose, more or less nodose, Bracteoles
variable in size, but not large and hiding the calyx, typically shorter
than the calyx tube and narrow, 1-3 (6) mm wide. Calyx tube bearing
conspicuous lobes, typically 4-13 (27) mm long. Legume flat to weakly
depressed between the seeds, pubescence typically uncinate with sparse
spreading to suberect trichomes, occasionally densely appressed with
uncinate trichomes lacking. Seeds usually suborbicular to slightly
longer than wide or slightly wider than long, rarely oblong, usually
thickened, 3-5 mm thick, occasionally compressed.
The members of the section Cauliflorae consist of a heterogenous
group of species that do not fit in the sections Bractearia, Flexuosa,
or Brachycalyx. As a result, they are more difficult to characterize.
However, members of section Cauliflorae can be recognized easily by the
climbing habit, the cauliflorous inflorescence (when produced),
thickened seeds, and the lack of the key characteristies of the other
three sections. This includes the lack of the conspicuously large
bracteoles which hide the calyx (characteristic of sections Bractearia
and Flexuosa) and the minute calyx lobes of 1-4 mm and sublomentaceous
fruits (characteristic of section Brachycalyx). A liana with narrow
bracteoles (1-6 mm wide) and conspicuous calyx teeth plus cauliflorous
inflorescences should indicate a member of the section Cauliflorae.
HOLOTYPIC SPECIES: C. sagotii Fantz

350
NOTES: With a heterogenous assemblage of species, the choice of
a type species was difficult. Clitoria sagotii has more of the common
characteristics (i.e. those characteristics found in a large number of
the species) as discussed below. Clitoria javitensis was considered as
the type because of its large degree of variability in the species, but
this species has some characteristics (i.e. fruits) that suggest
affinities with members of section Brachycalyx.
Two species, both endemics, stand out immediatey as distinct and
representing a divergent evolutionary line. Clitoria obidensis and
C. plumosa are both conspicuously hairy with long shaggy hairs, often
rufescent, occurring on the major axes of the plant, the calyx, and
both leaf surfaces. In addition, the calyx lobes and bracteoles are
extremely long compared to other members of this section. The anthers
are also large (2-3 mm x 0.7-1 mm), which is found only in one other
species, C^ pendens. The similarity of the fruits of C. obidensis
(absent in C^ plumosa), the calyx tube, and the androecium and gynoecium,
plus the cauliflorous inflorescences, indicate that it is more natural
to include them within the section Cauliflorae than in a section of
their own.
Clitoria coriácea is distinct with its ceriferous leaves, oblong
seeds in which the length is much longer than the width, and its shorter
fruits. The species has a shrub habit, apparently non-climbing. The
inflorescences are borne at defoliated nodes as in section Brachycalyx.
Yet this species has the conspicuous calyx lobes (7-12 mm), non-
sublomentaceous fruits, and typical uncinate pubescence of the calyx
and fruits, as found in most members of section Cauliflorae. Despite

351
the aberrant characteristics, this species appears to fit better in
section Cauliflorae than in section Brachycalyx, or in a section of its
own.
The remaining species are very similar in many ways, so much so
that historically, they have often been misidentified. Three names are
commonly used for this group of species. First, any liana with an
elongated inflorescence was given the name C. 1eptostachya. This name
has also been applied in some instances to specimens with longer
inflorescences and the distinctly large bracteoles which hide the
calyx. Most of the time it has been applied to plants from Guyana,
Surinam, and adjacent areas. Two distinct species are involved.
Second, a name commonly used for lianas was arborescens, the earliest
name applied for any woody climber that lacked the large primitive
bracteoles. Thus C. arborescens has been misapplied to many early
collections of woody vines. A third name applied was C. javitensis.
This name became more popular for lianas when some botanists pointed out
that arborescens had tomentose pubescence on the lower leaf surface,
whereas javitensis had appressed hairs. A number of specimens of
the eastern section of northern South America possessed glabrous
leaves. These were included as a separate variety of C. javitensis.
This broad concept of C. javitensis has been abandoned in this treatment.
Several distinct species are involved.
Variation of several characters plus misidentified specimens has
accounted for much of the nomenclatura! confusion. A synopsis of the
variable traits is useful. Within this group, inconspicuous uncinate
trichomes are typically found on the fruits, cálices, ovaries, and often
on the axes and branchlets of the inflorescence. Specimens matching the

352
type of C. javitensis lack this anatomical character. The macro-
trochomes of C. javitensis are appressed on the calyx, lower leaf
surface, and are densely exhibited on the fruits. The glabrous "variety"
typically lacked these trichomes or possessed scattered, subappressed
to suberect trichomes. Inflorescences showed much variation. Some
specimens have inflorescences which are conspicuously short, subsessile
to 0.5 cm long, and are few-flowered. Others had longer inflorescences
of 1-6 cm, or occasionally longer to ca 16 cm and bore several flowers.
Bracteole length was very variable, but they are usually narrow. Most
bracteoles are much shorter than the calyx, although longer bracteoles
occurred. Calyx tubes and lobes are very variable in length, with most
members exhibiting tubes of 10-24 mm long and lobes of 4-11 (13) mm in
length. Ovary sizes are quite variable, from 10-22 mm. The sericeous
hairs are white, white with a yellow tinge, or tawny-rufescent; or they
are often lacking and then the pubescence is conspicuously densely
uncinate. Staminal tubes range from 24-49 mm long. The midrib of
the upper surface of the leaves is weakly to strongly raised in some
members. Flower sizes and floral parts are variable, especially in the
length of the claws, which is a good indicator character.
The key to the species of section Cauliflorae is constructed such
that those species which have been typically included as one species or
those species which have been frequently confused with each other, are
placed adjacent to each other. Couplet two provides less contrast than
desired, but the calyx pubescence is a reliable clue to a similar
pubescence on the fruit and ovary. Clitoria cavalcantei is an exception
by having an appressed pubescent calyx which lacks the uncinate
trichomes but which has the uncinate trichomes conspicuously observed

353
on the ovary and legume. Inflorescence length may be questionable, but
C. sel 1 oi is geographically isolated from the combined distribution of
the other species. Cl itoria selloi is found near the eastern coast of
Brazil, in the state of Bahia south of Rio de Janeiro. Other species
are distributed in the central and western part of the Amazon Basin
northward. Legume length and seed thickness appear very reliable for
those species in which the fruits are known, but often collections are
made which lack the fruits. Style length compared to ovary length proved
reliable, but this often required some dissection of the flower. By
providing a number of characteristics and noting the exceptions that
occur, this key was superior for distinguishing different types of
collections than a more artificial key of one to a few characters.
KEY TO THE SPECIES:
1.Bracteoles short, 2-15 mm; leaves glabrous above (excluding
occasional pubescence on midrib); calyx lobes 4-13 mm; calyx tube
10-24 mm.
2.Style length subequaling the ovary length; legume long, 16-23
cm, and broader, 18-25 mm wide; seeds compressed, 2-3 mm thick;
calyx pubescence (vidi 20 x) conspicuously appressed, moderate
to dense, uncinate hairs hidden (C. arborescens) to lacking;
inflorescence typically medium-sized, 1-15 (20) cm.
3.Bracteoles 10-15 mm long; calyx tube 12-17 mm; leaf
pubescence tomentose below; ovary 15-18 mm; style 16-19 mm;
stamina 1 tube 28-35 mm; flowers 4-6 cm; vexillum claw
8-12 mm; stipe 14-20 mm; legume pubescence uncinate plus a
few scattered, suberect trichomes (northern South America,
Guyana to Colombia) 18. C. arborescens

354
3. Bracteoles 2-7 mm; calyx tube 17-24 mm; leaf pubescence
below appressed to glabrate to glabrous, infreguently
pilose; ovary 18-22 mm; style 17-25 mm; staminal tube
36-49 mm; flowers (5.5) 6-8 cm; vexillum claw 15-19 mm;
stipe 24-33 mm; legume pubescence dense, rufus, appressed
to suberect (northwestern South America and Central America,
Venezuela to Peru plus mainly Panama). .19. javitensis
2. Style much longer than the ovary, typically 10 mm or more (only
3-6 mm longer in C. selloj^ and 1 eptostachya); legume smaller,
(6) 8-16 (18) cm long, 10-18 (20) mm wide; seeds thickened,
3-5 mm thick; calyx pubescence (vidi 20x) conspicuously uncinate
plus some subappressed to suberect hairs, moderate to scattered
(appressed with uncinate hairs lacking in C. cavalcantei);
inflorescence very short (subsessile to 1.5 cm) or very
elongated, 10-90 cm (medium-sized, 2-15 cm in C. selloi).
4. Flowers 4-8 cm; staminal tube 32-49 mm; carina blade
11-20 mm long, claw 23-36 mm; style 17-33 mm, much exceeding
ovary (except C. 1eptostachya); inflorescence very short
(subsessile to 1.5 cm) or very elongate (10-90 cm); plant
of Amazon Basin northward.
5.Bracteoles elongate, 5-15 mm long, width 1-3 (5) mm;
inflorescence very short, subsessile to 1.5 cm, few-
flowered.
6.Calyx pubescence appressed, uncinate hairs lacking;
style 28-32 mm; vexillum claw 14-18 mm; inflorescence
0.5-1.5 cm; calyx tube 16-20 mm, lobes 5-7 mm (Para,
Brazil) 20. C. cavalcantei

355
6.Calyx pubescence uncinate with scattered subappressed
to suberect hairs; style 21-27 mm; vex i 11um claw
6-10 mm; inflorescence subsessile, to 0.5 cm; calyx
tube 11-17 (18) mm, lobes (5-7) 7-13 mm (those with
shorter lobes have tubes 11-15 nun; those whose tubes
14-18 mm have lobes 7-13 mm).
7.Stipules 3-6 mm; bracts 2-5 mm; bracteoles 1-3 mm
wide; staminal tube 32-39 mm.
8.Leaves pubescent below, margin weakly to
strongly revolute, primary nerves of 6-8
pair; leaflets smaller, 2-5 (6) cm wide,
5-10 (13) cm long; flowers 4-6 cm; stipe
16-20 mm.
9.Bracteoles 5-8 (11) mm; vexillum claw
9-10 mm; staminal tube 32-39 mm; leaf
appressed pubescent below, ceriferous,
margin strongly revolute, midrib
impressed above; shrub (Venezuela) . .
21 . C^_ coriácea
9. Bracteoles 13-14 mm; vexillum claw 6 mm;
staminal tube 28 mm; leaf tomentose
below, non-ceriferous, margin weakly
revolute, midrib weakly raised above
(Brazil) .... 22. C. tunuhiensis
8. Leaves glabrous below, margin non-revolute,
primary nerves of 8-12 pair; leaflets large,
5-9 cm wide, 8-20 cm long; flowers (5)

356
6-8 cm; stipe 27-35 mm (French Guiana to
Venezuela plus adjacent Brazil) ....
23. sagotii
7. Stipules 8-19 mm; bracts 5-8 mm; bracteoles
3-5 mm wide; staminal tube 39-43 mm (Guyana)
24. kaieteurensis
5. Bracteoles short, 1.5-4 mm long, 0.8-1.2 mm wide;
inflorescence very elongated, 10-90 cm, multi flowered.
10. Flowers 6-8 cm; staminal tube 40-48 mm; ovary
18-20 mm; style 27-33 mm; bracteoles 3-5 mm; calyx
tube broad, 9-15 mm wide at throat, lobes 6-9 mm
Suriname, eastern Guyana) ... 25. C. pendens
10. Flowers 4-6 cm; staminal tube 32-37 mm; ovary
12-15 mm; style 17-19 mm; bracteoles 2 (3) mm;
calyx tube narrow, 7-10 mm wide at throat, lobes
4-6 mm (Guyana, Brazil) ... 26. C. leptostachya
4. Flowers small, 2-4 (4.5) cm; staminal tube 24-28 mm;
carina blade 8-10 mm, claw 18-20 mm; style 13-16 mm,
slightly exceeding the ovary length; inflorescence medium-
length, 3-15 cm; plant of eastern Brazil (Bahia to Rio de
Janeiro) 27. C. sel 1 oi
Bracteoles extremely long, 20-35 mm; leaves densely pubescent above;
calyx lobes 13-27 mm; calyx tube 20-30 mm.
11. Stipules 6-11 mm; inflorescence short, to 1 cm; bracts 6-11 mm;
bracteoles 4-6 mm wide; calyx lobes 20-27 mm, subequaling tube
length of 20-27 mm; ovary 13-17 mm; ala claw 18-20 mm; carina
claw 31-33 mm (Brazil) 28. C. obidensis

357
11. Stipules 19-29 mm; inflorescence elongated, 10-15 cm; bracts
13-16 mm; bracteoles 3-4 mm wide; calyx lobes 13-18 mm, nearly
half the tube length of 27-30 mm; ovary 21-22 mm; alae claw
23-25 mm; carina claw 34-38 mm (Colombia) . . 29. C. plumosa
DISTRIBUTION (Figure 10): The members of section Caul if1orae are
distributed mainly in the central and western Amazon Basin northward,
and in Panama. One species, C. selloi, is found outside this range,
along the eastern coastal forests of Brazil, from Bahia to Rio de
Janeiro-. A second species, C^ javitensis, has been collected
infrequently in countries of Central America north of Panama.
18. Clitoria arborescens R. Brown, in Aitón Hort. Kew, ed. 2,
4: 302. 1814. (=C. arborescens Ait., l.c., err, cit.
£ro )
Clitoria poitaei DC., Prod. 2: 234. 1825. (C. poiteaui DC.,
orthogr. pro ; C. poitaei Dec. , Steudel Norn. Bot. ed. 2,
1 : 386. 1841 , err. cit. pro; C^ poiteai Benth., Ann.
Wein. Mus. Nat. 2: 115. 1837, err, cit. pro ).
Cl itoria amoena Miguel, Nat. Verh. Moll. Mat. Wet. Haarl.
7: 24. 1851.
Ternatea arborescens (Ait.) Kuntze, Riv. Gen. PI. 1_: 210. 1891.
Scandent scrub, typically tall climbing liana, or an erect shrub
with climbing branches, or infrequently an erect shrub to 5 m tall and
8 cm diameter. Braches 2-8 mm thick, pith hollow, juvenile branches
compressed-quadrangular, becoming nearly terete, longitudinally striate,
pubescence of uncinate and rufus, densely appressed hairs, trichomes

358
becoming whitish, the appressed hairs becoming less dense and the
uncinate hairs more conspicuous; bark grayish-brown, splitting in
longitudinal strips; axillary buds 3-4 mm. Leaves 3-foliate,
subcoriaceous, leaflets basically elliptic, becoming oblong-elliptic,
obovate, oval to suborbicular, apex broadly acute to obtuse, more or
less mucronate, base broad cuneate to rotund, rarely retuse, margin
thick-veined, midrib impressed above, primary nerves of 9-14 pair,
upper surface dark green, glabrous, juvenile leaflets with minute hairs
along midrib, lower surface green, tomentose, trichomes rufus, whitened
with age, less dense, lamina (5) 8-15 (21) cm long, 3-9 (13) cm wide.
Petiole compressed-quadrangular to subterete, weakly caniculate
adaxially, weakly twisting, 4-15 (19.5) cm; pubescence of conspicuous,
rufus, appressed to suberect hairs with scattered, inconspicuous
uncinate trichomes; rachis 1-2.5 (3) cm. Petiolules subquadrangular,
4-9 mm, pubescence dense, rufus, spreading. Stipules deltoid to
deltoid-lanceolate, semipersistent, acute, 6-9 mm long, 3-6 mm wide at
base, pubescence uncinate and appressed, rufus; stipels linear, acute,
pubescence as stipules, 4-8 mm long, 1-2 mm wide. Inflorescences
axillary, racemose, nodose, multiflowered, crowded in bud, lower flowers
aborting as axis elongates, axes subquadrangular, laterally compressed,
5-16 (22) cm long, 2-6 mm thick, pubescence of uncinate and densely
appressed, rufus hairs. Pedicels 4-9 mm, becoming 3-4 mm thick in
fruit. Peduncle 1.5-4.5 cm, straight; rachis very weakly flexuous,
internodes 4-10 (15) mm. Bracts persistent, often spreading to reflexed
with age after flower aborts, lanceolate, acuminate, 5-9 mm long,
1.5-2 (3) nun wide, pubescence uncinate and appressed, rufus, appressed,
outer bract slightly shorter than middle pair, inner bracts caducous.

359
Bracteoles lanceolate to linear, acuminate, often spreading, more or
less slightly twisted, almost subequal the calyx tube, (8) 10-15 mm
long, 2-3 (4) mm wide, inserted 1.5-2 min below the calyx base,
pubescence uncinate and appressed, rufus. Flowers medium-sized, 4-5.5
(6) cm, bicolored, blue to violet and/or white; vexillum with a white
to pale violet center, reddish to purplish veined, blue to violet toward
margins; ala and carina white or white with violet tinge, veins light
purple. Calyx yellowish-green, often longitudinally furrowed, veins
inconspicuous, pubescence inconspicuously uncinate beneath moderate to
densely, appressed, rufus trichomes, tube 12-17 mm long, 4-7 mm wide at
base to 7-11 mm wide at the throat, lobes deltoid, acuminate-subulate,
5-8 mm long, 2-3 mm wide at base, ventral lobe linear to subulate,
6-11 mm. Vexillum pubescence moderate to dense, rufus, appressed,
blade 2.5-3.5 cm wide, claw 8-12 mm. Alae extended beyond the carina
5-8 mm, blade 18-25 mm long, 5-10 mm wide, with scattered uncinate
pubescence, claw 14-18 mm. Carina falcate, blade 9-13 mm long, 4-5 mm
wide, claw 23-27 mm. Stamina! tube with uncinate pubescence near
apex, 28-35 mm, free filaments 2-4 mm long, with moderate to densely
uncinate pubescence; anthers lanceolate, 1.5-2 mm long, 0.6-0.8 mm
wide. Gynophore 1-3 mm; pubescence densely appressed, rufus; ovary
15-18 mm long, 1-1.5 mm wide, pubescence rufus, sericeous; style 16-19
mm, geniculate 5-7 mm from distal end; stigma subcapitate, ca 0.5 mm
diameter. Legume short-stipitate, shortly exceeding to slightly
included within the calyx, brown, flat, pubescence of uncinate with
scattered suberect trichomes, valves 16-23 cm long (rarely with matured
fruit 5.5-13 cm), 18-26 mm wide; stipe brown, 2-3 mm thick expanding to
near 6 mm at apex, 14-20 mm long, pubescence uncinate with sparse

360
scattered trichomes; beak when present, to 6 mm; dehiscence causing
valve to twist one-half of a turn. Seeds black, smooth, lenticular,
face suborbicular to slightly longer than wide, base truncate to
slightly oblique, 11-13 mm long, 10-13 mm wide, typically 7-11 seeds per
pod; hi 1 urn oblong, 2 mm x 1 mm. Figure 42.
The Arborescent Cljtorja^ is characterized as a scandent shrub with
elongate inflorescences bearing medium-sized flowers subtended by
narrow bracteoles that subequal the calyx tube, or bearing long fruits
with an' uncinate pubescence, and with leaves that are tomentose below.
The tomentose 1 eaves and 1ong bracteoles are good indicators of this
species.
PHENOLOGY: Flowers are generally collected from July to early
December, although single collections were made in January, February,
and May. In Venezuela, a few collections were made in April. Fruits
have been collected from July to December, with most maturing from
September to November. Despite the similarity in flowering and fruiting
season, individual collections usually bear only fruits or the flowers.
TYPE COLLECTION: Cult, from Trinidad, 1804, James Mere s.n.
(LECTOTYPE: BM).
This species generally goes by the name C. arborescens Ait, in the
literature. Stafleau (1967, p. 4) reported that Aitón was the editor
of Hortus Kewensis, edition two, and that most of the new descriptions
were written by other contributing authors. Stafleau credits R. Brown
for classes XIII to Cryptogamie, which included class XVII (Diadelphia)
where the genus Clitoria was placed. Therefore, the correct citation
for the species should be C. arborescens R. Brown in Ait.

Figure 42. Clitoria arborescens. (a) leaflet, x 1; (b) inflorescence,
x 1, (cT flower, x 1; (d) calyx, x 1, (e) vexillum, x 1;
(f) ala and carina, x 1; (g) androecium, x 1; (h) gynoecium,
x 1 ; (i) legume x 1; (j) three views of seed, x 1.
(Melinon 51, NY: a,i. Melinon 347, BM: b. Perrottet s.n.,
G-279-80: c-h. Williams 1 2557", FI 034942: jTJ

362
1 cm
a

363
R. Brown cited type specimens when he published the name C.
arborescens. The type syntypes were "Alexander Anderson and James
Vere (cult. 1804)." The scant data given for the types plus a
description that is very generalized, such that it could fit most
members of subgenus Bractearia, make it difficult to characterize
Brown's species. However, this species has been cultivated, and more
information can be obtained from later publications. In 1832, the
"Curtis Botanical Magazine" published a description of the species along
with an'illustration. However, problems of identification persist.
John Curtis made an illustration of the flowering specimen in 1822. Ten
years later, S. Curtis reported that he added a leaf and the dissected
flower parts. The leaf appears to represent C^ javitensis based upon
its long acuminate apex, wider-spaced and fewer primary nerves, along
with the leaf description of the pubescence agreeing with javitensis,
not C^ arborescens. The stamens were noted as glabrous, whereas in
C. arborescens, the filaments are densely uncinate pubescent. The rest
of the description is so general that it could fit either species,
although the flower color agrees with £. arborescens more than
C. javitensis.
In 1825, de Candolle published a detailed description of a new
species, C. poitaei (Type: Poiteau s.n., Guiana Gallica). In 1851,
Miquel published C. amoena (Type: Kappler 1933). In his synopsis of
the genus Clitoria, Bentham (1858) described C. arborescens Ait, and
included the names C^_ poitaei DC. and C^ amoena Miq. in synonymy.
Bentham cited several collections, including "Anderson (Br. Guiana),
Hostmann 50 & 1 097 , Kappler 1 933 , Perrottet (Fr. Guiana), Schoinburgk 849,
Rich. Schoinburgk 1331 , Lockhart, & Guilding (St. Vincent)." This author

364
has seen specimens representing all these collections, except the last
two. They are all of the same species.
Of the two syntypes, the Anderson collection appears to be the
better selection for the lectotype, since it was cited by both Brown
and Bentham. But it is uncertain that they examine the same Anderson
collection. Brown noted the plant from Trinidad. The specimen in the
Bentham herbarium (K-417!) bears the label data "frutex elegans, flore
odorato, St. Vincent's, Anderson" on the printed label titled "hb.
Forsyth'. 1835." The term "St. Vincent's" was crossed out by a line
drawn through it and "Guiana" inked in above it. This may or may not
be the specimen, or a duplicate of the specimen, that Brown referred to
in his publication.
The Mere collection was not cited by Bentham, but it is a more
probable syntype or isosyntype. This specimen is placed in a type
folder marked "C. arbórea" and bears the data "horti, from Ins.
Trinitatis, Dr. Mere" (BM!). At the bottom of the sheet, in faint
pencil lead, are the words "Clitoria arbores Hort. Kew no. 2." This
specimen is not the type of C. arbórea Benth. (1837) whose name was
not published in the "Hortus Kewensis." Nor has the Mere specimen ever
been cited for that species. The only species published in "Hortus
Kewensis" was C. arborescens. The incomplete identification on the
Mere collection misled the individual who placed the sheet in an
individual folder and marked it "C. arbórea." This Mere specimen is
probably from an individual plant growing in his garden, and the plant
referred to by Curtis (1832) who indicated that a cultivated stock of
the species (Curtis was discussing C. arborescens) was introduced to
Mr. Vere's garden from Trinidad in the year 1804. Brown probably

365
examined this specimen or .it least based his description in part on the
plant in Dr. Vere's garden which this specimen represents. Since the
Mere collection was the more probable type collection of the two
possible syntype collections, it is selected as the lectotype.
The lectotype is not as representative a collection as the Anderson
or type collections of the synonyms. The leaf pubescence is reduced and
is located mostly along the veins, the pubescence common for the older
leaves. The bracteoles are slightly shorter than the typical range of
variation (10-15 mm).
The name C. poitaei was published by de Candolle (1825) based upon
a collection of Poiteau from French Guiana. The collection labeled
"Cayenne, Guyane, Jul 1824, Poiteau s.n. , hb. Gay" (GH! K-30!) is the
probable type. These specimens agree with the Anderson and Vere
collections. Bentham was correct in synonymizing the name C. poitaei
DC. with C_. arborescens R. Brown in Ait. A. P. de Candolle based his
specific epithet on Poiteau's name, but altered the spelling
(intentionally?) to C. poitaei. A. P. de Candolle made corrections in
his "Addenda et Corrigenda," but he did not correct C. poitaei to
C. poiteaui.
Miquel published the name C^ amoena in 1851 citing one collection,
Kappler 1933 (G-281! M 14681! S! U 37642A! W 88619!). Only the Utrecht
specimen was identified as C. amoena, whereas all others were identified
as C, poitaej DC. The Utrecht specimen is the more probable type and
the other specimens are duplicates of the type.
One additional nomenclatura! note. Bentham (1858) cited two
additional collections fot C_. arborescens in addition to those listed
above. Cuming 1142 and Sinclair, both of Panama, represent

366
misidentifications of a pilose form of C. javitensis. Cl i tor i a
arborescens has been applied historically to two Panamanian species,
one a tree (=C. glaberrima) and the other a liana (C. javitensis).
Clitoria arborescens may have been applied to the Panamanian species
because of the early misidentification of these two collections. The
names for these Panamanian specimens were recently corrected by Croat
(1 974).
VERNACULAR NAMES: SURINAME: Haiaribal 1i (Arawak), Archer 2854.
(MAROWIJNE): Kroemenoeroe (Car.), Lanjouw & Linderman 990. (NICKERIE):
Koeroemoe enoelae (Kar.), Stinkvogel-oogen, Kav-ai, (N.E.), Gonqqryp
2179. (SARAMACCA): Kourai (bigie nonloe), Pulle 222.
ECONOMIC IMPORTANCE: This species has been cultivated in the past,
introduced into Trinidad, St. Vincent, and botanical gardens in England.
Archer 2547 (Guyana) noted the plant was used as a fish poison.
NOTES: Clitoria arborescens has close affinities with C^ javitensis.
Both species have larger fruits, appressed pubescent cálices, and styles
subequaling the ovary length, unlike other species which have smaller
fruits, uncinate pubescent cálices, and styles longer than the ovaries.
Clitoria arborescens can be distinguished easily from C^ javitensis by
the presence of uncinate hairs on the calyx and legume, smaller flowers,
and shorter reproductive structures (ovary, style, staminal tube), and
conspicuously longer bracteoles.
The leaves are variable within the species. Suborbicular leaflets
appear quite different from the elongated elliptic leaflets. But a
series of intermediates from the basic elliptic leaflet can be observed,
often within the same collection. Two collections, Maguire 24866 and

367
Jenman 361 were reported as trees, although otherwise they agree with
other collections labeled as lianas or vines.
DISTRIBUTION (Figure 43): This species is distributed in northern
South America, from French Guiana to eastern Venezuela, with isolated
collections from Colombia and northern Venezuela. The species is not
known from Panama nor other Central America countries, although it has
been reported there in the literature. The erroneous reports were based
upon misidentified collections. Plants are typically found in open
forest areas or along the forest margins of a river bank. This species
has been collected in gallery forests and reported from walaba forests,
mora forests, and palm scrubs. Soil is generally reported as lateritic
or sandy. Several Suriname collections note its occurrence with granite
or volcanic slabs. The species has been collected at elevations of
65-420 m, although most collectors did not record the altitude.
FRENCH GUIANA. 1820, Perrottet s.n. (G-2 sh.). GUYANE:
Cayenne, Anonymous s.n. (W 216971 ); l.c., Ju 1 1824, Poiteau s.n. (type
of synonym C. poitaei: GH, K-30); Mana, 1854, Melinon 51 (F,GH,NY)
and 154 0E); l.c., 1855, Sagot 754 (BM,BR,W); St. Laurent, Stat.
Crique Balatie, 22 Nov 1948, Bena 4338 (U,VEN); Maronia R. , 1861,
Mélinon 51_ (BM); 1 .c., 1876, Mel i non 347 (BM).
SURINAM. Anonymous s.n. (M 12774); Kappler 90 (S); 1842-3,
Hostmann 1097 (BM); Tigri-heddeval in Tanjimana R., Mennega 437 (U-2
sh.); Scotelweg, 2 & 12 Nov 1934, Archer 2854 (U). MAROWIJNE: ad flum.
Marowyne, Jul 1846, Kappler 1933 (type of synonym C. amoena: G-281,
M 14681, S,U 37642A,W 88619); Marowijne River near Armina Falls, 12 Aug
1933, Lanjouw 527 (U); fluv. Marowijne N of Albina, 18 Sep 1948,
Lanjouw & Underman 285 (NY,U); l.c., 1 Sep 1 948, Lanjouw & Linderman

Figure 43. South American distribution of four species of section
Caul iflorae. Clitoria cavalcantei C. coriácea (★)
C. arborescens (%); C. tunuhiensis ().

369

370
990 (NY,U); fluv. Marowyne prope Albina, Jul 1 904, Versteeq 532 (U);
fluv. Marowyne prope Galibi, 15 Oct 1901, Went 421 (U-2 sh.). SURINAME:
fluv. Suriname, 8 Jul 1900, Tresling 57 (U); Kabel station, 30 Aug 1914,
Stahel 41_8 (M0,U); Groningen station, 10 May 1916, Samuels 170 (NY).
SARAMACCA: Raleigh Falls, Coppename River alongside granite sheet of
Voltzberg, 4 Dec 1 956, Linderman 7753 (U,US); Coppename River near
Raleigh Falls, 10 Dec 1933, Lanjouw 735 (U); fluv. Coppename prope mont
Voltzberg, 23 Aug 1920, Pulle 286 (U); Coppename, Raleighvallen, 31 Jul
1923, Gonggryp & Stahel 6349 (U-2 sh.); l.c., Kelk road, 25 Sep 1954,
Mennega 142 (U); 8 km Saramacca R., 29 Sep 1944, Maguire 24866 (F,GH,
NY,RB-2 sh.,U,US); Saramacca R. between Posoegronoe & Grasi Falls,
11 Oct 1944, Maguire 24936 (F,NY,U-2 sh.,US); Paramaribo, 1910, Pulle
222 (U); l.c., Wul1schlagel 88 (W). NICKERIE: around Kabalebo airstrip,
9 Jan 1965, Florschutz & Mass 2669 (U-2 sh.); Fluv. Corantyn, Kaboeri-
kreek, 27 Jun 1916, Gonggryp 2179 (U); second hill, fluv. Corantijn,
7 Sep 1920, Pulle 467 (U); Corantijn, 1 dag stroompo v. rred.
Hendri-vallen, 31 Aug 1935, Rombouts 163 (U); Akwansa, f1. Nickerie,
25 Sep 1916, Stahel & Gonggryp 3562 (U).
GUYANA (BR. GUIANA). Anderson s.n. (probable syntype,
K-417); Jenman 2157 (NY-2 sh.); 1837, D. Schomburgk 2:82 (BM); 1837,
R- Schomburgk 825 (K); Penal Settlement, 14 Dec 1919, Hitchcock 17239
(GH). BERBICE: Corentyne River, Greale, Oct 1879, Jenman 361 (K,NY);
Governor Falls, Corentyne R. , Sep, Forest Dept. Br. Guiana 6487 (NY);
Powis ck., I.R. upper Courantyne River, 6 Nov 51, Forest Dept. Br.
Guiana 6776 (MICH,NY). DEMERARA: envoi des Demerara, 1824, Parker s.n.
(G). ESSEQUIBO: Harina Sub. Exp. Stat., Koriabo River, 16 Sep 1934,
Archer 2537 (F,US); Mabaruma Hill, 18 Sep 1934, Archer 2547 (US);

371
Mazaruni Station, 27 Sep 1937, Sandwith 1604 (U); l.c., 22 Dec 1937,
Forest Dept. Br. Guiana 2583 (NY); Parika, 18 mi W of Georgetown on
Essequibo River, 19 Nov 1919, Hitchcock 16797 (GH,NY,S) and 16805 (GH,
NY,S); Roraima, R. Schomburgk 849 (BM-2 sh., CGE.W).
VENEZUELA. AMAZONAS: Lower Orinoco, 1896, Rusby 8 Squires
5•n. (NY) and 184 (NY). BOLIVAR: Abededores de El Dorado, Nov 1948,
Aristequieta 3750 (US,VEN); Auyan Tepui, Tannier & Schwabe s.n. (VEN);
La Paragua, 70 m, 10 Mar 1940, Will jams 1 2556 (F,VEN) and 1 2557 (A ,F,
VEN); San Mateo, Bajo Paragua, 75 m, 10 Apr 1940, Williams 12836 (F,
MICH,UC,VEN- 2 sh.); Laja, S of Rio Chiguirete, 420 m, 11 Oct 1 954,
Wurdack 8 Guppy 26 (NY,US,VEN). DELTA AMACURO: downstream from Rio
Yarakita River, Rio Amacuro, Sierra Imatura, 65-80 m, 9 Nov 1960,
Steyermark 87425 (NY,VEN); vie. sawmill between mouth Rio Cuyubini 8
1st main fork at Hacienda Caicarocoro, 14 Nov 1960, Steyermark 87526
(F,NY,US) or 87562 (F). MONAGAS: Reserva Forestal de Guarapiche, Cano
Colorado, Jul 1969, Aristeguieta, Liogier, & Guevara 7165 (US,VEN);
Chapopota, area circa de Quiriquire, 20 Nov 1955, Lasser 8 Vareschi
4151 (VEN); ca 2 km SSW of Jusepin, 125-150 m, 24 Mar 1967, Pursell,
Curry, 8 Kremer 8533 (NY); between La Toscana 8 Chaguaramal, 100-125 m,
11 May 1 967 , Pursell, Curry, 8 Kremer 8838 (NY,US); 6 km E of la Hormiga,
14-50 m, 24 Sep 1955, Wurdack & Monacino 39454 (NY). SUCRE: prov.
Mariguita, 1852, no. 6635, herb. Triana 4316 (BM). MIRANDA: San Jose
de los Altos, 8 Feb 1975, Zambrano & Medrano 117-147 (FLAS). ZULIA:
±58 S of Machigues near Mission Los Angeles del Tukuko, 200 m, 15 Oct
1966, Bruijn 1192 (NY,U-2 sh.,VEN).

372
C O L O M B I A_. Rio Casanare, Esmeralda, Los Llanos, 130 m,
19-20 Oct 1938, Cuat recasas 3809 (F-2 sh.) - BOYACA: Orocue, Rio Meta,
350 m, 23-4 Sep 1965, Garcia-Barriga, Hashimoto, & Ishikawa 18534
(GH,NY).
S T. V I N C E N T S. 1818, Caley s.n. (CGE,NY,W-3 sh.).
LOCALITY UNKNOWN. Guiana, Panama, Trinidad,
28 Oct 1970, Tuinen [sic] 4245 (U 254533B).
19. Cl i tori a javitensis (H.B.K.) Benth., Journ. Linn. Soc. 2: 42.
1858. emend. Fantz
Neurocarpum javitense H.B.K., Nov. Gen. Sp. 6: 409. 1823.
Clitori a grandifolia Ducke, Arch. Jard. Bot. Rio de Jan.
5: 141. 1930.
Clitoria portobellensis Beurl., Kingl. Vetenskaps Acad.
Handl. p. 119, 1854.
Ternatea javitensis (H.B.K.) Kuntze, Riv. Gen. PI. 1_: 210.
1891.
Liana, tall, climbing to treetops, or less commonly an erect shrub,
2-4 in tall, polymorphic. Branches 3-9 mm thick, pith hollow
longitudinally striate to sulcate, terete, juvenile branch pubescence
uncinate and short, rufus, pilose, becoming glabrate; bark dark brown,
peeling in longitudinal strips exposing lighter layers beneath. Leaves
3-foliate, coriaceous, leaflets variable in shape, size, and pubescence,
mainly narrow to broadly elliptic-oblong, occasionally elliptic,
lanceolate-elliptic, lanceolate-ovate, ovate, or oval, apex acuminate
from an acute to obtuse lamina, acumen 1-3 cm, more or less mucronate,

373
base broadly cunéate to rotund, midrib impressed above, often more or
less short pilose, rarely pilose on primary nerves, primary nerves of
9-13 (15) pair, upper surface green to dark green, glabrous, lower
surface green, pubescence typically appressed, short, slightly
conspicuous, becoming glabrate to glabrous, or occasionally pilose
becoming glabrate with the spreading trichomes confined to nerves,
lamina 7-18 (23-28) cm long, 3-9 (18) cm wide. Petiole subterete,
longitudinally striated to caniciluate, often gradually twisted,
4-17 (25) cm; pubescence uncinate and with rufus, appressed to suberect
trichomes, becoming glabrate; rachis 1-4.5 cm. Petiolules subquadrate,
4-9 (10) mm, pubescence uncinate and pilose, rufus. Stipules deciduous,
ovate to lanceolate, acute, 3-8 (10) mm long, 1-3 mm wide, pubescence
uncinate and appressed, densest towards apex and more or less ciliate;
stipels semipersistent, linear to subulate, acute, lateral stipels
typically 3-7 mm long, 1-1.5 mm wide, terminal stipels 1-3 (4) mm long,
0.5-1 mm wide. Inflorescences axillary, solitary, and caul if1orous,
solitary to few-fascicled, several to multi-flowered, racemose, nodose;
axes pubescence dense, short-pilose, more or less rufus, axes typically
(0.5) 1-5 cm long, occasionally more elongate, 4-18 cm long. Pedicels
4-9 mm, thickened to 3-4 mm in fruit. Bracts ovate, acute to acuminate,
spreading to reflexed in age, pubescence uncinate and appressed; middle
pair and outer bract persistent, 2-6 mm long, 1-3 mm wide. Bracteoles
minute to short, ovate to lanceolate, broadly acute to short-acuminate,
2-6 mm long, 1-3 (9) mm wide, inserted 1-2 nun below calyx base or
occasionally subopposite to alternately inserted, 2-5 mm below the
calyx, pubescence appressed, more or less ciliolate. FIowers large,
showy, (5.5) 6-8 cm, various shades of pink to rose, less commonly lilac

374
or lavender to pale violet; veins of vexillum dark pink to red or
purplish within; alae and carinas whitish to pale pink to rose. Calyx
tinged purple, pubescence appressed, uncinate trichomes lacking, tube
(15) 17-24 mm long, 5-10 mm wide at base to 8-13 mm wide at throat,
lobes broadly deltoid to ovate, acute to short acuminate, 4-8 mm long,
3-4 mm wide at base, ventral lobe subequal, 5-9 mm long, 1-1.5 mm wide.
Vexillum pubescence moderate to dense, appressed, rufus becoming tawny,
blade 3.5-5 cm wide, claw 15-19 mm. Alae extended well beyond carina
7-10 mm, blade 22-31 mm long, 7-12 mm wide, claw uncinate pubescent,
18-27 mm. Carina falcate, blade 11-18 mm long, 3-6 mm wide, claw
30-38 mm. Staminal tube glabrous, (34) 38-49 mm long, free filaments
2-5 mm; anthers lanceolate, 1.8-2 mm long, 0.5-0.8 mm wide. Gynophore
5-7 mm, pubescence dense, tawny-rufus; ovary 17-22 mm long, 1.5-1.8 mm
wide; pubescence dense, appressed, rufus to tawny; style 16-25 (28) mm
long, geniculate 6-9 mm from distal end, pubescence uncinate and
densely bearded; stigma subcapitate, ca 0.7-0.9 mm wide. Legume long-
stipitate, exerted beyond the calyx, flat, green to brown tinged reddish,
pendant, weakly raised around seeds, pubescence dense, appressed to
suberect, rufus, becoming thinned with age, valves 18-24 cm long, 18-25
mm wide; beak when present typically 3-7 mm; dehiscence causing valve
to twist one-quarter to one-half of a turn; stipe 2-3 mm thick, (19)
24-37 mm long, pubescence uncinate and stigose to pilose, rufus. Seeds
smooth, thick-lenticular, face suborbicular to slightly longer than
wide, dark brown to black, 9-12 mm Ion, 9-10 mm wide, 3-4 mm thick,
7-11 seeds per pod; hilum 2 mm x 1 mm. Figures 44 and 45.
The Yavita Clitoria is characterized as a scandent shrub with large
pinkish flowers, appressed pubescent calyx, long-clawed vexillum, and

Figure 44. Cl i tor i a javi tensis - I. Var. javitensis f. javitensis:
(a) leaflet, x 1; (b) inflorescence .txis, x 1; (c) flower,
x 1; (d) vexillum, x 1; (e) caiyx x I; (f) ala and carina,
x 1; (g) androecium, x 1; (h) gynoecium, x 1; (i) legumes,
x 1; (j) three views of seed, 1. (Wurdack & Adderley
41111 . NY: a,c. Maguire & Pol ¡ti 28296, NY 3597”: b,d-h.
Maguire, Cowan, & Wurdack 30420, NY: i. Williams 15676,
VEN: j.)

376

Figure 45. Cl i tor i a javitensis - II. Var grandifolia: (a) leaflet,
x 1; (b) inflorescence, x I. Var. klugii: (c) inflorescence,
x 1; (g) calyx and bracteoles, x 1. Var. 1ongi1oba:
(d) flower, x 1. Var. javiten is f. bracteosubten'da:
(e) calyx and bracteoles, x 1. Var. portobeí1 ensis:
(f) calyx and bracteoles, x 1. (Ducl-e 20400,'RB' 21)400: a-b.
Klug 173, US 14551 72: c,g. Te.-.smann 3782, S: d. Wurdack &
Hoñachi no, 41 093: e. Piper 5107, US 1 165847: f. )

378

379
large fruits with conspicuous rufus trichomes. This species can be
easily recognized from all other members of the section Caul if1orae by
the larger fruit size and reddish usually fairly dense pubescence, and
from most members of the section by the long vexillum claw and lack of
uncinate hairs on the calyx and fruit.
PHENOLOGY: This species has a long flowering period, from
September to flay. Most of the collections made of flowering specimens
occurred from November through February. Fruits were usually collected
from late December to May.
TYPE COLLECTION: VENEZUELA. Amazonas: Yavita, Humboldt &
Bonpland (HOLOTYPE: P-not seen).
The type was not included in the loan specimens to this author.
An attempt to obtain this particular specimen, or a photograph of it,
proved fruitless as the specimen could not be located in 1975. A
representative concept of the type was obtained from three main sources.
First, the original description was very detailed. Second, Sandwith
(1931) discussed the leaves, calyx, and bracteoles based on his
examination of the type at Paris and specimens cited by Bentham in
1858 when he transferred the species from the genus Neurocarpum to the
genus Clitoria. Sandwith annotated some of these specimens, such as
Spruce 2320-A,^ noting that the indumentum of the calyx and vexillum,
4. This specimen at Kew is a mixed collection. Spruce 2320-A
represents the branch with leaves mounted at the left, the branch with
flowers mounted on the right, and the packet of loose flowers. Spruce
2320-B represents the branch with leaves mounted in the center plus
the individual mounted flower. Pencil lines drawn around these were
probably made by either Brown or Sandwith, the two whose annotation
notes appear on the sheet. The letters were placed by this author to
distinguish the parts of the collection and for reference points in
citing the material. A material is C^ javitensis. B material is
C. saqotii.

380
bracteole size, and general appearance of the calyx "agree well with
those of the type of C. javitensis at Paris." Sandwith selected Spruce
s._n^ (San Carlos, Venezuela, Aug 1853: K!) as the best match to the
Paris type from those specimens in Herb. Benth. and Herb. Hook. Third,
a collection collected ca 20 km from Yavita (En las malezas de Maroa,
Rio Guainia, 127 m, Hi 11 jams 14344) agrees with the description by
Humboldt, Bonpland, and Kunth. These two specimens (F 1193498, VEN 4037)
also agree with the description by Sandwith (l.c., p. 357) of the type,
and match the indumentum of the calyx and vexillum, bracteole size, and
general appearance of the calyx and leaves of Spruce 2320-A. Thus
Hi 11 jams 14344 and Spruce s.n. served as a substitute for the type.
The concept of this species has changed with time. Bentham (1858)
transferred the species from the genus Neurocarpum to the genus Clitori a.
In addition to the type, he cited several additional specimens, which
included Spruce 1877 and 2320 from northern Brazil, and Rob. Schomburgk
1000 and Rich. Schomburgk 1 723 of British Guiana. This produced a
much broader interpretation on the species. Sagot (1882) described a
variety (C_. javitensis var. glabra) based upon his collection, Sagot 120,
from French Guiana. In the mid 1920's, N. E. Brown annotated the Kew
sheets of the Bentham "types" and concluded (as indicated by his
annotations on these sheets) that most of these specimens were distinct
from each other. He referred to them as distinct species, but never
named them. Sandwith (1931) concluded that C. javitensis had much
variation in length and position of the bracteoles, in the size of the
calyx, corolla, and leaves, and in leaf pubescence. He concluded that
there were two basic groups, which he treated as varieties. The
typical variety included the Humboldt & Bonpland (type) and Spruce 1877

381
specimens. The Sagot variety (var. glabra) included Sagot 120,
Schomburgk 1000, Schomburgk U23, and Spruce 2320, plus additional
specimens cited by Sandwith for the first time. Based upon the
identifications of uncited specimens examined by the author of the
present study, Sandwith's treatment generally has oeen accepted.
The present treatment disagrees with Sandwith's concept, and
corresponds closely to N. E. Brown's conclusions. The author of this
study has examined those specimens cited in the early literature plus
many more that fit the broad interpretation of javitensis as viewed
by Sandwith. They contain a mixture of morphologically variable plants
that can be separated into distinct groups by both morphological and
geographical characters. The concept of C. javitensis in this treatment
has been amended to exclude Sagot's variety, (i.e. exclude C. javitensis
var. glabra) and the Spruce 1877 specimen. In addition, two new names,
C. portobellensis Beurl. and C^_ grandifolia Ducke, have been placed in
synonymy.
A new species has been established for those specimens included
under the concept of C_._ javitensis var. glabra. This species has been
given the name sagotii (the name glabra would cause confusion with
the existing species, C. glaberrima). Clitoria sagotii differs from
C. javitensis by many characteristics, including morphological
differences in the vegetative structures, the inflorescence and its
associated structures, the floral parts, the reproductive structures,
the fruits, and the seeds. Table 8 summarizes some of these differences.
In general, C. sagotii is found in the Guianas and rarely in Venezuela
or Brazil, with C. javitensis distributed in Venezuela, Colombia, Peru,
western Brazil (Amazonas), and Central America. Sagot 120 and French

382
Table 8. A comparison of some structures between C. javitensis
(H.B.K.) Benth. and C. sagotii Fantz.
CHARACTER
C. JAVITENSIS
C. SAGOTII
LEAVES:
midrib above*
Impressed
Raised
INFLORESCENCE:
1ength*
pubescence*
1-6 cm (occ. to 20 cm)
rufo-strigose
Subsessile to 0.5 cm
predominately uncinate
CALYX:
tube*
pubescence*
lobes
17-24 mm
appressed
4-8 mm
11-16 mm
uncinate predominately
5-8 or 8-13 mm
BRACTEOLES:
1ength
insertion
2-3 mm or 4-6 mm
1-2 mm (occ. 2-5 mm)
4-7 mm or 8-11 mm
2-5 mm
VEXILLUM:
claw*
pubescence*
15-19 mm
tawny-rufo appressed
6-9 mm
uncinate, few appressed
ALAE:
beyond carina
blade length*
8-10 mm
24-30 mm
8-13 mm
17-21 mm
CARINA:
shape*
claw*
falcate
30-38 mm
nearly straight
25-32 mm
FLOWER SIZE
(5.5) 6-8 cm
5-6 cm (occ. 6-8 cm)
ANDROECIUM:
column
free filaments
connective
(34) 38-49 mm
2-5 mm
acute
32-39 mm
4-7 mm
apiculate
GYNOECIUM:
ovary length*
ovary pubescence*
ovary width*
17-22 mm
rufus to tawny
1.5-1.8 mm
10-14 mm
white tinged yellow
1-1.2 mm
LEGUME:
1ength*
width*
pubescence*
stipe
dehiscence*
18-24 cm
18-24 mm
dense rufo-appressed
24-33 mm
1/4 to 1/2 turn
9-15 cm
15-18 mm
uncinate + scattered pilose
30-37 mm
1-2 turns

383
Table 8. - Continued
CHARACTER
C. JAVITENSIS
C. SAGOTII
SEEDS:
1ength*
9-12 mm
6-7 mm
width*
9-10 mm
7-8 mm
thickness
3-4 mm
5 mm
longer axis
1ength
width
seeds per pod
7-11
4-8
*major differences

384
Guiana specimens represent one variety (C. sagotii var. sagotii).
Schomburgk 1000 and Schomburgk 1723 and other Guyana collections
represent a second variety (C. sagotii var. caniculata). Spruce 2320-B
and a couple of additional collections of the western Amazon basin
represent a third variety (C. sagotii var. sprucei). Additional data
on the species are found under the treatment of C. sagotii species
number 23.
Spruce 1877 appears similar to javitensis in general appearance,
except for the longer calyx lobes. Upon examination, there are several
conspicuous differences, including floral and reproductive structures.
This specimen agrees with Cavalcante 2570 which has uncinate pubescent
fruits, which are not characteristic of javitensis. Additional data
are found under the species C_. cavalcantei, species number 20.
Of the early cited collections for javitensis, only Spruce 2320-A
and Spruce s.n. are retained with the type collection in the concept of
C, javitensis in this treatment.
The type specimen for the name portobel 1 ensis is at Stockholm
(Panama: In silvis monitium, Porto-Bello, Apr 1826, Billberg s.n. !).
This specimen and other Panamanian collections differ from the type of
C. javitensis by the longer stipules, bracts, and bracteoles, and in
having the bracteoles subtending the calyx base. Specimens in fruit had
stipes that averaged somewhat shorter in length. The specimens agreed
highly in all other structures. These few differences suggested that
one species was involved. Thus C. portobel1ensis Beurl. is synonymized
with C. javitensis and reduced from a species to the varietal level as
C. javitensis var. portobel1ensis.

385
Ducke cited three collections (syntypes: Ducke 20400! Ducke
20399! Kuhlmann 1493!) from the Botanical Garden Herbarium at Rio de
Janeiro when he published the name C. grandifolia. These specimens
differ conspicuously from the type collection of _C^_ javitensis by the
elongate inflorescence and the larger leaves (herein called the
"grandifoliate" leaf). The inflorescences were multiflowered, with
fruits on one collection and aborted flowers on the other. The
"grandifoliate" leaf was much larger in size, long-petiolate, and had
longer petiolules. Leaflets varied from glabrous to rarely pilose below,
with both extremes occurring in the same collection. The fruit and
seeds match those of javitensis, as do the calyx, floral structures,
and reproductive structures. The calyx lobes, stipules, and bracteoles
were longer than the type specimen of .javi tensi s, but their size
agreed with the variation in additional specimens that matched in most
structures, including the inflorescence and leaves. Within the C.
javitensis complex, specimens exist with shorter inflorescences and
"grandifoliate" leaf types. Also, some specimens occasionally will
exhibit slightly longer inflorescences, up to 10 cm, especially in some
of the cultivars associated with the Central American material. In
other species (i.e. C. arbórea, C. polystachya), there exist a number
of collections with elongate inflorescences and others with contracted
inflorescences. Since most structures agree with those found in
C. .javitensis specimens, and the only major differences that occur are
in the size of the inflorescence, petiole, petiolule, and lamina of the
leaf, the name C. grandifolia Ducke is synonymized with C. javitensis.
As with the Central American name, CW_ grandifolia Ducke is reduced from
the species level to the varietal level, as C. javitensis var. grandifolia.

386
VERNACULAR NAMES: VENEZUELA. (AMAZONAS): Yanomono (Indios
Guaicas), Aristeguieta & Lizot 7444.
ECONOMIC IMPORTANCE: This species has been cultivated for its
climbing habit and its large, showy, pink flowers, When grown in open
places, the plant assumes a shrubby habit. The flowers are important
to the insects in the community. In Bolivar, Venezuela, Steyermark
74763 reported that the flowers harbor stinging ants. From the Barro
Colorado Island in Panama, Duke 15039(2) reported that the flowers
are very attractive to ants (Paraponera), bees, and thrips. In
Colombia, Hermann 11208 reported the roots had the odor of cucumber.
Some of the species of subgenus Neurocarpum are reported to have
odoriferous roots, and are noted for lack of insect and nematode damage.
NOTES: Clitoria javitensis is a polymorphic species which exhibits
a range of variation in its habit, bracteole insertion below the calyx,
and in the size of the leaflets, stipules, petioles, inflorescence,
bracts, bracteoles, and calyx lobes. This complex needs further study
incorporating experimental methods. From a morphological standpoint,
certain combinations of characters will consistently occur in plants
from defined geographic areas. These distinct groups are treated
nomenclaturally. They are easily recognized and segregated in a key.
Few exceptions occur, and where they do, these difficulties are noted
in the key.
Comments will be made first regarding the variation in structure,
and secondly, in the combination of characters that occurs within a
single geographic region. The habit of the species is usually reported
as a woody vine or a liana. Occasionally, the species is noted as an
erect shrub 2-4 m tall, or as a scandent shrub. Johnston (1949)

387
expanded on the habit of individuals on San Jose Island noting it was
a woody vine in dense forest and an erect shrub in thickets and in more
open areas. Rarely, tree habits are reported (i.e. Kill ip 34304),
although morphologically these specimens are indistinguishable from the
climbing specimens collected. The only exception was the two
arborescent specimens from Peru which were segregated morphologically
and treated as var. klugii.
The leaves and its associated structures are highly variable.
Leaves generally can be divided into two groups, called the "typical
leaves" (those which match the type specimen) and the "grandifoliate
leaves." The typical leaf is short-petiolate, has a short petiolule,
and a smaller sized leaflet. The petiole usually is 4-12 cm long
sometimes elongating to 13-15 cm on larger leaflets. The petiolules
are usually 4-6 mm long to rarely 7-8 mm on larger leaflets. The
typical leaflet is elliptic to elliptic-oblong or lanceolate-elliptic,
7-18 cm long, 3-7 (9) cm wide. The grandifoliate leaf is long-
petiolate, has a long petiolule, and a larger sized leaflet. The
petiole is usually 10-20 (25) cm long, even on younger leaves. The
petiolule is usually (6) 7-10 mm. The grandifoliate leaflet is usually
ovate, ovate-elliptic, to lanceolate-ovate, becoming very broad,
15-28 cm long, 6-15 cm broad. The typical leaf is generally associated
with shorter inflorescences, whereas the grandifol iate leaf is often
found on individuals with elongate inflorescences. But some exceptions
do occur, particularly in Meta, Colombia and parts of Amazonas,
Venezuela. Therefore, the typical leaf and the grandifoliate leaf type
are helpful as supportive characters, but not reliable as diagnostic
characters.

388
The stipules of those members of Central America and most of South
America are generally 6-10 mm long. In Amazonas, Venezuela, and adjacent
geographical areas, the stipules are consistently 3-5 mm long, with
rarely a stipule of 6 mm. Stipules are deciduous, and not always
observed on a particular specimen.
The most common inflorescence is short, (0.5) 1-6 cm long, few-
flowered. Rarely an individual may possess an inflorescence 8 cm long
or in cultivars to 10 cm. The uncommon inflorescence is elongated,
usually 4-18 cm long, multi-flowered. Problems may arise in
identification with those specimens whose inflorescences are 4-7 cm
long, as in a number of Meta, Colombia, collections. Pedicels can be
used as a supportive character. The short inflorescence type has
pedicels commonly 3-7 mm, although some individuals may bear longer
pedicels. The long inflorescence type has pedicels of 6-11 mm.
Bracts are typically minute, 2-3 (4) mm long, highly consistent in
South American members. Bracts of Central American and the adjacent
Colombian members are noticeably longer, rarely 3 mm long, commonly
4-6 mm. The bract length refers specifically to the middle pair of
bracts which are more persistent, usually the largest, and are concave
and subtending the pedicel. The outer bract is more or less persistent,
and if present, may be slightly shorter.
Bracteoles are commonly 4-7 mm long. Specimens from Amazonas,
Venezuela, and adjacent areas have minute bracteoles, typically 2-3 (4)
mm long. In addition, these minute bracteoles often are inserted
2-5 mm below the calyx base whereas most bracteoles subtend the base,
inserted 1-2 mm below it.

389
Calyx lobes ore commonly 4-6 mm long, broadly deltoid with the
length subequal in- the width or to 1.5 times the width at the base. The
apex is acute or sometimes short-acuminate. This type of calyx lobe is
typical of Central American members and of most of those from South
America. In Peru and the western Amazon basin, calyx lobes are elongate,
the length nearly twice the width, with an acuminate to subulate apex.
This lobe type is mostly associated with individuals having the longer
inflorescences.
The stipe is 19-37 mm long. Specimens with elongated inflorescences
have stipes 19-24 mm. Central American specimens have stipes that
averaged 24-32 mm. Other South American specimens with shorter
inflorescences have stipes 28-37 mm.
A comparison if the patterns of variation and combinations of
structures that are associated together suggest five distinct groups.
The five varieties recognized in this treatment are summarized in
Table 9.
The typical variety must be the one that includes the type from
Amazonas, Venezuela. This group is quickly recognized by the minute
bracteoles inserted well below the calyx, and the short stipules. The
inflorescence is short with flowers bearing short calyx lobes, and fruits
1ong-stipitate. The leaves are of the typical type. This group is
treated as variety javitensis.
A second group is very similar, although the bracteoles are 3-4 mm
long and inserted at the calyx base. This difference is a minor
variation from the typical bracteole form, and therefore, this group is
treated as a form, javitensis f. bracteosubtenda.

Table 9. A comparison of some structures between the varieties of C. javi ten sis.
CHARACTER
JAVITENS IS
PORTOBELLENSIS
L0NGIL0BA
GRANDIFLORA
KLUGII
INFLORESCENCE
Short
Short
Short
Long
Long
LEAF TYPE
Typical
Typical
Typical
Grand!foliate
Typical
PETIOLE
Short
Short
Short
Long
Long
PETIOLULE
Short
Short
Short
Long
Short
STIPULE
Short
Long
Long
Long
Long
BRACTS
Short
Long
Short
Short
Short
BRACTEOLE LENGTH
Short
Long
Medium
Long
Mediurn
BRACTEOLE INSERTION
Subtend or
Subtend
Subtend
Subtend or
Subtend
Below 2-5 mm
Below 2-3 mm
CALYX LOBES
Short
Short
Long
Long
Long
Short
Lonq
Typical
Grandifoliate
KEY: Inflorescence
(0.5) 1-6 cm
4-18 cm KEY:
Leaflet length
7-18 cm
15-28 cm
Petiole
4-10 (14) cm
10-20 (25) cm
Leaflet width
3-7 (9) cm
6-15 cm
Petiolule
4-6 (8) mm
6-10 mm
Leaflet basic
Stipule
3-5 mm
6-10 mm
shape
Obiong-ellip.
Ovate
Bracts
2-4 mm
4-6 mm
Bracteole
2-4 mm
4-7 mm
Calyx lobes
4-6 (7) mm
(5) 6-9 mm
390

391
A third group is easily recognized by the elongate bracts and its
Central American distribution. In addition, the stipules and bracteoles
are longer than the typical variety, and the stipe is shorter. This
group is treated as var. portobel1ensis, making a new combination from
the prior specific epithet for this group. The leaves have appressed
to glabrate pubescence on the lower surface as in the typical variety,
but a frequently appearing form has a pilose pubescence, an uncommon
pubescence type in the species. These members are treated as var.
portobel1ensis f. pilosa. There also occurs a rare leaf shape in this
group in which the leaflet apex is truncate. This apex type is unknown
elsewhere in the subgenus Bractearia, and rarely occurs in other
species distributed in Africa (e.g., C. kaessneri) or Southeast Asia
(e.g., C. javanicd). This form is trated as var. portobel1ensis f.
truncata.
A fourth group was noted that has the typical leaf type and short
inflorescences, but much longer calyx lobes. The stipules and
bracteoles are also longer than the typical variety, although the bracts
agree. Fruits in this group were not observed, thus the stipe length is
unknown. These members are distributed in Peru. This group is treated
as var. longiloba.
A fifth group is quickly recognized by its elongated inflorescences
and the "grandifollate" leaf type. Calyx lobes are longer than the
typical variety as well as the stipules and bracteoles. This group is
treated as var. grandifolia, making a new combination from the prior
specific epithet applied to this group.
The placement of the last group is somewhat problematic. Most
characters agree within the variation found in the javitensis complex.

392
The elongate inflorescence and 1ong-petiolate leaves suggest var.
grandifolia, but there are several differences, notably in the
pubescence of the ovary, habit, and size of the petals, style,
petiolules, and leaflets. The two collections comprising this group
are reported as trees (only Kill ip 34304 is reported as a tree from the
complex) and, atypically, have an ovary pubescence that is white with
a yellow tinge toward the trichome apex. Members of the C. javitensis
complex have reddish-brown to yellowish-brown trichomes. This suggests
that these two members are distinct from C. javitensis, although all
other characters are in more or less agreement. Species with whitened
pubescence on the ovary typically have uncinate pubescent fruits, but
fruits are unknown in this group. Both specimens are from Peru,
whereas those species with fruits bearing uncinate trichomes are
typically found in the Guianas and adjacent Brazil. With a lack of
segregational characters, this group is presently treated as var. klugii,
rather than as a new species.
Cl i tor ia javitensis has close affinities with arborescens from
which it can be easily distinguished by the non-tomentose leaflets, the
lack of uncinate trichomes on the calyx, the much shorter bracteoles,
and larger flowers.
DISTRIBUTION (Figures 46 and 47): The species has a wide
distribution in South America, north and west, from the western Amazon
Basin with an extension into Central America. Members are commonly
found in southern Venezuela, western Brazil, Peru, Colombia, and
Panama. Those plants known as C. javitensis var. glaibra in the Guianas
and parts of Brazil, belong to other species. The species is found in

Figure 46. South American distribution of Clitoria javitensis, section
Caul i florae. Var. javitensis f. javitensis Té) 1 var.
portobellensis f. portobellensis (★)'» var. longiloba
iorr var. grandifoí ia (^j; var. kiugi i .

394
l_-

Figure 47. Central American and northern South American distribution
of Cl i tori a javitensis, section Cauliflorae. Var.
javitensis f. bracteosubtenda (A1; var. portobellensis
f. pilosa (¿u*); var. portobel 1 ensis f. truncata ( ).

396

397
tropical humid forest, infrequently in thickets or scrub, usually
associated with sany soil, at elevations up to 500 (800) m.
KEY TO VARIETIES AND FORMS:
1.Inflorescence short, 0.5-6 cm long, few-flowered; leaves typically
short-petiolate, petiole 4-10 (15) cm; petiolule 4-6 mm, rarely
longer (to 8 mm); stipe elongate, 24-37 mm long.
2.Calyx lobes with length subequal to 1.5 times the width,
4-6 mm long, apex acute to short-acuminate.
3.Bracts minute, 2-3 (4) mm; stipules short, 3-5 (6) mm;
stipe averaging longer, 28-37 mm (Central and Western
Amazon Basin) 19a. var. javitensis
4.Bracteoles minute, 2-3 (4) mm long, subopposite to
alternate on the pedicel, inserted 2-5 mm below calyx
base, apex below calyx base or more or less reaching
it but below the disc . . . . 19aa. f. javitensis
4. Bracteoles typically longer, (2) 3-4 (5) mm,
subopposite, subtending calyx base, inserted 1-2 mm
below, apex more or less reaching the swollen calyx
base represented internally by the disc
19ab. f. bracteosubtenda
3. Bracts large, 4-6 mm; stipules long, (5) 6-9 (10) mm; stipe
averaging shorter, 24-32 mm (Central America and adjacent
Colombia; N. Venezuela) 19b. var. portobel1ensis
5. Leaflet apex tapering, acuminate
6.Leaves with pubescence on lower surface inconspicu¬
ously appressed to glabrate, trichomes on nerves
appressed, non-spreading . 19ba. f. portobel1ensis

398
6. Leaves with pubescence on lower surface
conspicuously pilose to glabrate, trie home s
on nerves spreading to suberect
19bb. f. pilosa
5. Leaflet apex truncate, more or less retuse . . .
19bc. f. truncata
2. Calyx lobes with length nearly twice the width, (5) 6-9 mm long,
apex acuminate to short-subulate (Peru) . 19c. var. longiloba
1. Inflorescence elongated, 4-18 cm, multi-flowered; leaves typically
long-petiolate, petiole commonly 10-20 (25) cm; petiolule typically
6-9 mm, rarely shorter (var. kjugij); stipe short, 19-24 mm.
7. Ovary pubescence reddish to yellowish brown; style 27-30 nun;
ala blade 25-33 mm long; carina blade 15-18 mm long;
petiolules 6-10 mm; rachis typically 3.5-5 cm; scandent shrub
(Western Amazon Basin) 19d. var. grandifolia
7. Ovary pubescence white, tinged yellow; style 22-25 mm; ala
blade 21-25 mm; carina blade 12-15 mm; rachis 2-4 cm;
petiolule 4-6 mm; tree (Peru) 19e. var. kl ugi i
19a. Clitoria javitensis (H.B.K.) Benth. var. javitensis
Leaflets typically small, 7-18 cm long, 3-7 (9) cm wide. Petiole
short, 4-10 (14) cm. Petiolule 4-6, rarely to 8 mm long. Stipules
short, 3-5 mm. Inflorescences few-flowered, short, (0.5) 1-6 cm,
rarely to 8 cm. Pedicels commonly 3-7 mm, occasionally to 9 mm. Bracts
short, 2-4 mm. Bracteoles small, 2-4 mm. Calyx lobes 4-6 (7) mm,

399
length subequal to 1.5 times the width, apex acute to short-acuminate.
Stipe 28-37 mm.
DISTRIBUTION (Figures 46 and 47): This variety is distributed in
the central and western Amazon Basin which includes southern Venezuela
and Colombia, and Amazonas, Brazil. The bulk of the typical form is
centered in Amazonas, Venezuela. Most members of form bracteosubtenda
are found in Colombia, Bolivar, Venezuela, and Amazonas, Brazil.
19aa. Clitori a javitensis var. javitensis f. javitensis
Bracteoles minute, 2-3 mm long, rarely 4 mm, subopposite becoming
alternate on pedicels, inserted below calyx base 2-5 mm, apex below the
calyx base or more or less reaching it, but below the disc.
DISTRIBUTION (Figure 46): Hort. Kew, 1850, Forsberg s.n. (S-2 sh.).
VENEZUELA. Croizat 922 (NY) and 928 (NY); 17 May 1942,
Hi 11 jams 1 5650 (VEN). AMAZONAS: along Cano Maminare, a back-water
caño emptying eastward just north of Caño Darati, 140 m, 2 Apr 1953,
Maguire & Wurdack 34783 (NY); Playa alta, 3 Nov 1950, Maguire, Cowan,
& Wurdack 29363 (NY,VEN); along Rio Padamo ca 15 km above mouth, 150 m,
27 Mar 1 953, Maguire & Wurdack 34716 (NY,VEN); Capibara, arboredas del
Casiquiare, 120 m, 29 May 1 942, Williams 15676 (F,VEN); Brazo
Casiquiare above Capibara, 20 Jan 1931, Holt & Blake 661 (US); at mouth
of Rio Pamoni and Rio Casiquiaro, 130 m, 12 Feb 1954, Maguire, Wurdack,
& Bunting 37677 (NY); Mavaca, alto Orinoco, Jan 1970, Aristeguieta &
Lizot 7444 (VEN); El Temblador, Rio Orinoco, entre San Antonio y Ki,
6 May 1971, Foldats 252-A (VEN); along Rio Orinoco just above Tamatama,
150 m, 22 Jun 1 959, Wurdack & Adderly 43121 (NY); Orinoco R. , southern

400
part of Isla de Ratón, 5°2,N-67°46'W, 90 m, 19 Nov 1965, Breteler 4736
(U, VEN); abajo de Esmeralda, 143 ni, 13 May 1942, Williams 15348 (E,G,
NY,RB,US,VEN); above Cano Guasuriapan, Rio Atabapo, 1 km above San
Fernando de Atabapo, 120-130 m, 16 Nov 1 953, Maguire, Wurdack, &
Bunting 36249 (N Y - 2 sh.,RB); Moroa, Rio Guainia, 9 Feb 1 942, Will i anís
14214 (VEN); l.c., 127 ni, 14 Feb 1 942, Will i anís 14344 (F,VEN); San
Carlos, Aug 1853, Spruce s.n. (K-432); prope San Carlos, 1853-4,
Spruce 3543 (NY-mixed; non G,GH,CGE,F,K,RB,W); above Raudal Pacure, Rio
Cunucunuma, 9 Nov 1950, Maguire, Cowan, & Wurdack 24903 (NY,VEN);
Culebra Savanna, Rio Cunucunuma, 25 Dec 1950, Maguire, Cowan, &
Wurdack 30420 (NY); along Rio Saipa between Rio Casiquiare and mouth of
Cano Hechimoni, 120 m, 25 Jul 1959, Wurdack & A_dderly 4361 6 (NY); along
Rio Siapa near mouth, 130 m, 4 Apr 1953, Maguire & Wurdack 34833 (NY);
vie. base camp, Cerro Sipapo (Paraque), 17 Jan 1 949, Maguire & Pol iti
28419 (BM,NY,US); l.c., 10 Jan 1 949, Maguire & Po1 iti 28296 (NY,US); vie
base camp, Cerro Yapacana, entre el Campamento y la Sabana Grande,
3°45'N-66045'W, 125 m, Steyermark & Bunting 103211 (VEN); Isla de
Callare, raudal del Trapichóte, delta del Ventuari, 125 m, 21 Apr 1942,
Williams 15002 (F,S,VEN).
B R A Z J__L. AMAZONAS: Rio Negro, Uacaria, 22 Apr 1947, Froes
22195 (U); Inambu, 17 Nov 1952, Romero-Castañeda 3600 (M0).
COLOMBIA. VAUPES: Rio Kuduyari, middle and lower course,
1°15'N-70°5'W, 700-800 ft, 16 Oct 1952, Schultes & Cabrera 17862 (GH,
NY,US); Bocas del Carurú, orillas caño, 230 m, 27 Sep 1 939, Cuatrecasas
7063 (US); Rio Cuduyari orillas, afluente del Vaupes, 200 m, 15 Sep
1939, Cuatrecasas 6822 (US); Yapoboda, 10 Dec 1943, Allen 3247 (M0).

401
META: Llanos Oriental is, selva higrofila del cano de Quenare, 400 m,
22-23 Feb 1941, Dugand & Jaramillo 2904 (US-2 sh.); Cano del "Tigre" en
los desenchocondura del Rio Cubuyarito, Feb 1935, Garcia 4973 (US);
Sierra de la Macarena, Rio Guapaya, 500 ni, 2 Dec 1 949, Phi 1 ipson,
Idroho, & Fernandez 1669 (BM,F).
19ab. Clitoria javitensis (H.B.K.) Benth. var. javi tens is f.
bracteosubtenda Fantz, f_^ noy.
Bracteoles small, (2) 3-4 (5) mm long, subopposite on pedicel and
subtending the calyx base, inserted to 2 mm below calyx base, apex more
or less reaching the abruptly swollen area of calyx base (representative
of internal disc).
TYPE COLLECTION: VENEZUELA. Bolivar: Rio Parguaza, between mouth
and LI Carmen, 50 km upstream, 80-110 m, 3 Jan 1956, Wurdack & Monachino
41093 (H0L0TYPE: NY. Isotypes: US 2167581 & 2167582, VEN 40651).
The New York specimen has a number of flowers with bracteoles of
different sizes, but all subtending the calyx as is typical of this form.
DISTRIBUTION (Figure 47):
VENEZUELA. BOLIVAR: Rio Nichare (afluente Rio Caura)
arriba de la desembocadura con el Rio Cicuta (Icuta), 6°151N-65°5'W,
200-250 m, 25 Apr 1966, Steyermark 95733 (NY,US,VEN); entre el Rio
Paramichi y el Salto de Chal imano, sureste hacia la frontera Ven.-
Brazil, N de la Serranía Pia-soi (Pia-shauhy, 525-650 m, 5 Jan 1962,
Steyermark 90594 (VEN); Rio Paragua, Salto de Auraima, 10 Apr 1943,
Ki11 ip 37374 (VEN; vie. camp 1 along Rio Abacapa, NW slopes of Abacapa-
Tepui, 420 m, 30-31 Mar 1 953, Steyermark 74763 (F-3 sh. ,NY-3sh.,VEN).

402
AMAZONAS: Fish Creek, Esmeralda, 350 ft, Aug 1928-Apr 1 929, Tate- 964
(NY); Murciélago Falls, Sipapo River, 17 Nov 1948, Maguire & Politi
27314 (NY,US); Sabana de Asisa, 11 Feb 1949, Phelps & Hitchcock 458
(NY); la Isla Laolao, Bajo Cano San Miquel, Rio Guainia, 127 m, 24 Mar
1942, Will jams 14863 (F,US); San Carlos de Rio Negro, 100 m, 6 May 1 942,
Will jams 14667 (VEN).
BRAZIL. AMAZONAS: Rio Negro Marabitonas, 20 Apr 1947,
Pi res 483 (NY-2 sh.); Porto Curucuhy, S. Gabriel, Rio Negro, 6 Oct
1945, Froes 21116 (NY); Sao Gabriel, May 1852, Spruce 2320 (K-mixed).
R0ND0NIA: Rio Abuna 3 km above confluence of Rio Negro, north bank,
16 Nov 1968, Prance et al. 8508 (NY).
C 0 L 0 M B I A. AMAZONAS: Rio Igaraparana, las arredadores de
La Chorrera, 180 m, 4-10 Jun 1942, Schultes 3954 (F,K). CAQUETA:
Florencia, Dec 1930, Arbelaez 7_34 (US). VAUPES: Rio Negro, San Felipe
and vie., below confluence Rio Guainia and Rio Casiquiare, 1°50'N-
67°0'W, 600 ft. 25 Oct 1952, Schultes, Baker, & Cabrera 18044 (NY,US).
META: Llano de San Martin, Quename, 250 m, Jan 1356, no. 6635, hb.
Triana 4314 (BM); Quetame, Bogota, Rarjnen (?) s.n. (W); Sabanas de
San Juan de Arana, margen izquierda del Rio Guejar, alrededores del
aterrizaje “Los Micas," 500 m, 22 Jan 1951, Idrobo & Schultes 1261
(U,US). NARIN0: Barbacoas Sta., 15 Oct 1862, Hayes s.n. (BM-mixed).

403
19b. Clitoria javitensis (H.B.K.) Benth. var. portobellensis
(Beurl.) Fantz, comb. nqv_.
Clitori a portobel1ensis Beurl. , Kongl. Vetenskaps Acad.
Handl. p. 119. 1854.
Leaflets typically small, 7-18 cm long, 3-7 (9) cm wide. Petiole
short, 4-10 (14) cm. Petiolule 4-6 (8) mm. Stipules long, 6-9 (10) mm.
Inflorescences few-flowered, (0.5) 1-6 (8) cm. Pedicels 3-7 (9) mm.
Bracts .long, 4-6 mm. Bracteoles long, 4-7 mm long, inserted 1-2 mm
below the calyx base, subtending it. Stipe 24-32 mm.
TYPE COLLECTION: PANAMA. In silvis montium, Porto-Bello, Apr
1826, Bill berg (LECTOTYPE: S-herb. Regnell.
Beurling did not select a type specimen. He published the locality
data as "In silvis montium." The name was published as "Clitoria
(Vexillaria?) portobel1ensis" which agreed with the identification on
the Billberg specimen deposited in a type folder at Stockholm. Based
upon the label data and the agreement of the specimen with the published
description, the Bill berg s.n. specimen (S) was the probable specimen
examined by Beurling upon which the name was based, and therefore, it
was selected as the lectotype.
DISTRIBUTION (Figures 46 and 47): This variety is frequently
collected in Panama and the adjacent provinces of northwest Colombia.
Additional collections have been made infrequently in northern
Venezuela and Costa Rica.

404
1 9ba. Clitori a javitensis (H.B.K.) Benth. var. portobel1ensis
(Beurl.) Fantz f. portobellensis
Leaves with pubescence on lower surface appressed, becoming
glabrate, trichomes then confined to major nerves, more or less
inconspicuous, appressed.
DISTRIBUTION (Figure 46): A few collections are leafless, and
designation of a form name is impossible. These collections are cited
with the more common, typical form, and marked with the symbol
before the institutions.
C 0 L 0 M B I A. ANTIOQUIA: swampy area on peninsula ca 1 km W
of Turbo, ca 8 '5'N-76°431W, sea level, 31 Mar 1 962, Feddema 2106 (MICH),
CHOCO: Bahia Solano, Puerto Mutis, 0-20 m, 4 Jan 1973, Gentry &
Forero 7]_79 (MO); Truando's Falls, Feb 1858, Schott 8 (*F,NY);
corregimento de Truando, orillas del Rio Chintado, a 3 km de su
confluencia con el Truando, mon. de. Riosucio, 200 m, Oct 1956,
Castaneda 6080 (NY); Rio Truando, 18 May 1967, Duke 11129(2) (NY);
Cupica, ca 50 m, 10 Feb 1947, Haught 5556 (M0).
PANAMA. LOCALITY UNKNOWN: Sinclair s.n. (K); Mar 1905,
Cowell 419^ (NY); Panama or Colombia?, 1831, Cuming 1142 (GH,K-2 sh.;
not BM); Cana and vie., 2000-6500 ft, 17 Apr-8 Jun 1908, Williams 755
(NY). DARIEN: Mannene to mouth Rio Cuasi, Kirkbride & Bristan 1402
(MO); Quebrada Sie rra from Rio Tuira to track of Chevy Expedition,
between los Rios Cube & Punusa, 27 Feb 1972, Gentry 4438 (MO); Lucuti,
Chepigana Dist., 50 ft, 5 Mar 1940, Terry & Terry 1372 (A,F,M0).
PANAMA: between Mt. Hope & Santa Rita Trail, 25 Feb 1905, Cowell 68
(NY) and 82 (NY); sabanas near Chepo, 30 m, 20 Jan 1935, Hunter & Allen

405
92 (G,M0). COLON: 10 mi SW of Puertobelo, 2-4 mi from coast, 10-200 m,
24 Mar 1973, Liesner 1084 (M0); near Salamanca, 8 mi E of Transisthmian
Highway, 100 m, 19 Dec 1972, Gentry 6704 (M0); Peluca, ca 27 km from
Transisthmian Highway on road to Nombre de Dios, trail to Rio Boquerón
& along bank, 25 Feb 1973, Kennedy 2642 (M0). CANAL ZONE: San Jose
Island, East Harbor, 2 May 1945, Erlanson 35 (G,NY,US); East Point near
Quarry of San Jose Island, Johnston 670 (GH-2 sh.,M0); Frijoli station,
6 Feb 1862, Hayes 492 (BM,E,K,M); l.c., 2 Feb 1971, Croat 13223 (M0);
pipeline road, 2-4 mi N of Gamboa, 100 m, 15 Dec 1972, Gentry 6531
(M0); between Alhajuela & El Vigía, 12 Jan 1911, Pittier 2368 (US-
frag.); Las Cascadas Plantation, 4 Jan 1924, Standley 29557 (US); W.
arm of Quebrada Salamanca, 70 m, 16 Dec 1934, Dodge, Steyermark, &
A1 l_en 1 7007 (G-2 sh.,GH,M0); hills around Aqua Clara Reservoir near
Gatun, 20-30 m, 5 Feb 1911, Pittier 2652 (F,NY,S,US). CANAL ZONE
(BARRO COLORADO ISLAND): 1931, Aviles 85a (M0); Jan 1939, Brown 101
(F) and 136 (F); 15 Jan 1968, Duke 15039(2) (NY); 24 Jan 1968, Dwyer
8452 (GH,M0); 30 Jan 1929, Frost 82-1/2 (F); Jan 1930, Zetek 15002
(F-frag.); laboratory clearing, 10 Jan 1969, Croat 7077 (M0); l.c.,
19 Sep 1973, Montgomery 74 (MO); above rock cliff at N end of laboratory
clearing, 10 Jan 1969, Croat 7071 (MO); cove N of lab, 1 Feb 1932,
Woodworth & VestaJ 478 (F); shore N of lab, 2 Feb 1 932, Woodworth &
Vestal 347 (A,F); shore near lab, 5 Feb 1932, Woodworth & Vestal 735
(A,F); l.c., 15 Feb 1931, Wilson 23 (F,MO); shore near Gatun Lake, E
of laboratories, 24 Nov 1 943, Ki11ip 40030 (MO-2 sh., US); electric line
clearing near dock, 10 Apr 1968, Croat 4615 (MO); E of dock, 3 Nov 1931,
Shattuck 312 (F,M0); clearing N of dock, 6 Feb 1969, Croat 7732 (MO);
N of boat dock, 6 May 1956, Stoutamire 2051 (*MICH); shore cove W of

406
Drayton House, 16 Feb 1932, Woodworth & Vestal 598 (A,F): William
Morton Wheeler Trail, 600 m, 31 Jul 1929, Wetmore & Woodworth 64 (A,F);
l.c., 800 m, 20 Dec 1931, Wetmore & Abbe 6 (F,GH,M0); junction Fairchild,
Wheeler, & Lathrop trails, 13 Feb 1969, Croat 7929 (M0); Lathrop trail,
130 m, 9 Mar 1969, Croat 8497 (M0); Tom Barbour Trail 1000, 25 yds S
of trail along ridge, 10 Dec 1 967, Croat 4216 (M0,NY); shoreline near
end of Armour Trail, 28 Feb 1969, Croat 8289 (M0); Donato Trail near
Lutz creek, 3 Apr 1970, Croat 9310 (*M0); Zetek trail, 1600 in ca,
31 Jul 1929, Wetmore & Woodworth 71 (A,F); Van Tyne Trail, 6 Jan 1940,
Hunnewel 1 16424 (*GH); forest n of animal cages, 10 Jan 1969, Croat
7091 (*M0); cliffs in straight at Fairchild Point, 16 Apr 1968, Croat
4853 (F,M0); steep rocky wall along Buena Vista Reach S of Fairchild
Point, 1 Oct 1968, Croat 6605 (M0); steep slope on east Pena Blanca
Point, 4 May 1968, Croat 5333 (M0-2 sh.); ME corner of S cove N of
Burrunga Pt., 16 May 1968, Croat 5595 (M0); shoreline east side Pena
Blanca Point across from front of #8 light, 5 Oct 1968, Croat 6734 (M0);
forest edge of clearing near stream, 20 Mar 1970, Croat 8905 (M0);
shore of 2nd point E of Pena Blanca, 23 Feb 1932, Woodworth & Vestal
675 (A,F,M0); between the station and Point Salud, 23 Dec 1963, Graham
204 (GH); NE corner of Orchid Island, 24 Apr 1968, Croat 5067 (M0); E
edge of Orchid Isle, 18 Dec 1970, Croat 12847 (M0); SW corner small
island at M Trail end, 19 Apr 1968, Croat 4967 (M0); corner small island
near lights, on way to Pearson Trail, 8 Sep 1929, Bangham 599 (A,F);
shoreline, large cove SW of Slothia Island, 19 Jan 1969, Croat 7340A
(M0).
C__0 S^ T _A R I C A. Boca Culebra, 50 m, 18-Jan 1898, Pittier
12342 (P). PUNTARENAS: ca 5 km W of Rincon de Osa, Osa Peninsula,

407
50-200 m, 8°42'N-33'31‘ W, 9-12 Jan 1970, Burger & Liesner 7247 (F);
Tinco Station, between Rio Esquinos & Palmar sur de Osa, sea level,
22 Dec 1950, Allen 5711 (FLAS); entre Parrita y Esterillos, 10 m,
30 Mar 1966, Jimenez 3901 (BM,F,GH).
19bb. Clitoria javitensis (H.B.K.) Benth. var. portobel1ensis
(Beurl.) f. pilosa Fantz, f. nov.
Leaf pubescence on lower surface moderately densely pilose,
becoming glabrate, trichomes then confined to major nerves spreading
to suberect.
TYPE COLLECTION: PANAMA. Canal Zone: vie. of Gamboa, 20 Dec
1946, Allen 3931 (H0L0TYPE: G-26, hb. Delessert. Isotype: M0:
1572352).
DISTRIBUTION (Figure 47):
MARTINI Q U _E. Introduce & cultive, Tarnasse, 1883, Puss
1074 (NY); cult. Jardin Botanique, Fort de France, 22 Dec 1938,
Stehle 3533 (NY).
V E N E Z U E L A. MIRANDA: prope coloniam Tovar, 1856-7,
Fend!er 2201 (K). CARAB0B0: between Valencia & Campanero, 2500 ft,
7-3-57, Anonymous (=Fend1er ?) 2201 (K). YARACUY: del Rio Guayabito,
em El Guayabito Playón, 15 km N de Marin, 150-250 m, 28 Nov 1971,
Steyermark & Bunting 105298 (VEN); cerca del Pueblo de Aroa, Mar 1959,
Aristeguieta 3841 (US-2 sh.,VEN).
PANAMA. LOCATION UNKNOWN: Panama, Cuming 1142 (BM; not GH,
K); Sotto Caballo, 28 Dec 1 957 , Smith Jr., Smith, & Arrang 3319 (PH).
DARIEN: vie. of Pinas, 2 Mar 1967, Duke 10617 (M0); Rio Sambu, 0-5 mi

408
above Rio Venado, 18 Jan 1967, Duke 9262 (M0,NY,US); Garachine, sea
level, Jan-Feb 1912, Pittier 5513 (US). PANAMA: Isla del Ray, 28 Jan
1967, Duke 9529 (M0); Rio Pasiga to above waterfall on 2nd main fork,
29 Oct 1971, Gentry 22_79 (MO). COLON: Juan Mina Plantation, Rio
Chagres, region above Gamboa, 25 m, 3 Feb 1947, Allen 4JJ4 (G,M0).
CANAL ZONE: Gamboa road, 3-2-33, Jones 268 (US); between Summit &
Gamboa, 20 Jan 1922, Greenman & Greenman 5239 (MO); Gamboa pipe line,
on Tropic Test Center lands, 2 km S from main dirt road, 28 Nov 1967,
Correa A. & Haines 481 (*M0); vie. Gamboa, 20 Dec 1946, Allen 3931 (G,
MO); around Gorgona, 15 Jan 1911, Guiliard 2405 (US: on Pittier label);
betweeen Gorgona & Mamei, 10-30 m, Jan 1911, Pittier 2405 (GH,NY);
Empire to Mandinga, 23 Feb 1923, Piper 5140 (US); Culebra, 50-150 m,
4 Jan 1911, Pittier 2231 (US); edge Chagres River near Gatun, Dec
1859, Hayes 679 (NY); Chiva-Chiva Trail, Red Tank to Pueblo Nuevo,
1 Mar 1923, Piper 5107 (US); Barro Colorado Island, shore in front of
Fuertes House, 16 Jan 1 932, Wetmore S A_bbe 6_A (GH-2 sh.,M0); Barro
Colo. Is., shoreline N of 1st cove S of #8 Rear Light, 1 May 1968,
Croate 5285 (M0-2 sh.). CANAL ZONE (SAN JOSE ISLAND): Guajaqui1-viken,
1852, Andersson s.n. (S); Area 7C, S of Spring Ravine on Canyon Rd.,
7 Feb 1946, Johnston 1351 (GH); Mareno Valley below Red Hill, 4 Apr
1945, Johnston 628 (GH); along M-area road, N half of island, 21 May
1945, Erlandson 177 (GH); airstrip, 24 Mar 1945, Johnston 538 (GH).
CHIRIQUI: Corotu, 6 m SW of airport of Puerto Armuelles, 17 Feb 1973,
Busey 399 (M0); Quebrada Manzanillo, 9 km S x SW of Puerto Armuelles,
3 Mar 1973, Busey 733 (M0).

409
19bc. Cl itori a jav 1 tens 1 s (H.B.K.) Benth. var. portobel1ensis
(Beurl.) Fantz f. truncata Fantz, f. nov.
Leaves obovate, glabrate below, trie homes along nerves subappressed
to spreading; leaflets 5-8 cm long, 4-5.5 cm wide, apex truncate to
truncate-retuse, base cuneate.
TYPE COLLECTION: PANAMA. CANAL ZONE: Barro Colorado Island,
Jan 1939, Brown 185 (H0L0TYPE: F 1004739).
This form is known only from the one collection. The leaf type is
unknown elsewhere in the subgenus Bractearia, and only from the Asian
species C. javanica (subgenus Neurocarpum) and from a couple of
leaflets of a t kaessneri specimen (subgenus Cl itoria) from central
Africa.
19c. Clitoria javiten sis (H.B.K.) Benth. var. longiloba Fantz,
var. nov.
Leaves glabrate to glabrous on lower surface, trichomes, when
present, sparse, subappressed, confined to nerves; leaflets typically
small or becoming broadened, 7-14 (19) cm long, 4.5-9 (12) cm wide.
Petioles short, 3-9 cm. Petiolule 4-6 mm, rarely 9-10 mm on larger
leaflets. Stipules deciduous, usually not observed, (4) 5-7 mm.
Inflorescences few-flowered, short, 0.5-2 cm. Pedicels 5-7 mm. Bracts
small, 3-4 mm. Bracteoles small, 3-4 (rarely 5) mm, subtending calyx
base and inserted 1-2 mm below it. Calyx lobes elongate, length nearly
twice the width, apex short-acuminate to subulate, (5) 6-8 (9) mm long.
Stipe unknown.

410
TYPE COLLECTION: PERU. Loreto: Stromgebiet des Maranon von
Iquitos aufwarts bis zur Santiago-Mundang am Pongo de Manseriche, ca
77°30'W, 1 924, Tessmann 3782 (HOLOTYPE: S. Isotypes: G-193 & 194,
herb. Delessert. Cotype: F 612407).
DISTRIBUTION (Figure 46): This varieity is an apparent endemic to
Loreto, Peru, collected in forests at elevations of 100-220 m.
PERU. LORETO: Iquitos and vie., 120 m, 12 Oct 1929, Williams
3691 (F); Mishuyacu, near Iquitos, 100 m, Oct-Nov 1929, Kl_ug 464 (F,
NY); Stromgebiet des Maranon von Iquitos, 1924, lessman 5347 (G);
Balsapuerto, 220 m, May 1 933, Klug 3044 (A,BM,F,G,GH,M0,NY,S); between
Yurimaguas & Balsapuerto, lower Rio Huallaga basin, 135-150 m, 26-31 Aug
1929, Killip & Smith 28104 (NY) and 28235 (NY).
I9d. Cl itori a javitensis (H.B.K.) Benth. var. grandifolia
(Ducke) Fantz, comb, nov.
Clitoria grandifolia Ducke, Arch. Jard. Bot. Rio de Jan.
5: 141. 1930.
Leaflets typically grandifoliate, large, typically (10) 15-28 cm
long, 6-15 cm wide. Petiole elongate, 10-25 cm. Petiolule long,
6-10 mm. Stipules long, 6-10 mm. Inflorescence elongate, 4-18 cm,
multi flowered. Pedicels typically 5-9 (11) mm. Bracts short, 3-4 (5)
mm. Bracteoles long, 4-6 mm. Calyx lobes elongate, (4) 5-8 mm,
ventral lobe 7-12 mm. Stipe short, 19-24 mm.
TYPE COLLECTION: BRAZIL. Amazonas: Tonantins, 8 Nov 1927,
Ducke 20399 (LECTOTYPE: RB-2 sheets).

411
Ducke cited three collections he had examined from the herbarium
at the Jardin Botánico de Rio de Janeiro. Each collection is mounted
on two sheets placed in separate folders. Of the three collections,
Ducke 20400 lacked the fruits which were described in the original
description of C. grandifolia. The Kuhlmann 1493 specimen lacked any
matured flowers. The remaining collection (Ducke 20399) has a
matured fruit, a bifurcate inflorescence of many flowers, and a packet
of fruit fragments, an intact flower, and a dissected flower. The best
specimen which matches the description published by Ducke, Ducke 20399,
and thus it is selected as the lectotype. The other two collections
cited by Ducke are selected as 1ectoparatypes for the variety.
DISTRIBUTION (Figure 46): This variety is distributed in the
Western Amazon Basin at elevations recorded as (based on Colombia
collections) ca 100-500 m.
V E N E Z U E L A. AMAZONAS: Rio Cunucununa, 9 Nov 1950,
Maguire, Cowan, & Wurdack 29412-A (NY,VEN); vie. base camp, Cerro
Sipapo (Paraque), 30 Dec 1 948, Maguire & Pol iti 28044 (BM,NY,US,VEN).
B R A Z I J_. camp de Dr. Vieiralves, Manaus, 20-0-56, Chagas 3350
(US); Benjamin Constant, alto Solimoes, 7 Sep 1962, Duarte 6936 (NY,
RB-2 sh.); Sao Paulo de Olivenca, 18 Oct 1927, Ducke 20400 (LECT0-
PARATYPE: RB-2 sh. I so1ectoparatypes: S,U)
C 0 L 0 M B I A. PUTUMAYO: del Rio Putumayo en Puerto Ospina,
230 m, 24 Nov 1940, Cuatrecasas 10779 (US). META: Monte, .25 km E
Puerto Lopez, 200 m, 5 Feb 1944, Hermann 1 1208 (NY,US); Ladorada
Marayal de San Martin, 16 km NW San Martin, 400 m, 3 Feb 1944, Hermann
11172 (F,GH,M0,NY-2 sh.,US); ca 20 km SE of Villavicencio, 500 m,

412
17 Mar 1939, Kill ip 34304 (US-2 s h. ); sabanas San Juan de Arana, Rio
Guejar, alrededores del aterrezaje "Las Micas," 500 m, 22 Jan 1951,
Idrobo & Schultes 1211 (GH,U); Sierra de la Macarena, Cano Entrada,
500 m, 13 Dec 1 949, Phi 1ipson, Idrobo, & Fernandez 1739 (BM,F,US):
Rio Guapaya, Sierra de la Macarena, 450 m, 7 Dec 1949, Philpson,
Idrobo, & Fernandez 1720 (BM,US).
PERU. Iquitos, 26 Feb 1924, Kuhlmann 1493 (LECTOPARATYPE: RB-2
sh. Isolectoparatype: U; non S = Canavalia dictyota).
19e: Clitoria javitensis (H.B.K.) Benth. var. klugii Fantz,
var. nov.
Leaflets typically small, 11-13 cm long, 6-10 cm wide. Petioles
elongate, 10-19 cm. Petiolules short, 4-6 mm. Stipules long, 5-7 mm;
inflorescence elongate, multi-flowered, 8.5-19 cm. Bracts small, 2-4 mm.
Pedicels 7-9 mm. Bracteoles 3-5 nun long, subtending the base. Calyx
lobes elongate, 5-7 (8) mm. Ovary pubescence white, tinged yellowish.
Stipe not observed. Tree (?).
TYPE COLLECTION: PERU. Loreto: tree 4 m tall, fls. dark lilac,
Mishuyacu, near Iquitos, forests, 100 m, Oct-Nov 1929, Klug 173
(H0L0TYPE: US 1455172. Isotypes: F 624041, NY).
DISTRIBUTION (Figure 46); This variety is known only from the
type locality.
PERU. LORETO: Mishuyacu, near Iquitos, 100 in, Oct-Nov 1929,
Klug 39 (PARATYPES: F 623967,NY).

413
20. Clitor id cavalcantei Fantz, sp. nov.
Liana. Branches 2-4 mm diameter, pith hollow, subquadrangular
becoming terete, pubescence of juvenile branches uncinate becoming
glabrate; bark dark brown, splitting initially longitudinally at angles,
light-colored beneath; axillary buds 2.5-3.5 mm, scales ovate, sparsely
strigose, inner scales ciliate. Leaves 3-foliate, coriaceous,
concolorous to slightly darker above, leaflets elliptic to elliptic-
oblong, rarely obovate, apex acuminate, acumen 1-2.5 cm long, more or
less mucronate, 2-6 mm wide, straight to arcuate, base broadly cuneate,
midrib subimpressed above, primary nerves of 7-9 pair, upper surface
glabrous, lower surface glabrous or rarely with a few subappressed
trichomes on nerves, lamina 10-17 cm long, 5-9 mm wide. Petiole dark
brown, subterete, 3-6 cm long; swollen base 6-7 mm, pubescence uncinate;
rachis 2-2.5 cm. Petiolules quadrangular-subterete, rugose, glabrate,
5-6 mm. Stipules deciduous, lanceolate-ovate, acute, aparsely pubescent,
5-6 mm long, 2 mm wide; stipels more or less persistent, linear, acute
glabrate, 3-6 mm long, 1 mm wide. Inflorescence axillary and
cauliflorous, few-flowered (4-8 flowers), racemose, more or less
nodose, solitary or paired at node, 0.5-1.5 cm long, axis pubescence
uncinate with moderately dense, falcate-spreading trichomes. Pedicels
5-8 mm, pilose. Bracts ovate, acute, strigose, more or less ciliolate,
middle pair 3-5 mm long, 1-1.5 mm wide, outer bract 2-3 mm long, 0.5 mm
wide. Bracteoles lanceolate-ovate, acute, 5-7 mm long, 1-2 mm wide,
subtending the base and inserted 1 (2) mm below it, pubescence moderately
dense, appressed, ciliolate. Flowers large, (5.5) 6-7 cm, rose-colored,
turning purplish. Calyx pubescence appressed, thinned, uncinate

414
trichomes not observed (typically found in other species), tube 16-20
mm long, 5-9 mm wide at base to 8-11 mm at throat, lobes deltoid-ovate,
acuminate, often broad-angled sinus (nearly 90°) between lateral and
dorsal teeth, 5-8 mm long, 3.5-5 mm wide, ventral lobe 7-10 mm.
Vexillum pubescence conspicuously uncinate toward apex and margins,
appressed trichomes moderate, more or less confined to veins, blade
3.5-4 cm wide, claw 14-18 mm. Alae extended beyond carina 6-11 mm,
blade 22-27 mm long, 4-7 mm wide, claw 17-22 mm. Carina falcate,
11-13 mm long, 4-5 mm wide, claw 27-29 mm. Stamens diadelphous,
vexillary stamen coherent near base, stamina 1 tube nearly straight,
39-46 cm long, free filaments 2-5 mm; anthers lanceolate, 1.6-2.4 mm
long, 0.5-0.7 mm wide. Gynophore black, suberect, 3-6 mm; pubescence
of uncinate with scattered white trichomes; ovary 12-16 mm long, 1 mm
wide; pubescence white, sericeous, hiding uncinate trichomes beneath;
style dark-colored, 24-32 mm long, geniculate 7-10 mm from distal end.
Legume unknown except in juvenile state (ca 9.5 x 0.6-0.7 cm) flat,
swollen around seeds; densely uncinate pubescent with scattered suberect
trichomes; stipe 19 mm. Seeds unknown, ca 9 per pod. Figure 48.
Cavalcante's Clitoria is characterized as a liana with glabrous
leaves and short inflorescences which bear large, reddish-purple flowers
from an appressed pubescent calyx, and later bearing uncinate pubescent
legumes.
PHENOLOGY: Flowers have been collected from mid-July to October
and mid-February to March 1. The only juvenile fruit seen was collected
in early March.
TYPE COLLECTION: BRAZIL. Para: Cipo lenhoso, flor na anteae,
rosea depois roxo, mata devastada, comum, Parque Ingigena do Tumucumaque,

Figure 48. Cl itoria cavalcantei. (a-b) leaflets, x 1; (c) inflores¬
cence, x 1; (dTFlower, x 1; (e) vexillum, x 1; (f) ala and
carina, x 1; (g) gynoecium, x 1; (h) androeciuin, x 1;
(i) calyx, x 1; (j) juvenile fruit, x 1. (Cavalcante
2570, NY: b-c,g-j. Spruce 1877, K: a,e-f. Ducke 23406,
S: d. )


417
Rio Paru de Oeste, Missao Tiriyo, arredores da Missao, 2°20'N-
55°45'W, 1 Mar 1970, Cava!cante 2570 (HOLOTYPE: NY. Isotypes:
F 1717517,S,U 256619B).
The specimen selected as the holotype is the only specimen which
has the juvenile fruit and its characteristic uncinate pubescence which
segregates this species from javitensis. The packet also contains
parts of at least two dissected flowers. Spruce 1877 has been cited
for C. javitensis, but the dissected flower with the characteriStic
white pubescent ovary and uncinate gynophore agree with Cavalcante 2570.
Spruce 1877 has larger flowers and leaves than the less mature holotype.
It has been selected as a paratype.
NOTES: Clitoria cavalcantei has close affinities with javitensis
which it superf icial ly resembles. However-, it differs from it by the
uncinate pubescent fruit, a style much longer than the ovary, and a
white pubescent ovary which bears uncinate trichomes beneath the
sericeous trichomes. These characteristics are common in most of the
species of section Cauliflorae found in the eastern part of northern
South America. However, these species have conspicuously uncinate
cálices, unlike C. cavalcantei which has appressed trichomes and lacks
any uncinate trie homes on the calyx.
Identification of C_. cavalcantei from javitensis var. javitensis
may cause some difficulty where specimens from the Amazon basin are
involved because the distribution of the two species overlap. Other
varieties of CL javitensis are excluded from the discussion because of
their elongated inflorescences or because of their range of distribution
is far removed from C. cavalcantei. Clitoria javitensis specimens from
the Amazon basin typically have minute bracteoles (2-4 mm long), more

418
elongated inflorescences (1-6 cm), short calyx lobes (4-6 mm) rufus
pubescent to tawny ovaries, styles subequaling the ovary, and vexillum
pubescence conspicuously appressed. The leaflets have more nerves
(9-12 primary pairs), and appressed trichomes below, at least along the
nerves, and shorter stipules (3-5 mm). Clitoria cavalcantei specimens
have longer bracteoles (5-7 mm), shorter inflorescences (0.5-1.5 cm),
longer calyx lobes (5-8 mm), ovaries with white pubescence and
gynophores with uncinate pubescence, styles much longer than the ovary,
and a more conspicuously uncinate pubescence on the vexillum with the
appressed trichomes congregated along the nerves and less dense. The
leaflets have fewer primary nerves (7-9 pair), a lack of pubescence, and
longer stipules (5-6 mm).
DISTRIBUTION (Figure 43): This species is found in the forests of
the Amazon Basin in Brazil. Elevations for collections have not been
reported.
BRAZIL. AMAZONAS: Rio Negro, Sapo, Oct 1851, Spruce 1877
(PARATYPE: K-407, hb^ Bentham); alto Rio Negro acimedo afflu.
Curicuriary, logar Tucaby, 19 Nov 1929, Ducke 23704 (RB); Manaos,
14 Jul 1 929, Ducke 2^3406 (G, K,RB ,S ,US): l.c., loco Cachseira Grande,
2 Feb 1946, Ducke 2029 (A,VEN). PARA: Santarem, 35 km da estrada do
Palhao, arredores do Acompomento do Igarape Curupira, 24 Aug 1969,
Silva & Sonza 2379 (F,NY,S,U).
21. Clitoria coriácea Schery, Fieldiana Bot. 28: 260. 1952.
Clitoria certifera Cowan, Mem. N.Y. Bot. Gard. 9: 349. 1957.
Clitoria emarginata Kill ip & Pittier, nom. in sched.

419
Virgate shrub, 1.5-2.5 m tall. Brandies 3-7 mm diameter, pith
hollow, internodes 2-10 cm, somewhat straight to slightly zigzag,
juvenile branches quadrangular, pubescence uncinate and strigose,
branches becoming subterete with uncinate pubescence to glabrate;
axillary buds 3 mm, scales ovate, acute, strigose; bark dark gray,
splitting infrequently in longitudinal slits, light-colored beneath.
Leaves 3-foliate, coriaceous, leaflets sometimes complicate, lanceolate,
ovate-lanceolate to oblong-lanceolate, apex short-acuminate, acumen to
0.5 (1)' cm long, 0.4-0.8 cm wide, more or less mucronate, base rotund
to weakly cordate, or subtruncate, margins conspicuously revolute,
midrib impressed above, primary nerves of 6-9 pairs, upper surface dark
green, glabrous, lower surface ceriferous, pale yellow-green to olive-
green, becoming green with age, lamina 5-13 cm long, (2) 2.5-5 (6) cm
wide. Petioles subquadrangular, moderately strigose, 1.5-6 cm; rachis
1-2.5 cm. Petiolules subquadrangular, dark-colored, rugose, 2-4 mm,
strigose pubescent to glabrate. Stipules persistent, lanceolate, acute
to short-acuminate, 3-6 mm long, 1.5-3 mm wide, pubescence appressed;
stipels linear-lanceolate, apex long-tapered to a point or short-
acuminate, rigid, 2-5 mm long, 1-1.5 mm wide. Inflorescence axillary
and at defoliated nodes, racemose, subsessile to 0.5 cm long, few-
flowered (2-6 flowers), axis pubescence uncinate and falcate, rufus.
Pedicels 5-9 mm, pubescence falcate, rufus, and uncinate, moderate to
dense. Bracts lanceolate, acuminate, pubescence subappressed, middle
pair persistent, 2.5-5 mm long, 1-2.5 mm wide, outer bract deciduous,
2-3.5 mm long, 1-2 mm wide. Bracteoles persistent, lanceolate, 5-8
(11) mm long, 1-2 mm wide, inserted 2-3 mm below the base, pubescence
uncinate and falcate, rufus, scattered. Flowers pale lilac to purple,

420
vex i 11 uni basally with fine red veins, 4-6 cm long. Calyx pubescence
conspicuously uncinate with moderately scattered, falcate, rufus
trichomes, tube 13-17 mm long, 4-6 mm wide at base to 9-11 mm wide at
throat, lobes deltoid-lanceolate, acuminate, with a broad-angled sinus
(nearly 90°) between the lateral and dorsal teeth, 7-12 mm long,
2.5-4 mm wide, ventral lobe narrow lanceolate to linear, 10-14 mm
long, 1-1.5 mm wide. Vexillum pubescence uncinate and sparsely pilose,
often the macrotrichomes confined to nerves, blade 2.5-4.5 cm wide,
claw 9-12 mm. Alae extended beyond carina 4-7 mm, white, blade oblong,
weakly falcate, 18-26 mm long, 5-9 mm wide, claw 13-19 mm, uncinate
pubescent. Carina white, falcate, blade 11-16 mm long, 4-6 mm wide,
claw 23-30 mm, with uncinate trichomes. Stamens diadelphous, vexillary
stamen coherent near base, tube 33-39 mm long, free filaments 2-4 mm;
anthers lanceolate, 1.5-2 mm long, 0.5-0.7 mm wide, connective acute.
Gynophore 5-8 mm; pilose pubescent; ovary 11-15 mm long, 1-1.3 mm
wide, pubescence dense, tawny-sericeous; style 23-27 mm, geniculate
7-10 mm from distal end; stigma flattened, subcapitate, 1 mm diameter;
ovules 15 (vide Cowan, 1957). Legume flat, upcurved at apex,
greenish-brown, pubescence conspicuously uncinate with scattered
suberect trichomes, base enclosed within calyx lobes, valves 6-12 cm
long, 9-15 mm wide; stipe enclosed within calyx tube, 16-20 mm long,
2-3 mm wide expanding to 4-5 mm at apex, pubescence dense, of uncinate
and suberect trichomes; beak 5-10 mm; dehiscence causing valves to twist
one to one and one-quarter of a turn. Seeds light to dark brown, smooth,
weakly lenticular, face suborbicular to slightly longer than broad,
6-8 nun long, 5-7.5 mm wide, 3 mm thick, 8-11 seeds per pod; hilum
oblong, 1.5 mm x 1 mm. Figure 49.

Figure 49. Cl i tori a coriácea, (a-c) leaflets, x 1; (d) inflorescence,
x 1; Je) flower, x 1; (f) calyx, x 1; (g) ala & carina,
x 1; (h) androecium, x 1; (i) gynoecium, x 1; (j) immature
fruit, x 1; (k) stipe, x 1; (1) dehisced fruit, x 1;
(m) three views of seed, x 1; (n) anther, x 4; (o) upper
style, x 3, (p) vexillum, x 1. (Maguire et al. 37635,
NY: c-i, n-p, Williams 15047, F: a. Maquire et al.
37580, VEN: b. Maguire et al. 41666, US: j-lJT

422

423
The wax bearing Clitoria is characterized as a virgate shrub having
leaves ceriferous below and with conspicuous revolute margins, subsessile
inflorescences of few, medium-sized, lilac-purplish flowers, and small
legumes.
PHENOLOGY: Flowers have been collected in February, April, and
May. The only collection of the fruit was made in early October.
TYPE COLLECTION: VENEZUELA. Amazonas: Arbusto trepador, flores
purpuras o de color de malva, en la Sabana Abierta, San Antonio, upper
Rio Orinoco, 121 m, 27 Apr 1 942, Llewelyn Will jams 1 5047 (HOLOTYPE:
F 1189036. Isotypes: US 1834849, VEN 25808).
Cowan's name (C. cerifera) employs an appropriate epithet for the
only wax-bearing species known, although Schery's name has priority
and is the correct name. Cowan was unaware of the publication by
Schery, as Cowan had identified the Smithsonian isotype of C. coriácea
using his name of C. cerifera. The Venezuelan isotype (Williams 15047,
VEN 25808) is the type specimen for the unpublished name, C. emarginata
Kill ip & Pittier. This specimen was annotated to "C. coriácea Schery,
n. sp. Tipo." The discovery of a published name for the specimen may
be the reason why Kill ip & Smith did not publish their name.
There are two types for the synonym C^_ cerifera. Cowan selected
the holotype specimen as Maguire, Wurdack, & Bunting 37580 (NY).
Maguire, Wurdack, & Bunting 37635 was selected as the paratype. Both
collections represent younger leaves than Schery's type. The type of
Cj_ coriácea has lost most of the wax from the lower leaf surface and
the leaflets have become darker in color.
NOTES: Schery correctly noted that the species had affinities to
C. javitensis, whereas Cowan placed the species near C. arborescens.

424
Clitori a coriácea differs from both specie^ by its subsessile
inflorescences, a style much longer than the ovary, smaller legumes,
and a conspicuously uncinate pubescent calyx. Leaves of Clitori a
coriácea are similar to these two species with the leaflet midrib
impressed above, and more similar to those of javitensis in shape.
However, coriácea is much more closely related to C^_ sagotii based
upon the subsessile inflorescence, fruit pubescence (predominately
uncinate), the appearance of the calyx (longer lobes and broader sinus
between'lateral and dorsal lobes), and the reproductive structures,
although it clearly differs vegetatively from C_^_ sagotii by the revolute
margins, ceriferous lower surface, and lack of a raised midrib.
The smaller leaflets, smaller fruits, and habitat are similar to
some members of the subgenus Neurocarpum. The leaflets and fruits of
C. coriácea are the smallest found in the subgenus Bractearia.
DISTRIBUTION (Figure 43): This species is found in the drier parts
of savannas in the Amazonas territory of Venezuela at elevations of
100-140 m.
V E N E Z U E L A. AMAZONAS: right bank Rio Pacimoni, 50 km above
mouth, 7 Feb 1954, Maguire, Wurdack, & Bunting 37580 (Type of synonym
C. cerifera: F,Gil,NY-2 sh., holotype, RB ,VEN); l.c., 100-140 m,
2 Oct 1 957 , Maguire, Wurdack, & Bunting 41666 (F,G,GH,NY,RB,S,US,VEN);
left bank Cano Hechimoni, 8 km above mouth, 100-130 m, 9 Feb 1954,
Maguire, Wurdack, & Bunting 37635 (F,NY); along Rio Cotua W of
campsite, Cerro Yapacana, between El Campamento base & Sabana Grande,
3°45IN-66°45‘W, 125 m, 7 May 1970, Steyermark & Bunting 103270 (VEN)
and 103272 (VEN).

425
22. CJjtoria tunuhiensis Fantz, sp. nov.
Trailing liana. Branch 2-3 mm diameter, pith solid, with
pubescence dense, puberulent, rufus, becoming whitened with age. Leaves
3-foliate, thick membranous, leaflets ovate to ovate-elliptic, apex
short-acuminate, acumen to 0.5 cm long, 3-6 mm wide, more or less
mucronate, base rotund, midrib subraised above, pilose, rufus,
moderately dense, primary nerves of 7-9 pair, upper surface dark green,
glabrous, lower surface reddish-green, fomentóse, trichomes rufus,
lamina 6.5-9.5 cm long, 4-5.5 cm wide. Petioles subterete, 2-3 cm
puberulent, rufus; rachis 1.5-2.5 cm. Petiolules subquadrangular, 4-5
mm, puberulent, rufus. Stipules lanceolate, acute, 4-5 mm long, 2 mm
wide; pubescence ciliate and strigose, densest towards apex, trichomes
rufus; stipels linear, acute, 3-4 mm long, 0.5-1 mm wide, pubescence
uncinate. Inflorescence axillary, solitary, racemose, few-flowered
(2-4 flowered), subsessile to 0.8 cm long, axis pubescence dense,
puberulent, trichomes rufus. Pedicels 5-7 mm. Bracts lanceolate,
acute, 5-7 mm long, 1.5-2 nun wide, pubescence uncinate, pilose, more
or less ciliate. Bracteoles lanceolate, acuminate, 13-14 mm long,
2.5-3 mm wide, inserted 2-3 mm below the calyx base, pubescence densely
uncinate and scattered-pilose, ciliate. Flowers medium-sized, ca 4.5 cm
(s 1 i gilt 1 y immature), color unknown. Calyx pubescence moderately dense,
conspicuously pilose and uncinate, tube 15-17 mm long, dark-colored
(violet-tinged ?), 4-5 mm wide at base to 7-9 mm wide at the throat,
lobes lanceolate, acuminate, 9-11 mm long, 3-4 mm wide, ventral lobe
linear, 10-12 mm long, 1 mm wide, arcuate towards apex. Vexillum
pubescence appressed, tawny, blade ca 2.5 cm wide, claw 6 mm. Alae

426
extended beyond carina 4 mm, blade spatulate, 20 mm long, 4-6 mm wide,
claw 13 mm. Carina falcate, 12 mm long, 4 mm wide, claw 21 mm.
Stamens diadelphous, vexillary stamen coherent near base, tube
glabrous, 28 mm long, incurved abruptly at apex ca 2 mm, free filaments
2-4 mm; anthers lanceolate, 2 mm long, 0.7-0.8 mm wide, connective
acute. Gynophore 4-5 mm; pubescence near apex yellowish-white; ovary
10 mm long, 1.2 nun wide; pubescence white, tinged yellowish; style 21
mm long, geniculate 5-6 mm from distal end, with a dense beard,
sericeous towards base; stigma flattened, subcapitate, puberulent
around base, ca 0.75 mm diameter. Legume unknown. Figure 50.
The Tunuhy Clitoria is characterized as a woody vine with ovate
leaflets, revolute margins, and tomentose below, and bearing short,
subsessile inflorescences of medium-sized flowers that have a
pilose-uncinate pubescent calyx with elongate lobes and bracteoles.
TYPE COLLECTION: BRAZIL. Amazonas: Vine or bare (?), Rio Negro,
Icana, Serra de Tunuhy, 5 May 1947, Fróes 22274 (H0L0TYPE: U 95456A).
PHENOLOGY: The only collection of flowers was made in early May.
NOTES: Clitoria tunuhiensis has close affinities with C. coriácea,
both of which have similar appearances to the flowers, inflorescence,
and leaves. Both species have revolute leaf margins, an uncommon
characteristic in the genus. Clitoria tunuhiensis is easily dis¬
tinguished from C. coriácea by the tomentose pubescence and the lack of
wax on the lower surface of the leaves, the longer bracteoles, the
shorter staminal tube and a denser pubescence of the vexillum.
Clitoria tunuhiensis also appears similar to C^ arborescens which
differs by the elongate inflorescence, appressed pubescent calyx, and
a style subequaling the ovary length.

Figure 50.
Cl i tori a tunuhiensis. (a) leaflet, x 1;
x IHcT flower, x 1; (d) calyx, x 1; (e'
(f) ala and carina, x 1; (g) androecium,
x 1; (i) style apex and stigma, x 8; (j)
(Froes 22274, U 95456A).
(b) inflorescence,
vexi11 urn, x 1;
x 1; (h) gynoecium,
anther, x 4.

1 cm

429
DISTRIBUTION (Figure 43): This species is known only from the
type col lection.
23. Cl i tori a sagotii Fantz, sp. noy.
Clitoria jav i tens is (H.B.K.) Benth. (p.p. sensu Bentham),
Journ. Linn. Soc. 2: 42. 1858; non Neurocarpum
javitense H.B.K., Nov. Gen. Sp. 6: 409. 1823.
Clitoria javitensis (H.B.K.) Benth. var. glabra_ Sagot, Ann.
Sc. Nat. Ser. 6, 1_3: 299. 1882.
Clitoria javitensis (H.B.K.) Benth. var. guianensis Sagot,
noiiu in sched.
Clitoria prostrata Spruce, nom. in sched.
liana, climbing or occasionally trailing. Branches 3-17 mm
diameter, pith solid, juvenile branches subquadrangular, with pubescence
uncinate, branches becoming terete, glabrate; bark brownish on juvenile
branches becoming grayish, nearly smooth to rough texture, splitting
in longitudinal strips. Leaves 3-foliate, coriaceous, concolorous
(green) to slightly light green below, leaflets basically elliptic-
oblong or elliptic to lanceolate, narrow to very broad, apex nearly
obtuse with moderately abrupt, acuminate tip, acumen 1-2 cm long, 2-7
mm wide, more or less mucronate, base cuneate to broadly cuneate, midrib
conspicuously raised above, primary nerves of 6-11 pair, upper and lower
surfaces glabrous, lamina (6.5) 8-21 (25) cm long, 3-8 (13) cm wide.
Petiole subquadrangular-terete, adaxially truncate to caniculate, 1.5-
8.5 cm, pubescence uncinate to glabrate; rachis 1.5-4 cm. Petiolule
quadrangular, dark-colored, rugose, 4-9 mm, pubescence sparse, uncinate

430
with scattered white trichomes, to glabrate. Stipules deciduous,
deltoid-lanceolate, acute, 4-5 (6) mm long, 1.5-2 mm wide; pubescence
uncinate and more or less appressed; stipels somewhat persistent, linear
to lanceolate, acute, 1-6 mm long, 1-1.5 mm wide, pubescence of uncinate
and more or less appressed trichomes. Inflorescence axillary, solitary,
or cauliflorous, fascicled, racemose, few-rlowered (2-4 flowered, rarely
6-flowered), subsessile to 0.5 mm long; axis pubescence uncinate.
Pedicels 6-11 mm, uncinate pubescent; pedicels of cauliflorous
inflorescences compressed, flattened below the bracteoles, 2-3 mm wide.
Bracts ovate, obtuse, pubescence uncinate and more or less minutely
ciliolate; inner pair caducous; middle pair persistent, 3-5 mm long,
2 mm wide; outer pair deciduous, 1.5-3 mm long, 1-1.5 mm wide.
Bracteoles linear-lanceolate, acute to short-acuminate, often spreading,
4-11 mm long, 1-2 mm wide, inserted 2-5 mm below the calyx base,
uncinate pubescent. Flowers odorous, 5-8 cm, vexillum pinkish-mauve
with darker-colored veins near center, alae and carina white. Calyx
chartaceous, pubescence conspicuously uncinate (vide 20 x) with scattered
subappressed to spreading trichomes, tube 11-16 mm long, 4-7 mm wide at
base to 8-13 mm wide at the throat, lobes deltoid-ovate to lanceolate,
acuminate, more or less arcuate, 5-13 mm long, 2-6 mm wide at base,
ventral lobe 8-14 mm long, 1-2 mm wide. Vexillum pubescence uncinate
with scattered appressed trichomes, blade 3-4.5 cm wide, claw 6-9 mm.
Alae extended well beyond carina by 8-13 mm, blade spatulate-falcate,
18-23 mm long, 8-12 mm wide, claw 17-26 mm. Carina nearly straight,
uncinate pubescent, blade 11-18 mm long, 3-6 mm wide, claw 25-35 mm.
Stamens diadelphous, vexillary stamen coherent at the very base, tube
glabrous, arcuate-falcate, 32-39 mm long, free filaments 4-7 mm; anthers

431
lanceolate, 1.5-2.1 mm long, 0.7-0.8 mm wide, connective acute, subequal
to slightly shorter than pollen sacs. Gynophore 2-4 mm; ovary 10-15 mm
long, 1-1.3 mm wide; pubescence white, tinged yellowish; style dark-
colored, 22-26 mm, geniculate 8-10 mm from distal end; stigma dark-
colored, capitate, 0.6-0.8 mm diameter. Legume (known from var.
caniculata only) exerted above calyx, flat, pubescence more prominent
along sutures, moderate, of uncinate trichomes with few, scattered,
suberect, white trichomes, valves 8.5-15 cm long, 15-18 mm wide; stipe
straight to weakly arcuate, 1.5 mm thick expanding to 3-4 mm at apex,
27-35 mm long, pubescence uncinate and pilose; beak when persistent,
10-15 mm; dehiscence causing valves to twist 1-2 turns. Seeds (observed
in damaged condition) dark brown to black, smooth, thick lenticular to
subglobular, 6-7 mm long, 6-8 mm wide, 4-5 mm thick, 4-8 seeds per pod.
Figures 51 and 52.
Sagot's Clitoria is characterized as a liana with glabrous leaves
and subsessile inflorescences bearing large, pinkish-mauve flowers from
an uncinate pubescent calyx, or bearing uncinate pubescent legumes
slightly exerted beyond the calyx.
PHENOLOGY: Flowers are collected from July through March with a
peak during November to February. Fruits have been collected from
November to February, and in July.
TYPE COLLECTION: FRENCH GUIANA. Frutex alte scandens, flores
rosei, suaveolentes, plurimi e ligno prodientes (?), fructum non vidi,
Karouany, 1857, Sagot 120 (H0L0TYPE: K-26, hb^ Benthu; S= photo of K.
Isotypes: GH,NY,S,U 37632A.W).
Sagot 120 is also the type for two other names, C. javitensis
(H.B.K.) Benth. var. guianensis Sagot, nom. in sched., and

Figure 51. Clitoria sagotii - I. Var. sagotii• (a) leaflet, x 1; (b)
inflorescence, x 1; (c) flower, x 1; (d) calyx, x 1;
(e) vex i 11 inn, x 1; (f) ala and carina, x 1; (g) androecium,
x 1; (h) gynoecium, x 1; (i) anther, x 4; (j) petiole and
rachis, x 1; (k) cross-section of petiole, x 2. (Sagot
120, K-hb. Bent ham: a-c,j-k. Cowan 38706, F 1446355:
d-i.)

433

Figure 52.
Cl i tori a sagotij^ - II. Var. caniculata: (a) leaflet,
x 11 TbT caniculate petiole, x 1; (cT~x-section petiole,
x 2; (d) flower, x 1; (e) vexillum, x 1; (f-g) legume,
x 1; (h) three views of seed, x 1. Var. sprucei:
(i) leaflet, x 1; (j) flower, x 1; (k) petiole and rachis,
x 1; (1) x-section petiole, x 2. (Forest Dept. Br.
Guiana 3933, K-413: a-c; NY 2689: f,h. Cruz 1642,
GH: g. Lang & Persaud 67, F 559005: d e. Spruce 3543,
K-433: i -17) '

435

436
C. javitensis (H.B.K.) Benth. var. glabra Sagot (1882). Sagot annotated
the specimens with one varietal name, and later published the variety
under a second name. He did not select a holotype. In recognizing this
group as a species, Sagot's published varietal name was rejected as a
specific epithet because of the possible confusion with the name
C. glaberrima. Sagot's unpublished name was also rejected because it
would become a homonym of C. guianensis (Aublet) Benth. Each specimen
of this collection examined is nearly equivalent in the vegetative
material and flowers it possesses. The Kew specimen is the only one
which has the complete label data, part of which was cited by Sagot.
The Kew specimen is selected as holotype for the new species.
VERNACULAR NAMES: SURINAM (MAROWIJNE): Kwatta Kamra, Jonker-Verhoef
& Jonker 359 (mixed collection with fruits non-Clitoria). GUYANA:
This species is referred to as a "bush rope" by several collectors
(i.e. Sand with 599, fillet & Tiljet 45273 , Jenman 4930), a term not used
for other Clitoria species. It is not clear as to whether this name is
a vernacular name, or a local expression used for a woody vine.
NOTES: This species includes those specimens that typically go by
the name C. javitensis var. glabra Sagot. This group of plants has been
excluded from the broad concept of C. javitensis as observed by Bentham
in 1858 (cf. C. javitensis, pp. 379-384). Clitoria sagoti can be
easily segregated from C. javitensis by the prominent uncinate
pubescence on the calyx, legume, and various plant axes, the raised
midrib on the upper leaf surface, subsessile inflorescences, short
vexillum claw, and smaller fruits. Table 8 (p. 382) compares additional
characteristics between the two species.

437
This species consists of three distinct varieties, nearly allopatric
in distribution. The typical variety is conspicuously different from
the other two varieties by the larger size of the leaflets, flowers, and
caly/ lobes. A second variety (var. caniculata) is easily recognized by
the smaller bracteoles and petiolules, and the caniculate petioles.
This is the only variety in which fruits have been collected. However,
often in the Guiarías, mixed collections have been made which contained
a mixture of non-Clitoria fruits and Clitoria flowers. The third
variety' (var. sprucei) is widely separated geographically from the other
two varieties, and superficially appears similar to C. cavalcantei
which has a sympatric distribution. Clitoria cavalcantei can be
distinguished from this variety by the impressed midrib on the upper
surface of the leaflet, the lack of any uncinate pubescence on the calyx,
and ty the longer inflorescence, vexillum claw, and calyx tube.
Several of these specimens have been discussed in the literature as
noted under the discussion of javitensis (cf. pp. 379-384).
Additional specimens often bear annotations by Sandwith, who treated
all of them as one species under the name of C. javitensis and by N. E.
Brown, who treated most of them as separate species, although Brown
never gave his species a name. Brown's concept of the groups is best
summarized by the annotation note on the sheet of Gleason 556 (NY):
5
c.j. Bth. in J. L. Soc. 2: 42 was founded upon Spruce
1877/2320 and Schomburgk 1000/1723 which are certainly in my
opinion all 3 perfectly distinct species and neither one of
5. c.j.Bth. = Clitoria javitensis (H.B.K.) Benth., Journ. Linn.
Soc. 2: 42. 1858. Bentham broaden the concept of the species when
he tiansferred it from Neurocarpum to Clitori a.

438
them can be possibly the same as N. jav. H.B.K., on which
the specific name is founded for that plant is stated to have
pub. leaves, while those quoted by Bentham are all perfectly
plab. beneath. Other parts of description are also at
variance.
The ptesent treatment nearly agrees with Brown that most of the cited
collections by Bentham are distinct from C. javitens is and should be
treated in a separate species. Spruce 232u is a mixed collection of
which one part belongs to javitensis. Spruce 1877 is placed in a
new species, C. cavalcantei. The remaining specimens along with Sagot
120, cited by Sandwith (1931) in his discussion of this group, represent
three distinct groups which are not distinguished enough from each other
to be treated as separate species. Sagot 120 represents the typical
variety of a new species, C. sagotii. The two Schomburgk collections
represent var. caniculata. The second part of the mixed collection of
Spruce 2320 represents var. sprucei.
Bentham (1858, p. 43) cited the two Schomburgk collections as
"Rob. Schomburgk 1000" and "Rich. Schomburgk 1723." The Schomburgk 1000
collections (BM,CGE) bear the printed or written label data as
"Roraima, Br. Guiana, 1842-3." The Kew specimen (Herb. Benthamian,
K-409) bears a printed label "Schomburgk, Br. Guiana, 1844" with the
number 1000/1723. The incorrect date and lack of locality (i.e.
Roraima) suggest that this specimen is not Schomburgk 1000. Is the
collection then Schomburgk 1723? Or is it a mixed collection (there
are two branches mounted on the sheet) of both Schomburgk collections?
Or is there another explanation for both numbers on the Kew specimen?
6. N. jav. HBK. = Neurocarpum javitense H.B.K., Nov. Gen. Sp.
6: 409. 1823.

439
the Guia ñas a í¡r the western Amazon B<: ü i n It is typical ly found as a
climber in forests or thickets, or cccasi.;» 11 y as a treiiiig vira 'r>
more open are s, and notably associated with sat-uy soil, Elevati ns ere
rarely given by the co1 lectors.
KEY TO VARIETIES:
1. Calyx lobes 8-13 mu; ova>y 13-15 mm; flowers i .rg?, b-3 cm; leaves
typically arger, leaflets 1-21 ;25) cm long 5-8 cm wide; petiole
longer, 5-8.6 cm (French Guiana) 23a. Vuo sngoti i
1. Calyx lobe 5-8 mm; ovary 10-12 run, flowers medium-large, 5-6.5 cm,
ra¡ .j 7-8 cm in an individual flower; "eaves typically smaller.,
leaflets (6.5) 3-15 (18) cm long, (2.5) 3-/ cm wide; petiole
shorter , 1.5 - 5.5 í7) cm.
2. Pelmole me. or less caniculate adaxially, t. acteoles 4-6 (7;
mm, insey,vv d 2 . irsn below the calyx, petiolules typically
4-6 (7) mm; prn,. nerves of 1 aflec of G-8 (9) pair (Guyai
Surinam)
23b. var canicular
2. Petiole more o'" less flattened daxially, lacking grooves;
b^acteoles (G) 7-10 mm, inserted , ¿ mm oelow calyx base;
petiolules typically 5-9 mm; primary nerves cf leaflet of
8 10 oairs (W stern Amazon Basin: SE Colombia, S Venezuela,
NW Brazil)
23c. var. spruce
-3a. Clitoria sagotii Fantz var. s:gotii
Clitoria jay i tens is (H.8.K.) Benth. var. glabra Sage .,
Ann. Sc. Nat. Ser. c, 13 29'u 1832.

Figure 53. South American distribution of three species of section
Caul iflorae. Clitoria sagotii var. caniculata (#), var.
sagotii (Á.), var. sprucei (Qj); C. kaieteurensis (☆);
C. 1 eptostachya var. 1 eptostachya var. fruticosa

441

442
Clitoria javiten sis (H.B.K.) Benth. jar. guianensis Sagot,
nom. in sched.
Primary nerves of leaflet 8-11 pair, leaflets (12) 15-21 (25) cm
long, 5-8 (13) cm wide. Petiole 5-8.5 cm long, flattened adaxially,
without grooves. Petiolules 6-9 mm. Flowers large, 6-8 cm. Calyx
lobes elongate, 8-13 mm. Bracteoles 6-10 mm. Ovary elongate, 13-15 mm.
The typical variety is easily recognized by the elongated calyx
lobes and more robust appearance.
PHENOLOGY: Flowers have been collected from December through
February.
DISTRIBUTION: This variety was collected from mountain forests of
French Guiana with only a single elevation recorded as 275 m.
F R E N C H GUIA N A. Rive gauche de la riviere Camopi upon
summit montagne Alikene, 17 Feb 1968, Olderman & Sastre 299 (NY,U); vie.
camp montagne de Kaw, 275 m, 11 Dec 1955, Cowan 38706 (F,NY,RB).
23b. Cl itoria sa^otii Fantz var. caniculata_ Fantz, var^ nov.
Primary nerves of leaflet of 6-8 pairs, leaflets (6.5) 8-16 (20) cm
long, (2.5) 3-7 cm wide. Petiole 1.5-5.5 (7) cm, with a shallow to
deep longitudinal groove adaxially. Petiole 1.5-5.5 (7) cm, with a
shallow to deep longitudinal groove adaxially. Petiolules 4-6 (rarely
7-9) mm. Flowers 5-6.5 (rarely 4-8) cm. Calyx lobes 5-8 mm. Bracteoles
4-6 (7) mm. Ovary 10-12 mm.
This variety is easily recognized by its smaller bracteoles and
caniculate petioles.

443
TYPE COLLECTION: GUYANA. Vine with woody stems either procumbent
in the open or twining up low bushes in secondary wh.[ite] sand forest,
sandhills, Demerara River, 22 Feb 1943, field no. FI 197, Forest Dept.
British Guiana 3933 (HOLOTY-PE: K-413 sheet marked no. 2! Isotypes:
K, sheet no. 1-not seen, NY!).
This collection is only one of three (including Cruz 1642 & Lanjou
1835) that possessed fruits, and the only one also possessing flowers,
in addition to the characteriStic caniculate petioles, shorter bracteoles,
and smaller calyx lobes found on other collections. The Kew specimen is
made the holotype because of dehisced fruits exposing the seeds.
PHENOLOGY: Flowers have been collected nearly year round, from
July through March. Fruits were collected from November through
February and once in July.
DISTRIBUTION (Figure 53): This variety is found in dense upland
woods and thickets or in open areas trailing along railroads, or it is
commonly collected in Surinam savannas. It grows on sandy soils of
Surinam and Guyana. Elevations usually are not reported by the
col 1ectors.
S_ U R I NAM, witzand savanne bij de nieuwe weg naar Hanover,
12 Feb 1961, Kramer, Hekking, & Tryon 2870 (U-2 sh.). MAROWIJNE:
savannebos by Albina weg naar Moengo, 23 Jan 1956, Jonker-Verhoef &
Jonker 35(9 (U-mixed). SURINAME: Brownsweg, 29 Jun 1924, Bonteweren
6549 (U); Jodensavanne, Mapore kreek area, Suriname R., 6 Oct 1953,
Lindeman 4888 (U). SARAMACCA: naar hot Wilhelinina gebergte, Boven Gran
Rio, 17 Mar 26, Anonymous 271 (U-2 sh.); Tibiti savanna, 15 Jan 1949,
Lanjouw & Lindeman 1835 (U-mixed); l.c., 11 Jan 1949, Lanjouw &
Li ndeman 17711 (NY,U).

444
ü U Y A N A. Butukari, 20-21 Jul 1921, Gleason 714 (GH,NY);
BERB1CE: between Demerra & Berbice R., 5 50'N, 15-19 Jul 1922, Cruz
2642 (F,GH,M0,NY). DEMERARA: Demara R., Nov 1888, Jenman 4930 (K).
ESSE0UIB0: Adaro R. mouth, Kukui R., 500 m, 50 Sep 1 960, Ti 11et &
Til let 45273 (NY-mixed,U); mile 20, Bartica-Potaro Road, 24 Dec 1943,
Atkinson 23 (BM); Kabakaburi, Pomeroon Dist., 10-15 Feb 1923, Cruz 3332
(GH,NY,US); Rockstone, 15 Jul-1 Aug 1921, G1eason 556 (NY,US) and 543
(GH.NY); 1.c., 31 Dec 1919-1 Jan 1 920, Hitchcock 12309 (GH,NY,S,US);
Kamakusa, 9 Nov 1922, Langa Persaud 67 (F); l.c., 15 Dec 1922,Lan¿&
Persaud 358 (F); Essequibo R., Moraballi Creek near Bartica, 12 Nov 1929,
Sandwith 199 (K,NY,RB); Roraima, 1842-3, Rob. Schomburgk 1000 (BM,CGE,
W); 1044, Schomburgk 1000/1723 (K).
V E N E Z U E L A. BOLIVAR: Rio Cuyuni, Feb 1949, Cardona 2796a
(VEN).
23c. C1itoria sagotii Fantz var. sprucei Fantz, var. nov.
C1itoria pros trata Spruce, nom. in sched.
Primary nerves of leaflet of 8-11 pairs, leaflets 7-18 cm long,
3-7 cm wide. Petiole 2-7 cm long, flattened adaxially, without grooves.
Petiolules 7-8 mm. Flowers 5-6.5 cm. Calyx lobes (5) 6-8 mm.
Bracteoles (6) 7-10 mm long, subtending and inserted 1 (2) mm below
calyx base. Ovary 10-12 mm.
TYPE COLLECTION: VENEZUELA [?]. Amazonas: AC. Tomo fl. Guaimie,
ubi in graminosis sylvanum humiliorum gregarie viget, 18 Aug 1854,
Spruce 3543 (H0L0TYPE: K-433, Hb. Benth.). PARATYPES: Adflum
Guainia v Rio Negro supra ostium fluv. Casiquiari, 1854, Spruce 3543

445
(G 195 & 196, GH, CGE-hb. Lindley, F 1546829, U 112030, RB 17245).
Prope San Carlos, ad Rio Negro Brasilia borealis, 1853-4, Spruce 3543
(BM,K-431, hb. Benth., W 18666. Isopara types : K-410, hb. Hooker,
mixed; NY-Cambridge Univ. Hb., mixed).
Spruce 3543 has the best flowering material and some dissected
flowers. This number includes three sets of data with all plants
collected within a small geographic area. The holotype specimen bears
Spruce's name C. prostrata, the probable original collection of the
three sets placed under one number by Spruce. The other two sets are
designated as paratypes, with two exceptions. A sheet at Kew (Hb.
(looker) and one sheet from New York (Hb. Cambridge Univ.) contain mixed
material, and thus are designated as duplicates of the paratype, or
i sopara types, and are considered as non types in the determination of
the variety.
PHENOLOGY: Flowers have been collected in August, November, and
late April through early June.
DISTRIBUTION (Figure 53): This variety is widely distributed in the
western Amazon Basin in woods or open areas.
V E N E Z U E L A. AMAZONAS: Sao Gabri1-caatingas, May 1852,
Spruce 2320 -S'7 (K-mixed); San Antonio, 121 m, Alto Orinoco, en selva
de Tierra firme, 27 Apr 1 942, Williams 1 5043 (F,VEN).
B R A Z I L. AMAZONAS: Rio Cauabury between mouth or Rio la and
Rio Maturacá, 3-7 Nov 1930, Holt & Blake 439 (NY); basin Rio Negro,
Cajuri-Miki, 31 Jan 1 942, Froes 1 2447/1 91 (A); 201 km Manaus-Itacoatiara
Rd., Rio Urubu, 10 Jun 1968, Prance et al. 5122 (NY).
7. Compare with footnote 4 (p. 379) tor an explanation of the
letter B.

446
C O L O M B I A. AMAZONAS: Rio Caqueta, La Pedrera and vie.,
2 May 1952, Schultes & Cabrera 16323 (NY).
24. Clitori a kaieteurensis Fantz, sp. nov.
Liana, trailing along ground or climber to 6-10 ft high. Branches
2-6 mm diameter, pith solid and sometimes dark-colored, juvenile
branches subquadrangular, pubescence scattered uncinate and ca 0.5-1.3
mm long, appressed, rufus trichomes, branches becoming nearly terete,
glabrate; bark dark-brown becoming dark grayish-brown, splitting in
longitudinal strips; axillary buds 2-3 mm, scales ovate, acute,
pubescence dense, appressed. Leavers 3-foliate, coriaceous, leaflets
oblong-elliptic, elliptic, ovate-elliptic or ovate to occasionally
suborbicular, apex obtuse to broadly obtuse, abruptly acuminate, acumen
2-8 mm long, 2-10 mm wide at base, more or less mucronate, base rotund
to weakly cordate, midrib impressed above, primary nerves of 8-10
pairs, upper surface dark green, glabrous, lower surface dull green,
glaucous, pubescence spreading to suberect, moderately dense, becoming
widely scattered with age, lamina 8-15 cm long, 4-9 cm wide. Petiole
reddish-brown, subquadrangular to nearly terete, weakly longitudinally
striated, occasionally with a shallow groove adaxially near the apex,
2-7.5 cm; pubescence scattered, uncinate and subappressed, rufus,
becoming glabrate; rachis somewhat more compressed laterally, 1-3 cm.
Petiolules quadrangular, dark-colored, rugose, 5-6 mm, pubescence
appressed, rufus. Stipules deciduous, large and conspicuous, lanceolate
to oblong-lanceolate, broadly acute, sometimes weakly arcuate, 8-19 mm
long, 4-6 mm wide, pubescence sparse, appressed, ciliolate; stipels

447
1 inear-lanceolate to lanceolate, acute, pubescence lateral, stipels
very conspicuous, weakly 5-nerved, 8-13 mm lone;, 2-3 mm wide, terminal
stipels smaller, 4-9 mm long, 1-2 mm wide. Inflorescence axillary and
cauliflorous, 1-4 peduncles per node, dark-colored, subsessile to 0.5
cm long, racemose, few-flowered (1-2, occasionally 3 pairs of flowers),
axis pubescence of uncinate trichomes and a few spreading to erect
trichomes. Pedicels 4-6 mm, pubescence predominately uncinate. Bracts
variable; middle pair conspicuous, ovate-lanceolate, acute, 5-8 mm
long, 2-3 mm wide, pubescence scattered, uncinate; outer bract hidden,
ovate, broadly acute, 1.5-2 mm long, 1.5-2 mm wide. Bracteoles large,
lanceolate, acute to subacuminate, 9-12 (15) mm long, 3-5 mm wide,
inserted 1.5-2 mm below the calyx, pubescence uncinate, scattered.
Flowery large, 6-7.5 cm, pale purple. Calyx pubescence moderately dense,
predominately uncinate with a few, scattered, subappressed trichomes,
tubes 14-18 mm long, 5-8 mm wide at base to 9-13 mm wide at throat,
lobes ovate, acuminate, apex rapidly narrowed to a triangular acumen
one-half the lobe length, 7-11 cm long, nearly one-half the tube length,
3-5 mm wide at base, ventral lobe 9-13 mm long, 1-1.5 mm wide. Vexillum
pubescence moderately dense, uncinate, with occasional appressed
trichomes, blade abruptly curved outward above claw, 3.4-4.5 cm wide,
claw 6-9 mm. Alae extended well beyond carina by 8-11 mm, blade
subfalcate, 17-21 mm long, 7-11 mm wide, claw 19-26 mm. Carina arcuate,
15-18 mm long, 4-5 mm wide, claw 27-33 mm. Stamens diadelphous,
vexillary stamen coherent near base, tube glabrous, falcate, 39-43 mm
long, free filaments 4-6 mm; anthers lanceolate, ca 1.7 mm long, 0.5 mm
wide. Gynoecium falcate; gynophore 5-8 mm; ovary 13-17 mm long, 1.2-
1.3 mm wide, pubescence white, tinged yellowish, style 23-27 mm,

448
geniculate 2-4 mm from discal end; stigma subcapitate, flattened.
Legume stipitate, base enclosed within calyx lobes, flat, brown,
pubescence of uncinate and scattered, rufus nearly erect trichomes,
valves 8-13 cm long, 10-16 mm wide, beak when persistent, 5-7 mm;
stipe 18-22 mm long, 2-3 mm thick expanding to 5 mm at apex; dehiscence
cuasing valves to twist one-half to one turn. Seeds brown, smooth,
subglobose, globose, face nearly orbicular, slightly compressed, 8-12
mm long, 7-11 mm wide, 4-5 mm thick, 7-9 seeds per pod; hilum 2 x 1 mm.
Figure 54.
The Kaieteur Clitoria is characterized as a trailing to scrambling
liana with large, conspicuous stipules, stipels, and bracteoles,
glaucous leaves, and a subsessile inflorescence bearing large purplish
flowers with a calyx having elongate lobes. Tie long and broad bracteoles
and stipules easily distinguish this species from all of its near
relatives in section Cauliflorae. Two other species within this section
(C^ plumosa and C. obidensis) have conspicuously large stipules and
bracteoles, but they are highly pubescent plants with large calyx
tubes (2-3 cm) and lobes (1.3-2.7 cm).
PHENOLOGY: Flowers have been collected from September, October,
February, and May. Fruits have been collected in February and May.
TYPE COLLECTION: GUYANA. Essequibo: locally frequent, scrambler,
stem woody, leaves glaucescent beneath, fls. pale purple, cauliflorous,
fr. greenish-brown, Kaieteur Savanna, 4 May 1944, Maguire & Fanshawe
23202 (H0L0TYPE: NY 3107. Isotypes: A, F 1280928, K-412, MO 1371651,
U 72863A, US 1950981, VEN 28042).
The New York specimen is chosen as the holotype because it is the
best representative specimen, which includes leaves, an inflorescence

Figure 54. Clitoria kaieteurensis. (a-b) leaflets, x 1; (c) stipule,
x 1; TdT stipels, x 1; (e) inflorescence, x 1; (f) flower,
x 1; (g) calyx, x 1; (h) vexillum, x 1; (i) ala and carina,
x 1; (j) androecium, x 1; (k) gynoecium, x 1; (1) bracteole,
x 1; (m) legume, x 1; (n) three views of seed, x 1.
(Maguire & Fanshawe 23203, NY 3107: a,f-n; K-412:
A: b.).
c-e;

*
0S17

451
with flowers, a fruit still attached to its stipe, seeds (partially
destroyed), and a dissected flower (found within the packet).
NOTES: The collections of this species have traditionally been
identified as C. javitensis or as javitensis var. glabra. Clitoria
kaieteurensis is easily distinguished from javiten sis by the smaller
fruits, uncinate pubescent calyx, white pubescent ovary, elongated
style much longer than the ovary, and the subsessile inflorescence.
Clitoria kaieteurensis has its affinities with CL sagotii, which
includes those plants previously known by the name of C^ javitensis var.
glabra. Cl itoria sagotii is distinguished from CL kaieteurensis by the
glabrous, non-glacuous leaves, smaller stipules and stipels, the raised
midrib on the upper leaflet surface, smaller bracts and narrower
bracteoles, a shorter androecium, and longer-stipitate, broader fruits
with smaller seeds.
DISTRIBUTION (Figure 53): This species is an endemic that is
locally abundant in the Kaieteur Savanna of Guyana, collected only once
outside this region. Plants grow in sandy soil at elevations reported
as near 500 m.
G U Y A N A. ESSEQUIBO: Kaieteur Savanna, Sep-Oct 1881, Jenman
1021 (K) and 1022^ (K,NY); forest bordering brun-bracken savanna at
Adaro R. mouth, Kukui River, ca 500 m, 5 Sep 1960, fillet & Til let
54273 (NY-mixed; non U); Membaru Creek, upper Mazaruni River, 15 Feb
1 939, Pinkus 230 (NY,US).
25. Clitoria pendens Fantz, sp, nov.
Liana, to 30 m long (teste Irwin 54552). Branches typically 2-8 mm
diameter, occasionally to 4 cm thick, pith hollow, branches

452
subquadrangular becoming 1 .rete, pubescence of branches uncinate with
scattered, appressed trickmies (macrotrichomes deciduous well before the
more persistent uncinate trichomes), then glabious; bark grayish;
axillary buds 2-3 mm, scales ovate, acute, pubescence uncinate with a
few, suberect trichomes. leaves 3-foliate, thick membranous to
chartaceous, leaflets elliptic to ovate-elliptic, apex obtuse, nearly
abruptly acuminate, acumen 1-2 cm long, 3-3 mm wide at base, mucronate,
base broadly cuneate to rotund, midrib weakly raised above, primary
nerves of 6-8 pair, concolorous (green) to slightly light green below,
upper surface glabrous, lower surface glabrous (juvenile with
pubescence erect, confined to nerves and midrib above), lamina 7-15
(27) cm long, 4-7 (10) cm wide, lateral pair of leaflets occasionally
asymmetrical. Petioles subquadrangular to terete, 5-9 (11.5) cm,
pubescence sparse to glabrous, trichomes uncinate; rachis (1.5) 2-4
(4.5) cm. Petiolules subquadrangular, rugose, 5-6 mm, pubescence
sparse, of uncinate trichomes with occasional erect trichomes near
leaflet base (juvenile leaflets). Stipules deciduous, lance-deltoid,
acute, inconspicuously striate, 3-4 (5) mm long, 1 mm wide, pubescence
uncinate with a few, appressed trichomes, to glabrous; stipels deciduous,
linear, glabrous, 3-5 mm long, 0.2-0.8 mm wide. Inflorescence usually
caul iflorous, multiflowered, racemose, unbranched or rarely branched
near the base, highly elongated, generally 12-45 cm long bearing flowers
to occasionally 95 cm long with fruit, nodose, primary lateral branches
subsessile to 1 (2) mm, rarely to 5 mm long; inflorescences fascicled,
inserted mainly near the base of the stem (teste Maguire 54265),
pendent (teste Irwin 54552), or terminal and bearing small, deciduous
leaves at the base; axis pubescence uncinate, dense, with some appressed,

453
rufus trichoines, becoming sparsely uncinate toward the base. Peduncles
2-10 cm, internodes 2-15 mm. Pedicels 3-5 mm. Bracts minute, inner
bract caducous, solitary, linear, acute, 1-2 mm long, 0.2 mm wide,
middle pair semipersistent, lanceolate, acute, 2-3 mm long, 1 mm wide,
outer bract solitary, narrowly lanceolate, acute, 2-3 mm long, 0.5-0.7
mm wide. Bracteoles short, inconspicuous, narrowly lanceolate, acute,
3-5 mm long, 1-1.2 mm wide, inserted 1 mm below the calyx base,
pubescence of sparse, appressed, rufus trichomes. Flowers large, 6-8 cm,
juvenile white as opening, becoming pinkish to pale violet with age
(teste Irwin 54552). Calyx pubescence predominately uncinate, with
sparse to few, appressed, rufus trichomes, tube green with reddish-
purple tinge, 17-24 mm long, 5-8 mm wide at base expanding to 9-15 mm
wide at the throat, lobes ovate-deltoid, acute, 6-8 (9) mm long, 4-6 mm
wide at base, ventral lobe 7-9 mm long, 2 mm wide. Vexillum pubescence
uncinate to glabrate, blade abruptly curved outward near the base above
the claw, 3.5-5 cm wide, claw 10-15 mm. Alae extended 4-6 mm beyond
carina, blade suboblong-spatulate, 23-27 mm long, 7-10 mm wide, claw
with uncinate pubescence, 23-26 mm long. Carina arcuate, 15-20 mm long,
5-7 mm wide, claw with uncinate pubescence, 34-36 mm long. Stamens
diadelphous, vexillary stamen coherent near middle and coherent to free
near the base, tube glabrous, weakly arcuate, 40-48 mm long, free
filaments 2-6 mm; anthers lanceolate, 2-3 mm long, 0.7-0.8 mm wide,
connective apiculate. Gynoecium weakly arcuate; gynophore 5-8 mm;
ovary 18-20 mm long, 1.5-1.7 mm wide, pubescence dense, trichomes white,
tinged yellowish; style 27-33 mm, geniculate 8-13 mm from the distal
end; stigma capitate, 0.5-0.7 mm diameter; black to brownish, puberulent
at base. Legume short-stipitate, base enclosed by calyx lobes to

454
slightly exerted, flat, pubescence uncinate with scattered subappressed
to spreading trichomes, valves 13.5-16 cm long, 10-15 mm wide; stipe
23-31 mm long, 2-3 mm wide expanding to 5 mm ai apex; beak 4-8 mm;
dehiscence not observed. Seeds not observed, ca 7-8 per pod.
Figure 55.
The Pendent Clitoria is characterized as a liana with highly
elongated, pendent, nodose inflorescences which bear large pinkish-
purple flowers from a broad calyx.
PHENOLOGY: Flowers have been collected during two periods, from
March to April and from July through August. Fruits were collected
during July and August.
TYPE COLLECTION: SURINAM. Nickerie: Liana, caul iflorous, flowers
mainly at base, calyx green and reddish, corolla lavender, wood yellow,
in mesophytic forest, 9 km north Lucie River, 12 km west of Oost River,
3°36' to 3°41 1 N-56lJ30' to 56°34'W, 275 m, 16 Jul 1963, Maguire et al.
54265 (H0L0TYPE: NY).
The specimen selected as the holotype is the only collection
observed which has both flowers and fruits. However, this specimen
does lack the vegetative structures. A second collection from the
same locality, Irwin et al. 54552, was selected as a paratype. It has
vegetative and flowering structures, some of which are in a dissected
state. Both type collections have the characteriStic elongated
inflorescence.
VERNACULAR NAMES: SURINAM. Maneho (Chuk.), Gonggrijp 3816.
(SARAMACCA): Gate-tete, (Mat.) Florschutz & Florschiitz 1344.
NOTES: This species has close affinities with sagoti i and C.
kaieteurensis, but easily distinguished from thorn by the highly

Figure 55. Clitoria pendens. Left side: (a) leaflet, x 1;
(b) inflorescence, x 1; (c) flower, x 1; (d) calyx, x 1;
(e) vexillum, x 1; (f) ala and carina, x 1; (g) androecium,
x 1; (h) gynoecium, x 1; (i) two views of anther, x 4;
(j) style apex and stigma, x 5. (Craz 15JJD, MO 897321:
a-b. GH: e,g,i. Irwin et al., F 1615263: c-d,f,h,j.)
Right side: (a) juvenile inflorescence, x 1/2; (b) old
inflorescence, x 1/2; (c) calyx and stipe, x 1; (d) legume,
drawn from broken pieces, x 1. (Maguire et al., NY: a,d.
Pul 1e 310, U 37663A: b-c.)

456

457
elongated inflorescences and small bracteoles. This species also shows
affinities with C. 1eptostachya. Historically, any specimen with very
elongated inflorescences and lacking the large bracteoles that hide the
calyx was identified as C. leptostachya, the most frequent name found
on specimens of this species. Clitoria pendens is similar to
C. leptostachya in pubescence, appearance of the inflorescence, and
fruits. But C_._ pendens is distinguished from C. 1 eptostachya by a large
number of characteristics, as summarized in Table 10. Clitoria pendens
is easily distinguished from C^ 1eptostachya by the larger flowers,
broader calyx with longer lobes, and the stouter inflorescence axis.
DISTRIBUTION (Figure 57): This species is found in mesophytic forest
of Surinam and eastern Guyana at elevations (reported elevation data
scanty) of 275-360 m.
SURINAM. 19 Apr 1922, Gonggrijp 3816 (U); fluv. Gonini,
Aug 1903, Versteeg 148 (U). SARAMACCA: line from Paka Paka (Saramacca
R.) to Ebbatop, near 20 km, 12 Feb 1951, Florschutz & FIorschütz 1344
(NY,U); fluv. Saramacca, Mar 1 903, Pul 1e 468 (U); Voltsberg, 23 Aug 1920,
Pulle 302 (U-2 sh.) and 310 (U). NICKERIE: hills 9 km N of Lucie
River, 12 km W of Oost River, 275 m, Irwin et al. (PARATYPES:
F 161 5263,GH,M0 1800897 , NY ,S,U 1 69942B ,US 2447103)-, Lucie R. , langslijn
Lucie-Julianatop, ca 360 m, 8 Aug 1963, Schultz 10348 (U).
26. Clitoria 1eptostachya Benth., Journ. Linn. Soc. 2: 43. 1858.
Ternatea 1 eptostachya (Benth.) Kuntze, Riv. Gen. PI. 210.
1891.
Liana or erect shrub with apex more or less scandent, glabrous.
Branches 3-10 mm thick with larger ones 2-3 cm in diameter, triquetrous,

458
Table 10. A comparison of some structures between C. leptostachya
Benth. and C. pendens Fantz.
CHARACTER
C. LEPTOSTACHYA
C. PENDENS
LEAVES:
Shape
El 1iptic-1anceolate
Elliptic, Ovate-elliptic
Length/width ratio
2-3.5 : 1
1.8-.2. : 1
Apex
Acumen 2-2.5 cm
Acumen 1-2 cm
Primary nerves*
8-11 pair
6-8 pair
Petiolules
3-5 mm
5-6 mm
Midrib above
Strongly raised
Weakly raised
INFLORESCENCE:
Slender
Stout
FLOWER SIZE: *
Medium, 4.5-6 cm
Large, 6-8 cm
CALYX:
Tube length
16-21 mm
17-24 nun
Tube width throat*
Narrow, 7-10 mm
Broad, 9-15 mm
Lobe length*
4-6 (rarely 7) mm
6-8 (9) mm
BRACTEOLES:
Length
2 mm (Guianas)
3-4 mm (Brazi1 )
3-5 mm (Guianas)
Insertion
2-4 mm below calyx
Subtending, 1 mm below
PEDICEL: *
5-7 mm
3-5 mm
VEXILLUM:
Claw
9-12 mm
10-15 mm
Pubescence
Glabrate
Uncinate
Curvature
Weakly outcurved
Strongly outcurved
ALAE:
Beyond carina
3-4 mm
4-6 mm
Blade length*
13-18 mm
23-27 mm
Blade width*
4-6 mm
7-10 mm
Claw*
19-22 mm
23-26 mm
CARINA:
Blade length*
11-15 mm
15-20 mm
Blade width*
3-5 mm
5-7 mm
Claw*
26-32 mm
34-36 mm
ANDROECIUM:
Column*
32-37 mm
40-48 mm
Anther length*
1.5-2 mm
2-3 mm

459
Table 10. --Contiried
CHARACTER
C. LEPTOSTACHYA C. PENDENS
GYNOECIUM:
Ovary length*
1 2-1 5 min 18-20 mm
Ovary width* 1.2-1.3 mm 1.5-1.7 mm
Style vs. ovary* Slightly longer Much longer
Style length* 17-19 mm 27-33 mm
Geniculate length* Last 6-9 mm Last 8-13 mm
LEGUME:
Width
12-20 min 10-15 mm
*major differences

460
longitudinally furrowed between angles, pith solid, juvenile branches
subquadrangular and becoming nearly terete to angled, pubescence of
minute, white, appressed trichontes, soon glabrous; bark brown. Leaves
3-foliate, chartaceous, concolorous to slightly lighter green below,
glabrous or when juvenile with occasional inconspicuous pubescence along
the nerves below, juvenile leaflet elliptic becoming elliptic-
lanceolate to elliptic-oblong, apex acuminate, acumen 2-2.5 cm long,
2-4 nun wide, occasionally arcuate distal ly, more or less inucronate,
base broadly cuneate to weakly rotund, midrib subraised to strongly
raised above, primary nerves of 8-11 pairs, lamina 7-15 (21) cm long,
4-7 (9) cm wide. Petioles subquadrangular-terete, occasionally
longitudinally striated, 5-9 (13) cm, glabrate, pubescence appressed,
more or less scattered; rachis weakly compressed laterally, 2.5-3.5
(4) cm. Petiolules darkened, rugose, 3-5 mm. Stipules deciduous,
linear, acute, 2-5 mm long, 1 mm wide, pubescence uncinate with few
appressed trichomes; stipels deciduous, linear to subulate, 1-4 mm
long, 0.1-0.5 mm wide. Inflorescence axillary and solitary, or
cauliflorous and fascicled (2-4 peduncles), slender, multiflowered,
juvenile axis pubescence of short, appressed trichomes, dense becoming
scattered, and exposing uncinate trichomes, peduncles 2-11 cm,
internodes 1-6 cm. Pedicels dark-colored, 5-7 mm, pubescence of
uncinate and moderately dense, appressed trichomes. Bracts variable,
middle pair elliptic to oblong, broadly acute, 3-4 mm long, 1.5-2 mm
wide, outer bract ovate, obtuse, 1.7-2 mm long, 1.5 mm wide, inner bract
caducous, subulate. Bracteoles short, inconspicuous, lanceolate to
ovate, acute, 1.5-4 mm long, 1-1.2 mm wide, pubescence of uncinate and
some appressed trichomes. Flowers medium-sized, 4.5-6 cm, pale violet

461
to rose to pink to white. Calyx pubescence of uncinate and scattered,
subappressed trichomes, tube 16-21 mm long, 5-/ mm wide at base expanding
to 7-10 mm wide at the throat, lobes ovate-deltoid, acuminate for ca
one-third of their length, 4-6 mm long, 2-4 mm wide, ventral lobe
5-7 nun. Vexillum pubescence glabrate to uncinate, blade curved outward
at base just above claw, 4-5 cm wide, claw 9-12 mm. Alae extended
beyond carina 3-4 mm, blade with uncinate pubescence, 13-18 mm long,
4-6 mm wide, claw 19-22 mm. Carina weakly falcate, 11-15 mm long,
3-5 mm 'wide, claw 26-32 mm. Stamens diadelphous, vexillary stamen
coherent near base, tube weakly arcuate, split at base, 32-37 mm, free
filaments 2-4 mm; anthers lanceolate, 1.5-2 mm long, 0.5-0.7 mm wide,
connective acute. Gynoecium weakly arcuate; gynophore 7-9 mm; ovary
12-15 mm long, 1-1.3 mm wide; pubescence dense, trichomes white, tinged
yellowish; style slightly longer than the ovary, 17-19 mm long,
geniculate 6-9 mm from the distal end; stigma dark-colored, capitate,
ca 0.5 mm in diameter. Legume long-stipitate, exerted slightly above
calyx, flat, pubescence of uncinate and scattered, spreading trichomes,
valves 13-18 cm long, 12-20 mm wide; stipe 23-31 mm long, 2-3 mm wide
expanding to 5 mm at apex; beak not observed; dehiscence causing valves
to twist 2.5-3 turns. Seeds rarely observed, black, weakly compressed,
face suborbicular, 10-11 mm long, 11-12 mm wide, 3-4 mm thick, hilum
oblong, 2.5 mm x 1 mm. Figure 56.
Slender-spiked Clitoria is characterized as a more or less scandent
shrub with highly elongated, slender inflorescences which bear medium¬
sized flowers from a narrow calyx.
PHENOLOGY: This species flowers from May through September with
fruits produced in August and September.

Figure 56. Cl i tori a leptostachya. Left side: var. 1eptostachya :
(a) leaflet, x 1; (b) juvenile inflorescence, x 1;
(c) inflorescence, x 1/2; (d) flower, x 1; (e) calyx and
bracteole, x 1; (f) vexillum, x 1; (g) ala and carina,
x 1; (h) androecium, x 1; (i) gynoecium, x 1; (j) anther,
x 5; (k) stigma and style apex, x 4; (1) calyx and
juvenile fruit, x 1; (m) inflorescence with matured fruit,
x 1. (Schomburgk s.n., K-28, hb. Benth. : a,c. Forest
Dept. Br. Guiana 7210, NY: b,d-m. ) Right, side: var.
fruticosa: (a) leaflet, x 1; (b) matured inflorescence,
x 1/2; (c) flower with bracteoles, x 1; (d) legume, x 1;
(e) three views of seed, x 1. (Ducke 1 7244, S: d-e.
US 1442540: b-c. Ducke 235, A: a7J~'

â– r>-M
463

464
TYPE COLLECTION: GUYANA. Last collection alpha, Rob. Schomburak
s.n. (LECTOTYPE: K-28, hb. Benth. ; S=photo of K).
Bentham cited two syntype collection'.- as "In Guiana anglica ad
flumen Corentyn superius prope fines Bras 11iencium, Rob. Schomburgk,
coll. ii. sine num., et in Surinamo, Hostmann." Bentham noted that the
Hostinann collection had racemes, "tripol 1 icare" flowers, but lacked any
leaves. Clitoria leptostachya is absent from Surinam and has medium¬
sized flowers. No Hostmann collection bearing a distinct, elongated
inflorescence, has been examined, but based upon the description by
Bentham and examination of other Surinam specimens, the Hostmann syntype
probably would be excluded from this species and be assigned to
C. pendens, which has large flowers and a wide distribution in Surinam.
The only Schomburgk specimen observed with elongated inflorescences
lacks the full data cited by Bentham, but the specimen in Bentham1s
Herbarium at Lew placed in a type folder matches the original description
of the leaves, stem, inflorescence, and flowers. The stem in particular
was described by Bentham as "caules saepe elevato-triquetri faciebus
sulcatis, raro teretes." The three-angled stem with a sulcate channel
between the faces was observed in only two Clitori a collections from all
those examined for the genus. These collections included Rodrigues 457
(US) and the Schomburgk specimen at Kew. Since the Schomburgk specimen
is the most probable collection examined by Bentham, and the only
collection which matches the description published by Bentham, it was
selected as the lectotype.
NOTES: This species shows close affinities with pendens, a new
species previously included within leptostachya. Clitoria 1eptostachya
is easily distinguished from C_._ pendens by the smaller flowers, more

465
slender inflorescence with longer internodes, and narrower calyx.
Table 10 summarizes other characteristies that segregate the two species.
DISTRIBUTION (Figure 53): This species is found in forests on
sandy soil of Guyana and eastern Amazonas to western Para, Brazil.
Elevations have not been reported by collectors.
KEY TO VARIETIES:
1. Bracteoles ovate, 2 mm; ovary 14-15 mm; anthers 1.5 mm x 0.5 mm;
legume broader, 15-20 mm wide; flowers pinkish-white to white-
mauve; leaflet with midrib weakly raised above; liana (GUYANA)
26a. var. 1eptostachya
1. Bracteoles lanceolate, 3-4 mm; ovary 11-13 mm; anthers 1.7-2 mm x
0.5-0.7 mm; legume narrower, 12-16 mm wide; flowers rose to lilac;
leaflet with midrib strongly raised above; shrub, apex more or less
scandent (BRAZIL) 26b. var. fruticosa
26a. Clitori a leptostachya Benth. var. 1eptostachya
Liana. Leaflet with midrib weakly raised above. Bracteoles
minute, ovate, 2 mm. Flowers whitish, pinkish-white to white-mauve.
Ovary 14-15 mm long. Anthers 1.5 mm long, 0.5 mm wide. Legume
15-20 mm wide.
DISTRIBUTION (Figure 53):
G U Y A N A. Mission Field, 24 Sep 52, Forest Dept. Br. Guiana
7210 (NY); along Berbice-Rupununi cattle trail, 10 Jun 1919,
Abraham 1 71-A (NY).

466
26b. Cl i tor i a leptostachya Benth. vai'. fruticosa Fantz, var. nov.
Shrub with apex occasionally climbing. Leaf with midrib raised
above. Bracteoles short, 3-4 mm long, lanceolate. Flowers pigmented,
rose- to 1ilac-colored. Ovary 11-13 mm long. Anthers 1.7-2 mm long,
0.5-0.7 mm wide. Legume 12-16 mm wide.
TYPE COLLECTION: BRAZIL. Para: prope medium fl. Tapajoz, loco
Quataguara, 15 Aug 1923, Ducke 17244 (H0L0TYPE: S. Isotypes: 6 209 &
210, hb. Barbey-Boissier, RB 17244,2 sh., U 63828; US 1442540).
The Stockholm specimen has the only fruits and seed along with
the inflorescences, flowers, and vegetative structures. It has the
characteristic structures and is a better specimen.
DISTRIBUTION (Figure 53):
B R A Z I L_._ PARA: Faro, 12 May 1911, Ducke 1 1 652 (RB-2 sh.).
AMAZONAS: Manaos, 17 Sep 1929, Ducke 23405 (RB); Reserva Florestal
Ducke, do Buiao, Manaus, 6 Aug 57, Rodrigues 457 (US); margen do
Igarape, do Buiao, Manaus, 23 Jul 56, Coe!ho 3930 (US); estrada da
Forquilha, Manaus, 4 May 56, Coelho 3807 (US); Manaos & estrada do
Alexio, 16 Jul 1 936, Ducke 235 (A,F,M0,NY ,S).
27. CljtoHa sel 1 oi Benth., Journ. Linn. Soc. 2: 42. 1858.
Terna tea selloi (Benth.) Kuntze, Riv. Gen. PI. 1_: 210. 1891.
Liana. Branches 3-8 mm thick, subterete, pith hollow, juvenile
branches with pubescence of uncinate and conspicuous, dense, pilose
trichomes, branches soon becoming glabrate then glabrous; bark reddish-
brown, peeling in longitudinal strips, light brown beneath. Leaves

Figure 57. South American distribution of four species of section
Caul iflorae. Cl i tori a obi dens is (★); C. pendens (•);
C_L pi umosa (Q); C . sel loT~(¿Ék).

8917

469
3-foliate, thick membranaceous, concolorous, leaflets oblong-elliptic
to oblong-lanceolate to oblong, apex obtuse, abruptly acuminate, acumem
usually 0.5-1 (2) cm long, 2-7 mm wide, mocronate, base broadly
cuneate, midrib weakly raised above, primary nerves of 8-11 pair, upper
surface glabrous, lower surface pubescent, trichomes subappressed to
appressed, dense, becoming moderately scattered, lamina generally
7-16 cm long, (2.5) 3-4.5 cm wide. Petioles subterete, more or less
weakly compressed laterally, 3-7 cm, pubescence pilose becoming glabrate,
trichomes rufus; rachis weakly compressed laterally, 1-1.5 cm.
Petiolules quadrangular, 3-5 mm, pubescence pilose, moderate to
scattered, trichomes rufus. Stipules ovate to short-lanceolate, acute,
weakly striated, 3-5 mm long, 2 mm wide, glabrate with sparse,
uncinate trichomes and scattered, subappressed trichomes; stipels
lanceolate, acute, weakly striated, glabrate, 3-5 mm long, 0.5-1.5 mm
wide. Inflorescence axillary or cauliflorous, usually solitary,
racemose, multi flowered, flowers crowded toward the apex; axis 3-11
(16) cm long, quadrangular, twisting, nodose, pubescence of uncinate
and rufus, spreading to erect trichomes, becoming glabrous. Pedicels
4-6 mm becoming to 8 mm in fruit. Bracts lanceolate, acute, pubescence
sparse to moderate, pilose, ciliate; inner bract linear, caducous,
3-4 mm long, 0.5 mm wide; middle pair persistent, reflexed in age, 3-5 mm
long, 1-2 mm wide; outer bract deciduous, nearly equal to the middle
pair. BracteoIes lanceolate, acute, 6-8 mm long, 1.5-2 mm wide,
inserted ca 2 min below calyx base, pubescence sparse to moderately
dense, pilose, ciliate. Flowers small, 3-4 cm, pale violet, darker-
colored medially on vexillum inner surface Calyx pubescence
conspicuously rufus, pilose, moderately dense, becoming scattered with

470
age and more conspicuously uncinate, tube 10-13 (15) mm long, 3-5 mm
at base expanding to 6-7 mm at throat, lobes ovate, 4-6 mm long, 2-3 mm
wide at base, apex rapidly narrowed to acuminate tip to 2 mm, ventral
lobe 5-7 mm long, 1 mm wide. Vexillum pubescence uncinate, blade
weakly to strongly outcurved above claw, 1.5-2 cm wide, claw 5-8 mm.
Alae extended beyond carina 3-6 mm, blade subfa1cate-spatulate, 13-18 mm
long, 3-7 mm wide, claw 9-15 mm. Carina falcate, 8-11 mm long, 3-5 mm
wide, claw (15) 18-22 mm. Stamens glabrous, weakly arcuate, vexillary
stamen coherent below middle, tube more or less split near base, 24-
28 mm, free filaments 1-3 mm; anthers 1.5-2 mm long, 0.5 mm wide,
connective acute. Gynoecium weakly arcuate; gynophore dark-colored,
4-7 mm; ovary pubescence white to occasionally slightly tinged
yellowish; 10-12 mm long, 1 mm wide, style 13-17 mm long, geniculate
5-7 mm from the distal end, beard dense, trichomes nearly 1 mm long;
stigma capitate, ca 0.5 mm in diameter. Legume unknown; stipe 13 mm
long, 2-3 mm wide expanding to 5 mm at apex. Figure 58.
Sel low's Clitori a is characterized as a liana with medium-sized
inflorescences bearing small violet flowers from a pilose calyx. The
major axes of the plant are densely and conspicuously rufus, pilose, as
is the calyx in the juvenile stages. This species is the only one in
the section Caul if1orae with small flowers, and it is also the only
species of the section found in the eastern coastal states of Brazil.
PHENOLOGY: Dates of collection usually were not recorded, except
in two instances (April and September), both with flowers. From the
general appearance of the plant's leaves, inflorescences, and flowers,
the blooming period, if continuous, is more likely to be September to

Figure 58. CI i tori a seiloi. (a-b) leaflets, x 1; (c) juvenile
inflorescence, x 1; (d-e) older inflorescences, x 1;
(f) flower, x 1; (g) calyx, x 1; (h) vexillum, x 1; (i) ala
and carina, x 1; (j) androecium, x 1; (k) gynoecium, x 1;
(1) style apex and stigma, x 10; (m) anther, x 5; (n) old
inflorescence with stipe, x 1. (Sell_o 903, K-25, hb.
Benth.: b-c,k-l,n; BM: f-j,m. Riedel ?A3, BM: a,d.
Blanchet 3093A, G hb^. cte Candol 1 e: eTJ

472

473
April rather than from April to September. The specimen collected in
April had larger leaves and older inflorescences.
TYPE COLLECTION: BRAZIL. 1815-17, F. Sel lo [or Sel low] 993
(LECTOTYPE: K-25, hb. Bentham; S=photo of K. Isotypes: BÍ1-2 sh.).
Bentham (1858) cited the type collected as "Sello, in Brasilia
(media? or australiore?)." No collection number was designated. There
are two Sel low collections known. Sel lo 1008-811 (B-not seen,
destroyed in war; F-fragment and photograph of Berlin specimen!) lacks
the voluble condition of the stem described by Bentham. Sel low 993 in
the Bentham Herbarium at Kew is highly voluble and agrees with the
original description. Bentham (1862) cited Sel low for C. selloi. The
Berlin and British Museum specimens lack the "w" in Sellow's name
whereas the specimen in Bentham's Herbarium (K-25) has "Sel 1ow" and is
the probable type specimen, thus designated as the lectotype for the
species.
NOTES: A specimen of C. selloi has been collected only once in the
last century, and rarely in the twentieth century. Specimens apparently
fall into two distinct groups. The two Sel 1ow collections are highly
rufus-pubescent, with short inflorescences crowded with flowers, and
the buds are conspicuously pilose. The Blanchet and Riedel collections
have elongated inflorescences, more widely spaced internodes, aborted
flowers, and sparse to scattered pubescence. It is believed that these
two groups represent collections made in early bloom and near the end
of the blooming season. The plant appears to be highly pubescent in
early stages of growth, the pubescence becoming less dense with age.
This is a common characteristic observed in a number of species of
Clitoria. Inflorescences are crowded with flowers in early blooming

474
stages of growth, losing aborted flowers with an elongation of the axis,
as is typical of other species with medium-siziJ inflorescences.
Bentham noted the affinities of sel 1 oi to be near C. arborescens,
but tie did not elaborate. However, the species shows closer affinities
to C. leptostachya which is similar in the flowering and vegetative
structures. Clitoria sellen can be easily distinguished from the other
liana species by its small flowers and disjunct geographical location.
DISTRIBUTION (Figure 57): This species is widely removed
geographically from the other species of section Cauliflorae. It is
found along the eastern coastal states of Brazil, from Bahia to Rio de
Jane i ro.
B R A Z 1 L. Campo Vitoria, Sello 1 (.'08-81 I (F-frag. from and photo
of B). BAHIA: partie meridionale, 1840, B lan diet 3039A (G-3 sh.);
árenos maritime pr. Ilheus, Sep 1821, Riedel 243 (BM,NY,W). ESPIRITO
SANTO: via Lagoa do Brazil, Rio Doce, 15 Apr 1934, Kuhlmann 21_9 (RB).
RIO HE JANEIRO: entre Lagoa do Peixe et Rio Bonito, env. Rio de
Janeiro, 1882, Glaziou 1 2548 (G,NY).
28. Clitoria obidensis Huber, Bol. Mus. Goeldi Para 5: 405. 1909.
Liana, voluble and tall climbing, shrubby base. Branches 2-6 mm
in diameter, nearly terete, pith solid, pubescence of uncinate trichomes
hidden by conspicuous, erect, rufus, stiff trichomes of 1-1.5 mm long,
very dense on juvenile branches becoming moderately spaced. Leaves
3-foliate, thick chartaceous, concolorous, leaflets elliptic to
elliptic-oblong or elliptic-lanceolate, occasionally obovate, apex
acuminate to cuspidate, acumen 1-2 (3) cm long, 2-4 mm wide, mucronate,

475
base broadly cunéate to weakly rotund, midrib and primary nerves
prominently raised dbove, densely pubescent, primary nerves of 6-8 (9)
pairs, widely spaced, upper surface conspicuously pubescent, trichomes
rufus, 0.5-1.5 mm long, subappressed to falcate, slightly stiff,
texture more or less rough to the touch, lower surface with similar
pubescent, more dense, lamina 8-17 cm long, 2-7 (9.5) cm wide.
Petioles subquadrangular to terete, 3-7 cm, pubescence of uncinate
trichomes hidden by hispid, rufus trichomes of 1-2 mm; rachis 1.5-2.5
(4) cm.' Petiolules subquadrangular, 3-6 mm; pubescence moderately
dense, hispid, rufus, and inconspicuously uncinate. Stipules large,
ovate-lanceolate, acute to acuminate, 6-11 (15) mm long, 2-3.5 mm
wide, pubescence uncinate, ciliate, with a few hirsute, rufus trichomes
at apex and base; stipels large, linear-lanceolate, 6-10 (15) mm long,
0.7-2 mm wide, pubescence as stipules, terminal stipel one-half to
two-thirds the length of the lateral stipels. Inflorescence terminal
and axillary at denuded nodes, racemose, few flowered (4-8 flowers)
short, 0.5-1 (1.3) cm, solitary, axis with pubescence uncinate hidden
by conspicuous, hirsute, rufus trichomes. Peduncles 2-6 mm; rachis
internodes 2-5 mm. Pedicels 3-5 mm. Bracts large, conspicuous, dark-
colored, lanceolate to oblong, more or less short-acuminate, middle pair
8-11 mm long, 2.5-3.5 mm wide, outer pair 6-8 mm long, 2-3.5 mm wide.
Bracteoles large, conspicuous, length subequalling to longer than calyx
tube, oblanceolate to oblong-lanceolate, apex acuminate, base
abruptly attenuate, 20-35 mm long, 4-6 mm wide, inserted 1-2 mm below
calyx base, pubescence of uncinate and rufus, hirsute trichomes at apex
and scattered elsewhere, blowers large, rose-colored, usually with
complicate vexillum 5-6.5 cm long and spreading vexillum at 7-8.5 cm.

476
Calyx chartaceous, nerves more or less conspicuous, pubescence of
uncinate and rufus, spreading trichomes primarily confined toward base
and dorsal and ventral margins, tube 20-27 mm long, 5-7 mm wide at base
expanding to 9-14 mm wide at the throat, lobes highly elongated,
lanceolate, long acuminate, apex often bent, ciliate, 20-27 mm long,
5-8 mm wide, ventral lobe 20-25 (occasionally to 35) cm long, 2-4 mm
wide; lateral lobes overlapping dorsal lobes at base, dorsal lobe
innermost. Vexillum pubescence of uncinate and a few scattered
trichomes, blade 4-5 cm wide, claw 6-8 mm. Alae extended beyond carina
4-7 mm, blade subfalcate-spa tul ate, 26-34 mm long, 6-10 mm wide, with
sparsely uncinate pubescence, claw 18-20 n.m. Carina falcate, blade
10-15 mm long, 4-7 mm wide, with sparsely uncinate pubescence, claw
31 -3d mm. Stamens diadelphous, vexillary stamen nearly free, tube
weakl/ arcuate, glabrous, abruptly incurred last 4-8 m, 40-50 mm, free
filaments 2-5 mm; anthers lanceolate, 2-3 mm long, 0.7-0.9 mm wide,
connective acute. Gynoecium nearly straight to weakly arcuate;
gynoihore 3-5 mm; ovary 13-17 mm long, 1-1.5 mm wide, pubescence
tawny-sericeous; style 27-35 mm, geniculate 9-12 mm from the distal
end, beard dense, trichomes to 2 mm long; stigma capitate, flattened,
ca 1 mm in diameter, with ring of minute white hairs at base. Legume
short-stipitate, (stipe to 1.5 cm fide Rizzini, p. 186, 1963; base of
legume thus well enclosed within calyx), valves flat, more or less
weakly raised around seeds, pubescence uncinate with scattered,
subappressed, stiff, rufo-tawny trichomes, denser along sutures, surface
obliquely striolate, valves 7-10 cm long, 10-15 mm wide; beak to 4 mm;
stipe not observed; dehiscence causing valves to twist one to one and
one-half turns. Seeds subglobose, weakly compressed on two sides, face

477
suborbicular, subconvex, black, 7-8 mm long, 8 mm wide, 4-5 mm thick,
8-12 seeds per pod;- hilum oblong, 2 mm x 1 mm. Figure 59.
The Obidos Clitoria is characterized as a conspicuously rufus-
pubescent liana with short inflorescences bearing large rose-colored
flowers from a long calyx tube with elongated lobes, and having long,
conspicuous stipules, stipels, bracts, arid bracteoles.
PHENOLOGY: Collections of this species are scarce, but those
collected have produced flowers from mid May to late June and fruits in
May.
TYPE COLLECTION: BRAZIL. Para: in silvis prope Obidos, 10 May
1905, Ducke 7215 (LECTOTYPE: MG 7215; F=photo of MG! Isotypes: BM!
RB 11846-2 sh.! Cotype: F 602328).
Huber did not select a holotype specimen. The specimen at the
Goeldi Museum is the probable specimen examined by Huber, as this
specimen has been photographed and distributed as the type, and Huber
published the description of the species in the bulletin of that
institution. This specimen has not been examined, although the
duplicates (BM,RB) and a dissected flower and a leaflet from the type
and its photograph (F) have been examined. The specimen deposited at
the Goeldi Museum has the best material and more nearly matches the
description of Huber; thus, it is designated as the lectotype.
NOTES: Huber noted that this species had probable close affinities
with C. stipularis. Clitoria obidensis is superficially similar to both
C. stipularis and C. densiflora in appearance in that all three species
are conspicuously pubescent, have large flowers, and prominent stipules
and bracteoles. However, C. stipularis and C^ dens if1 ora differ greatly
from C. obidensis by the viscid seeds, costate legume with convex

Figure 59. Clitoria obi dens is. (a-b) leaflets, x 1; (c) inflorescence,
x l;"Td) flower, x 1; (e) vexilluin, x 1; (f) ala and
carina, x 1; (g) calyx, x 1; (h) androecium, x 1;
(i) gynoecium, x 1; (j) stigma and style apex, x 4;
(k) anther, x 4; (1) legume, x 1; (m) three views of seed,
x 1. (Ducke 7215, F 602328: e-k; RB 11846: d,l-m.
Ducke 11861, BM: a; S: b; FLAS 122959: c.)

479

480
valves, small ovary, and 10-nerved calyx, all characteristics of the
subgenus Neurocarpum. Clitoria obidensis is far removed from
C. siipularis and C. densiflora.
Clitoria obidensis has affinities with C. plumosa which is
distinguished by its very elongated stipules, medium-sized inflorescence,
shorter calyx lobes, and the elongated, plumose, ventral calyx lobe.
DISTRIBUTION (Figure 57): This species is endemic to the forest
near Obidos, Brazil. Huber (1909) illustrated Ducke's collection sites
on a map and placed the forest from which Ducke had collected specimens
NNE of Obidos.
B R A Z I L. PARA: Obidos, 25 Jun 1912, Ducke 11861 (BM,FLAS,
G,S,U,US).
29. Clitoria plumosa Fantz, sp. nov.
Liana, high growing to forest crown (teste Haught 2075)
conspicuously pubescent, trichomes dense, rufus, spreading to erect,
more or less stiff, 1.5-3 mm long. Branches 5-8 mm thick, reddish-
brown, angular becoming terete, dense rufo-pubescent, pith hollow ca
one-half of the diameter. Leaves 3-foliate, conspicuously pubescent on
both surfaces with long, rufus trichomes, thick membranaceous, leaflets
oblong to elliptic-oblong, rarely slightly wider above the middle, apex
obtuse, abruptly acuminate, acumen 0.5-1 cm long, to 4 cm wide,
mucronate or with bristle tip of 3-6 mm, base broadly cuneate, margins
rufus-ci 1iate midrib weakly raised above, primary nerves of 10-12 pair,
upper surface dark green, moderately pubescent, lower surface green,
densely pubescent especially on nerves, 1 amina 9.5-16 cm long, 4-6 cm

481
wide. Petioles angular-terete with swollen base of 6-7 mm, 11-14 cm
long, densely pubescent at the base to moderately dense above,
trichomes spreading to erect, tawny-rufus; rachis 4-5 cm. Petiolules
dark-colored, quadrangular, 4-5 mm; pubescence of uncinate trichomes
hidden by dense, long rufus trichomes. Stipules highly elongated,
conspicuous, persistent, narrowly oblong-lanceolate, narrowing gradually
the last two-thirds of its length into a long acumen, 19-29 mm long,
2-4 mm wide; pubescence ciliate with outer surface glabrate; stipels
highly elongated, narrow stipule-like, terminal stipels 8-15 mm long,
0.3-1 mm wide, lateral stipels 12-20 mm long, 1-1.5 mm wide.
Inflorescence cauliflorous, multiflowered, crowded, axis 10-15 cm
long, conspicuously bracteated, pubescence of scattered, uncinate
trichomes hidden by conspicuous, dense, long, rufus trichomes. Pedicels
5-8 mm, densely pubescent. Bracts persistent, very conspicuous,
stipule-like, 13-16 mm long, 2-4 mm wide, pubescence of uncinate
trichomes with a few, scattered, long trichomes, and ciliate, lowermost
bracts short, 7-11 mm. Bracteoles highly-elongated, stipule-like, long
acuminate, more or less attenuate base, 22-25 mm long, 3-4 mm wide,
inserted 1 mm below calyx and often slightly spreading, pubescence
ciliate with uncinate trichomes and scattered, long trichomes. Flowers
very large, 7.5-9 cm, light yellow flushed red (teste Haught; yellow
is a rare color in the genus, but other species with this color are
white, fading to yellow with age). Calyx pubescence dense, spreading,
tawny-rufus, tube 27-30 mm long, 8-10 mm wide at base expanding to
10-14 mm wide at throat, lobes deltoid-lanceolate, rapidly narrowing to
long-acuminate tip, dorsal lobes 15-18 mm, lateral lobes 13-16 mm long,
4-5 mm wide, ventral lobe highly elongated with linear base and subíate

482
apex, plumose, 23-28 mm long, 0.3-1 mm wide. Vexillum pubescence
glabrate to scattered-uncinate, densest toward base and margins, blade
4.5-5 cm wide, claw 9-11 mm. Alae extended beyond carina 6-8 mm, blade
falcate-spatulate, 30-35 mm long, 5-10 mm wide, claw 23-26 mm. Carina
falcate, blade 13-17 mm long, 6-8 mm wide, claw 34-38 mm. Stamens
diadelphous, vexillary stamen free to base, tube nearly straight,
incuived slightly at base and apex, 49-53 mm long, free filaments 3-6
mm; anthers lanceolate, 2.7-3 mm long, 1 mm wide, connective apiculate.
Gynophore subsessile; ovary 21-22 mm long, 1-1.3 mm wide, pubescence
white, soft, spreading-ascending (to 3 mm wide including pubescence);
style 31-33 mm long, geniculate 13-14 mm from the distal end, beard
dense, triehomes long; stigma capitate, flattened, ca 1 mm in diameter.
Legume unknown. Figure 60.
The Plumose Clitoria is characterized by the conspicuously
reddish-pubescent appearance, medium-sized, conspicuously bracteated,
inflorescences bearing crowded, very large flowers with a calyx having
an elongated, plumose ventral lobe, and the narrow, elongated bracteoles,
bracts, stipules, and stipels.
PHENOLOGY: The one collection of the species had flowers mostly
in early stages of maturity in mid-November.
TYPE COLLECTION: COLOMBIA. Santander: Viscaina Creek, 32 km S
of El Centro, vicinity of Barranca Bermeja, Magdalena Valley, between
Sogamoso and Carare River, ca 100 m, 15 Nov 1936, Haught 2075
(H0L0TYPE: UC M 049493. Isotypes: F 1450138, GH,NY).
Both the specimens from Gray Herbarium and Field Museum are more
fragmentary, having an inflorescence, a couple of leaflets, and a flower
placed in the packet. The other two collections have much more material

Figure 60. Clitoris plumosa, (a) leaflet with stipels, x 1;
Xb-) stem with stipules, x 1; (c) inflorescence, x 1;
(c) flower, x 1; (e-f) two views of calyx, x 1; (g)
vex ill uni, x 1; (h) ventral calyx lohe, x 2; (i) ala and
carina, x 1; (j) androecium, x 1; (f) gynoecium, x 1;
(1) anther, x 3; (m) stigma and style, x 5. (Haught 2075,
UC 49493: a-b,d-m; NY: c.)

17817

485
and a dissected flower. Tfie specimen deposited at the University of
California has more mature leaves and stem, and therefore, is selected
as the hoiotype.
¡'JOTES: This species is easily recognized by the elongate, plumose
ventral calyx lobe and the elongated stipules, stipels, bracts, and
bracteoles. Clitoria plumosa has close affinities with C. obidensis
which is distinguished by its short inflorescences, elongated calyx
lobes subequal to the tube 1ength, and shorter bracteoles, bracts, and
stipule's.
DISTRIBUTION (Figure 57): This species is known only from the type
locality in Santander, Colombia.

486
Subqenus Clitoria
II. Cl itoria L. subgenus Clitoria.
Ternatea Tourn., Mem. Math. Phys. Acad. Roy. Sci. p. 103.
1706; nom. in va 1.
Clitori us Petiv. ex Dill., Hort. Elth. p. 90. 1732; nom.
inval.
Clitori a L., Gen. PI. 216, no. 572. 1 737; nom. inval.
Ternatea Tourn. ex Mill., Gard. Diet. ed. 4, 3; TERNATEA.
1754.
Vexi1laria Eaton, Man. Bot. p. 82. 1817; nom. superfl.
Ternatea Tourn. ex Kunth, Nov. Gen. Sp. 6: 415. 1824.
Clitoria L. sect. Ternatea (Tourn. ex Kunth) DC., Prod. 2:
233. 1825.
Nauchea Desc., Mem. Soc. Linn. Par. 4: 7. 1826; nom.
superfl.
Clytoria Presl, Rostí. 3; 196. 1835; nom. ill eg. (?orthoqr.
pro. Clitoria L. 1753)^
Clitoria L. subgenus Ternatea (DC) Baker, FI. Br. India
p. 208. 1879.
Suffrutescent herbs, erect or twining, or occasionally a liana.
Leaves commonly 5- or 7-foliate, infrequently 9- or 11-foliate, rarely
3- or 1-foliate; leaflets small, 2-9(12) cm long, 0.1-5 cm wide,
midrib weakly to strongly raised above, upper surface pubescent,
either microscopically uncinate or macroscopically strigose, trichomes
ca 1 mm, scattered, upper surface occasionally becoming glabrate with
age, lower surface usually strigose. Petioles shorter than rachis,

487
typically 0.5-4 cm long. Petiolules short, 0.7-4 mm. Stipules
narrow, 0.1-2 mn wide, 1-5 nerved; stipels very narrow, 0.1-0.4 mm
wide. Inflorescence axillary, usually solitary, commonly bearing 1 or
2 flowers at apex of peduncle, occasionally racemose, few-flowered;
axis typically short, 0.2-1 (1.5) dm. Bracts typically 1-7 mm long,
0.2-4 mm wide. Bracteoles ovate to nearly orbicular, often one-half
to occasionally subegualling the calyx size and hiding it, membrana¬
ceous, subpellucid, 3-, 5-nerved, inserted at the base of the calyx.
Flowers chasmogamous, small to medium-sized, 2.5-6 (6.5) cm,
papilionaceous or very rarely actinomorphic (5 vexillum-like petals
and some to all stamens free). Calyx subpellucid, (5-) 10-nerved,
somewhat shrinking in fruit but persistent; tube short, 0.4-1.4 cm
long, pubescence of uncinate and scattered, short, subappressed to
spreading trichomes; lobes typically subequal to tube length,
occasionally shorter, typically (2) 4-11 mm long. Vexillum
short-clawed, 1-5 mm, occasionally longer to 8 mm, commonly glabrate.
Alae claw 7-14 mm, blade variable. Carina 5-10 mm long, 2-6 mm wide,
claw short, 10-18 mm long. Stamens diadelphous, vexillary stamen
more or less connate near base; stamina! tube short, (0.8) 1-2 cm;
anthers small, 0.7-1.5 mm long, 0.5-1 mm wide. Gynophore short,
1-1.5 mm; ovary small, 0.5-0.8 cm long; style short, 0.7-1.8 cm,
beard dense below stigma. Legume subsessile, base coneate, enclosed
within persistent calyx, valves flat, ecostate, thickened on sutures,
nearly straight to weakly curved at apex, ventral margin sometimes
undulating; stipe 1.5-2 mm. Seeds smooth, compressed, subreniform,
small (2-6 mm long x 3-8 mm wide), usually 1-10 seeds per pod. Seed
germination epigean. (x=8).

488
The members of the subgenus Clitori a can be characterized as
typically suffrutescent herbs with 5- to 11-foliate leaves,
subpellucid, 5- or 10-nerved cálices which have lobes nearly equal
to the tube length, subpellucid, ovate to orbicular bracteoles,
1- or 2-flowered pedunculate inflorescences, and with flat, ecostate
fruits bearing compressed, reniform seeds lacking a viscid coat.
HOLOTYPIC SPECIES: C. ternatea L.
Nomenclaturally, this subgeneric group has historically been
known by the name Ternatea. A. P. de Candolle (1825) segregated this
group from other Clitori a species primarily by the number of leaflets
g
(i.e. 5 or 7 leaflets). He gave this group sectional status where
it has been known historically as Clitoria L. section Ternatea DC.
A. P. de Candolle (1825) cited Kunth as the author of Ternatea. Kunth
(1824) cited Tournefort as the author of Ternatea. Tournefort (1706)
published the generic name Ternatea, but being pre-Linnaean (i.e.
published before 1 May 1753), his name is invalidly published in
accordance with the rules established by the International Code of
Botanical Nomenclature (Stafleu, 1972). Kunth has been recognized
historically as the first to validly publish the name Ternatea, thus
the full citation would be Ternatea Tourn. ex Kunth. When de Candolle
transferred the genus Ternatea to be sectional level, it became
Clitoria L. sect. Ternatea (Tourn. ex Kunth) DC. when fully cited.
Recently, the Gray Herbarium Card Index noted that Miller had validly
8. The other characters used (i.e. tubular calyx and spurless
vexillum are common to all Clitoria species, but were used to
segregate section Centrosema. Today, Centrosema is recognized as a
segregate genus from Clitoria.

489
published Tournefort's name in 1754. Neither Kunth nor de Candolle
referred to Miller's publication.
Baker (1879) elevated de Candolle's section to the subgeneric
level as Cl itoria L. subg. Ternatea (DC) Bak. He used the fruit
(i.e. pods flat, ecostate) as the segregating structure. Baker's
treatment has not been followed by other botanists. Most commonly
de Candolle was followed and this group treated (in floras) as
section Ternatea. Baker's treatment of this group at the subgeneric
level has probably been overlooked in the literature, as opposed to
not being accepted by his botanical colleagues, because his
publication is never cited by later authors.
The name Euclitoria was introduced by de Candolle (1825) as a
sectional name and applied to a different group of Clitoria species.
These species, historically known as the "true clitorias," belong to
the subgenus Neurocarpum. The name Euclitoria has been misapplied
historically because the "true clitorias" should be that group
including C. ternatea, the generitype which has never been included
under the name Euclitoria. In accordance with present nomenclatura 1
rules under the International Code of Botanical Nomenclature (Stafleu,
1972), the epithet of a subdivision of a genus containing the type
species is either of the same form as the generic name, or a plural
adjective agreeing in gender (Art. 21). In addition, the epithet is
not to be formed by the prefix "Eu-." The name Euclitori a is
illegitimate even if had been applied to the correct group. Baker's
subgeneric name Ternatea must be rejected since this subgenus contains
the type species of the genus, and in accordance with the Code, the
epithet must be of the same form as the generic name. Therefore, the

490
correct name for this subgenus would be Clitori a L. subg. Clitori a.
In a vernacular sense, the species of this subgenus should be
considered as the "true clitorias."
NOTES: Members of the subgenus Clitori a have several character¬
istics shared by the primitive subgenus Bractearia, including flat,
ecostate fruits which bear smooth, compressed seeds, epigeal seed
germination, and the lack of cleistogamy. Members of subgenus
Clitori a also have a number of characteristics shared by the advanced
subgenus Neurocarpum. With the exception of C. lasciva, the habit is
suffrutescent. Reduction in size has occurred in a number of
structures, including fruit length and width, stipe length, seed
dimensions, leaf size, petioles, leaf blade dimensions, the gynoecium,
and the androecium. Seeds have become reniform in shape. Calyx
nerves have become more distinct, with each of five nerves, leading
to the apex of one lobe, and five additional nerves more or less
distinct, with each nerve leading to a sinus and dichotomously forking
such that each branch becomes marginal in adjacent lobes. Inflores¬
cences bear few flowers, typically one or two at the apex of an
axillary peduncle. Subgenus Clitoria appears to represent an
evolutionary line that departed early from that of Neurocarpum.
Members of subgenus Clitori a exhibit several characteristics that
are distinct from the other two subgenera. Leaves are typically 5- to
11-foliate. Bracteoles and calyx are subpellucid, membranaceous,
translucent to somewhat clear, and bearing raised, prominent nerves.
In the dried state, the nerves appear to hold the structure intact
since the tissue between nerves easily tears when manipulated with a
probe. It also appears that this tissue quickly shrinks somewhat in

491
size and shrivels slightly when specimens are collected and pressed,
as evidenced by the specimens examined. The same condition is
characteristic of the bracteoles and cálices of fruiting material.
Shrinking in size and shriveling of the calyx and bracteoles do
not occur typically in the other two subgenera, neither of which have
subpellucid cálices nor bracteoles, but have persistent cálices which
retain nearly the same size during fruit stages as in the flowering
stage. Fruits are subsessile with the stipe typically 1-1.5 mm.
Although very limited research has been done on the karyogamy of
species of Clitori a, from the available evidence, it appears that
x=8 may be typical for the subgenus.
Bracteoles in this subgenus infrequently exhibit a primitive
condition observed in sections Bractearia and FIexuosa of subgenus
Bractearia. Some bracteoles become quite large, nearly orbicular in
shape, hiding the calyx, 11-15 mm in diameter (bracteoles are
commonly 5-10 mm), and are very conspicuous when compared to others
within the species. Examination of additional nodes of the same plant
or additional material of the same collection deposited at different
institutions usually reveals that both the typically more advanced and
the primitive bracteole types exist, and there is often a gradient of
sizes and shapes. The primitive bracteole type is observed most
commonly in African specimens of ternatea.
The African species C. kaessneri would appear to be misplaced if
only the vegetative form is considered, having 3- or 1-foliate leaves
typical of a number of Neurocarpum species. Subgenus Neurocarpum is
commonly absent from Africa except for two introduced species
(C. falcata and C. laurifolia) which have become naturalized in

492
isolated populations. Examination of C. kaessneri indicates that the
species has non-viscid seeds, flat, ecostate fruits borne on a
subsessile stipe, and subpellucid cálices, characteristics associated
with subgenus Clitoria, not with Neurocarpum, the species of which
have viscid seeds, convex, turgid, costate fruits, and subcartilaginous
cálices. Asian members of subgenus Neurocarpum have ecostate fruits,
but have long calyx lobes, which C. kaessneri does not possess.
Clitoria kaessneri agrees with the size gradient of a number of
flower structures when compared to other species within subgenus
Clitoria. Its affinities lie with those species of subgenus Clitoria,
not with those of subgenus Neurocarpum. Thus, it is the only species
within this subgenus with less than five leaflets.
Clitoria lasciva is unique as the only woody member (a liana) of
the subgenus, and in exhibiting an elongated, racemose inflorescence
bearing several flowers. Clitoria ternatea is unique in producing a
variety with "double" flowers that have actinomorphic petals and free
or nearly free stamens.
Within the subgenus there is a trend toward reduction in size of
a number of structures, including the flowers and their associated
structures, the leaves (particularly in leaflet size), and the fruits.
These trends can be observed within the largest species, C. ternatea,
a polymorphic species. Clitoria heterophylla exhibits the maximum
reduction reached by this evolutionary line. Many of its structures
are the smallest known for the genus.
DISTRIBUTION (Figure 11): Members of subgenus Clitoria are
pantropical. The large range of the subgenus is due predominantly
to one species, C^ ternatea, which has been transported to foreign

493
lands by man at least since the late Seventeenth Century. The native
distribution of the subgenus was probably in the paleotropics, which
included Africa (excluding western Africa?), Madagascar and the
Mascarene Islands, and the Indian peninsula (and Ceylon?). Clitori a
ternatea has been introduced as a cultivar, and often establishing
itself natively as a weed in the neotropics, in numerous Pacific
Islands, in Australia, and probably in Indonesia and Southeast Asia.
KEY TO THE SPECIES:
1.Leaves 5-, 7-, 9-, or 11-foliate; legume 4-11 mm broad; seeds
compressed, 0.5-2 mm thick; stipules narrow, 0.3-1.5 mm wide;
petiolules short, 0.7-3 mm long.
2.Legume 6.5-11 cm long, 8-11 cm wide (ca 2.5 cm x 5 mm in rare
C. ternatea var. pilosula); staminal tube 13-20 mm, or three
to all stamens are free; anthers 1-1.5 mm long; ovary 6-9 mm;
style 12-18 mm; flowers (3) 3.5-6.5 cm, papilionaceous or
rarely double with five vexi1lary-1ike petals, actinomorphic;
alae extending beyond carina 6-10 mm, blade 16-28 mm long;
carina 7-10 mm long.
3.Inflorescence 4-15 cm, 4-8 flowered; flowers 5.5-6.5 cm;
calyx tube 6-8 mm; larger bracts 4-6 mm; ovary 6-7 mm;
leaves 7-, 9-, and 11-foliate; liana (Madagascar) ....
30. C^ 1asciva
3. Inflorescence 0.5-1.5 cm, 1 (2) flowered; flowers (3)
3.5-5.5 cm; calyx tube 8-14 mm; larger bracts 2-4 mm;
ovary 7-9 mm; leaves 5- and 7-foliate; suffrutescent
herb forming tangled mats or climbing (Pantropical)
31. C. ternatea

494
2. Legume 2-5 (6) cm long, 4-8 mm wide; staminal tube 8-11 mm;
anthers 0.7-1 mm long; ovary 5-6 mm; style 7-12 mm; flowers
2.5-3.5 cm, papilionaceous; alae extending beyond carina
2-4 mm, blade 8-11 mm; carina 5-7 mm long.
4. Herbaceous vine; stipules 1-3 mm; stipels 1 mm; bracteoles
2-3 mm long, 0.4-1.0 mm wide; calyx lobes 2-3 mm;
peduncles l-(2) flowered; leaves dimorphic, type I oval to
orbicular, type II linear (a branch may produce only 1
type), pubescence above uncinate (Madagascar & Mascarene
Islands) 32. C_^ heterophylla
4. Erect herb; stipules 6-11 (16) mm; stipels 3-7 (9) mm;
bracteoles 6-11 mm long, 3-5 mm wide; calyx lobes 4-8 mm;
peduncles 2-flowered; leaves monomorphic, leaflets
basically lanceolate, pubescence above scattered strigose
(India) 33. C. biflora
Leaves 3- and 1-foliate; legume 10-15 mm wide; seeds 2-3 mm thick;
stipules broad, 1.5-2 mm wide; petiolules 3-4 mm (Angola, Zaire,
Zambia) 34. kaessneri
30. C1itori a lasciva Bojer ex Benth., Ann. Wein. Mus. Natur. 2:
114. 1837.
Clitoria lasciva Bojer, Hort. Mauritianus p. 92, 1837;
nom. nud.
Ternatea lasciva (Boj. ex Benth.) Kuntze, Riv. Gen. PI. 1:
210. 1891.

495
Liana, climbing in shrubs and trees, conspicuously pubescent.
Stems 1-1.5 m (teste Bojer 76), lignose, longitudinally striate,
twining, 2-4 mm in diameter, pith hollow, ca 1-1.5 mm in diameter,
pubescence short, densely pilose, becoming puberulent with age.
Leaves 7-, 9-, and 11-foliate, thick-membranaceous, leaflets basically
ovate, becoming elongated with age, oblong-ovate, lanceolate-ovate,
or elliptic-ovate, to lanceolate, mottled with black spots on both
surfaces (vidi 8 x), more conspicuous on upper surface, apex acute,
variable from short- to long-tapered apex, base rotund to weakly
retuse, midrib raised above and often bearing minute remnants of
trichomes along midrib and major nerves, primary nerves of 6-8 pairs,
to 11 pairs on elongated leaflets, upper surface dull, dark green,
with pubescence moderately strigone on juvenile leaflets, becoming
widely scattered to glabrate with age, lower surface pale green,
with sericeous pubescence on juvenile leaflets, trichomes becoming
scattered, subappressed, lanima 2-6 cm long, (1) 1.5-2.5 (rarely 3)
cm wide. Petiole caniculate above (more rarely sulcate), shorter
than the rachis, 0.5-4 cm; pubescence dense to medium, short-pilose;
rachis 4-11 cm long, internode segments 0.5-2 cm long with terminal
segment clearly the shortest; petiolules subterete, 2-3 mm long,
pubescence spreading, moderate to dense. Stipules linear, conspic¬
uously narrow, long-tapering to an acute apex, prominently 5-nerved
at base becoming 3-nerved near midway and 1-nerved near apex,
5-8 mm long, 1-1.5 (2) mm wide; pubescence short-pilose; stipels
linear, extremely narrow, 1-3 mm long, 0.1-0.2 mm wide. Inflorescence
axillary, solitary, racemose, (2) 4-6 (8) flowered, predominantly
4-9 cm long occasionally to 15 cm long in robust specimens; axes

496
pubescence short, pi 1ose-puberulent; peduncles subquadrangular,
longitudinally striate, 3.5-6 (12) cm; rachis internodes 0.4-1.5 cm.
Pedicels 4-7 mm. Bracts pubescent, trichomes short, spreading; inner
bracts not observed; middle bract lanceolate, persistent, often hidden
by outer bract, 3-4 mm long, 1.5-2 mm wide; outer bract large, showy,
bracteole-1ike, ovate to broadly ovate, more or less apiculate,
conspicuous palmately nerved, 4-6 mm long, 3-4 mm wide. Bracteoles
broad ovate, translucent, membranaceous, short-acuminate, 5-8 mm long,
3-6 mm wide, inserted 1 mm below calyx, pubescence spreading-pilose,
ciliolate. Flowers blue, in dry state pale blue to purplish-blue with
yellowish middle on vexillum; perianth base and outer surface mottled
black spotted. Calyx short-tubular in flower, becoming campanulate-
tubular in fruit as upper lobes spread apart from lower lobes and
tube shrinks slightly, 5- or 10-nerved, five nerves prominent, each
entering a lobe and extending to the apex, five nerves somewhat
prominent, each leading to a sinus, dichotomatously forking with each
branch extending from one-third to one-half of the way marginally into
adjacent lobes, pubescence of uncinate and conspicuously appressed to
spreading trichomes, tube 6-8 mm long, 3-4 mm wide at base expanding
to 5-7 mm wide at throat, lobes oblong to ovate-oblong, apex obtuse,
abruptly acuminate, ciliate, subequalling the tube length, 6-8 (9) mm
long, 3 mm wide, vexillary lobe similar, 7-9 mm long, 3 mm wide.
Vexillum glabrate, (5) 5.5-6.5 cm long, 4-4.5 cm wide, claw minute,
3-4 mm. Alae extended well beyond carina by 6-10 mm, blade 17-21 mm
long, 6-10 mm wide, claw 8-12 mm. Carina subfalcate, 8-10 mm long,
3-5 mm wide, claw 12-15 mm. Staminal tube bearing pilose to ascending
hairs towards base, vexillary stamen nearly free, tube 13-17 mm long,

497
free filaments 4-7 mm; anthers lanceolate, 1.2-1.5 mm long, 0.8-1.0 mm
wide. Gynophore nearly lacking, 1 mm long, 0.7-0.8 mm wide, whitish-
sericeous; ovary 6-7 mm long, 1-1.3 mm wide; pubescence whitish to
yellowish-white, densely sericeous; style 12-16 mm long, weakly
geniculate 4-6 mm from the distal end, beard dense below stigma for
3-4 mm; stigma capitate, ca 0.7 mm in dia. Legume subsessile, linear
to weakly arcuate towards apex, base cuneate, flat, ecostate, becoming
weakly raised around seeds, pubescence conspicuously pilose and
inconspicuously uncinate, valves 6.5-9 cm long, 9-11 mm wide; stipe
1.5-2 mm, enclosed along with base of fruit by persistent calyx;
beak 4-10 mm; dehiscence by separation and spreading of valves,
twisting action not observed in specimens examined. Seeds black,
reniform, compressed (1 mm thick) except near center (to 2 mm thick),
smooth, 4 mm long, 7-8 mm wide, 4-6 seeds per pod; hilum 3 mm x 2 mm.
Figure 61.
Bojer's Clitoria is characterized as a liana with 7- to 11-fo1iate
leaves and elongated inflorescences bearing medium-sized blue flowers.
PHENOLOGY: Bojer (1837) reported the flowering dates as
November to December in Madagascar and March to April in Mauritius.
Collection data of those specimens examined do not support the
November-December dates. Flowering specimens were collected from
April to September with fruits collected from June to September.
However, those specimens collected in Mauritius lacked the collection
date as well as a number of specimens from Madagascar.
TYPE COLLECTION: MADAGASCAR. Scandit in arboribus et in
frútices ad margines fluviorum provincia Bé-tani-menae ora oriental is,

Figure 61. Cl i tori a lasciva: (a) portion of stem with leaf and
inflorescence, x 1; (b) portion of older stem with flower
and immature fruit, x 1; (c) calyx, x 1; (d) vexillum, x 1
(e) ala and carina, x 1; (f) androecium, x 1; (g) anther,
x 7; (h) gynoecium, x 1; (i) style apex and stigma, x 6;
(j) mature fruit, x 1; (k) three views of seed, x 1.
(Bojer 76, W: a-b. Taulay s.n., W 9622: c-i. Thompson
s.n., BM.)

66fr

500
Flo. Sep. Oct. 1830, Bojer v.76 (LECTOTYPE: W. Isolectotypes: K-hb.
Bentham, M 12439).
Bentham did not designate the depository of the type specimen.
The specimen in his herbarium was a duplicate from Wein which was
added in 1836, but like the Miichen specimen, lacked any data beyond
"Madagascar, Bojer." Only the Wein specimen has the habitat and
locality data cited by Bentham when he published Bojer's name.
Bentham did not cite the collection number, a practice commonly
followed in his 1837 treatment. The Wein specimen has leaves with
seven, nine, and eleven leaflets as described in the original
description by Bentham. The specimen in Bentham's herbarium has few
leaves, and lacks 9- and 11-foliate leaves. Stafleu (1967) noted that
Bentham borrowed specimens from other institutions on loan, and that
his types may be in other institutions besides Kew. The Wein specimen
was probably examined by Bentham on loan because it is a better match
to the original description, and bears the cited collection data.
Thus, it was selected as the lectotype instead of the specimen in
Bentham's herbarium.
The Heisenberg collection deposited at the British Museum may be
a duplicate of the type collection. There are two labels. One bears
the data "Hab. in prov. Be-tani-mena ins. Madagascar. Nomen madag.
Vahé-bourissika quod characterem hujius generis Linnaéanum bene
experimit. Clitori a sp." which closely parallels the data of the type
collection (W), not cited above, of "Clitoria lasciva Boj. Scandit
. . . orientalis insula Madagascar. Flo. Sep. Oct. Norn,
madagascariensis Vahe-m-burisika. quod Linnei characterem bone
experimit." This phraseology does not occur in the literature, and

501
has been found only on the label data of these two specimens. The
second label bears "Madagascar, Heisenberg" written in a different
ink and handwriting. The British Museum specimen lacked a species
name and Bojer's name in the data, which was present in the Kew and
Muchen duplicates. Is this specimen a duplicate added to Heisenberg's
herbarium before Bojer named the species, or is there another
explanation for the close similarity in label data?
VERNACULAR NAMES: MADAGASCAR: Vahe-m-burisika, Bojer v.76;
Vahe-bourissika, Heisenberg s.n.; Vahytsikomba, Hochrentrier (1908);
Famahyfary, Hochrentrier (1908); Fanefary, Thompson s.n.
ECONOMIC IMPORTANCE: This species is cultivated in gardens
locally in Madagascar and Mauritius for its climbing habit, foliage,
and the deep blue flowers.
NOTES: Clitoria lasciva appears to have its closest affinities
with C. ternatea which has similar fruits and flowering structures,
but is distinguished by the smaller inflorescence bearing usually one
flower, the larger calyx tube, the suffrutescent habit, and fewer
leaflets.
The origin of the specific epithet appears to be the Latin term
"1ascivus" which means "wanton" or "licentious."
DISTRIBUTION (Figure 62): Clitoria lasciva is an endemic in
Madagascar found in the forested margins of rivers, and cultivated in
adjacent island gardens.
MADAGASCAR. Belanger s.n. (G); 17 Ju1 1863, Blackburn
s.n. (GH.K.NY); 29 Mar 1941, Decary 1698 (K,US); Humbolt 151 (K,W);
Lyal1 93^ (K); Aug-Oct 1838, MacWi11 iams s.n. (G); Thompson s.n. (BM);
central Madagascar, Baron 1418 (BM,K) and 2589 (K); coast to capital,

Figure 62. African distribution of three species of subgenus Clitoria.
Cl i tor i a kaessneri (•); lasciva (▲}; C. heterophylla
var. peduñculata Tit), var. heterophylla (ft).

503

504
May 1880, Kitching s.n. (K-2 sh.); Bedilo, Majunga, 27 May 1912,
Afzelius s.n. (K,NY,S-2 sh.); valle de 1'Ikopa, 14 Mar 1930, Decary
7534 (G-2 sh.,M0,NY,S); NossiBe, Sep 1879, Hildebrandt 3128 (BM,G-2
sh.,K,M,W); l.c., 1888, Anonymous s.n. (NY,W-2 sh.); Diego-Suarez,
Jun 1887, Taulay s.n. (NY,W); l'interieur de la baie de Antongil,
Maroa, 1897, Mocquery 189 (G); Isle Ste. Marie, Jun 1887, Taulay s.n.
(W-2 sh.); l.c., 1846-52, Voyage Boivin s.n. (G); l.c., Boivin 1917
(G); Foulepointe, 1863, Bouton s.n. (G); Tamatave, Jul 1887, Taulay
s.n. (W-2 sh.); Vatomandry, 15 Apr 1904, Guillot 112 (G-2 sh.);
Massif ne Ankara, 29 Jul 1929, Decary s.n. (BM,K); Tsingy de Namoroka,
8e Reserve, 4 Apr 1933, Service Forestier, Anonymous 20 (BM);
Be-tani-mena, Heisenberg s.n. (BM).
MAURITIUS, cult., Bouton s.n. (CGE); cult., hb. Lemann
s.n. (CGE); cult, gardens only, Telfair s.n. (K).
31. Clitoria ternatea L., Sp. PI. ed 1. 2: 753. 1753.
Flos clitoridis ternatensibus Bre.yn., Cent. 76, t. 31.
1678; nom. ill eg.
Phaseolus foliis pinnatis Riv., Terr. 40. 1691 ; nom. ill eg.
Phaseolus indicus glycyrrhyzae foliis, flore ampio,
caeruleo, pleno Comm., Hort. Med. Amstel. 47, t. 24.
1697; nom. illeq.
Phaseolus indicus caeruleus, glycerrhizae foliis alatis,
flore ampio, cl i torio PIuk., Aim. 294. 1 700; nom.
.IU.1SL--
Ternatea flore simplici, caeruleo Tourn., Mem. Math. Phys.
Acad. Roy. Sci. Amsterdam 105. 1706; nom. illeq.

505
Ternatea flore simplici albido Tourn., l.c.; nom. ill eg.
Ternatea flore pleno, caeruleo Tourn., l.c.; nom. ill eg.
Flos clitorius, flore caeruleo Burm., Zeyl. 100. 1737;
nom. illeq.
Flos clitorius, flore albo Burm., l.c.; nom. illeq.
Clitori a foliis pinnatis L., Hort. Cliff. 360. 1737; nom.
illeq.
Flos coeruleus Rump., Amb. 5: 56, t. 31. 1747; nom. inval.
Clitoria foliis pinnata L. B Flos clitori us, flore alba
(Burm.) L., Fl. Zeyl. 130. 1747; nom. illeq.
Cl itoria ternatea L., Hort. Upsal. 214. 1748; nom. inval.
Ternatea flore simplici, caeruleo Tourn. ex Mill., Gard.
Diet. ed. 4, 3: TERNATEA. 1754; nom. illeq.
Ternatea flore simplici, albido Tourn. ex Mill., l.c.; nom.
illeq.
Ternatea flore pleno, caeruleo Tourn. ex Mill., l.c.; nom.
illeq.
Clitoria ternatensium Crantz, Inst. Rei. 2: 59. 1766.
Lathyrus spectabilis Forsskal, Fl. Aeq. 135. 1775.
Clitori a ternatea L. B eadem foliiolus obtusioribus, f1 ore
albido Lam., Ency. Bot. 2: 50. 1786; nom. illeq.
Clitoria bracteata Poir. in Lam. Ency. Bot. Supp. 2: 301.
1811.
Ternatea vulgaris H.B.K., Nov. Gen. 6: 415. 1823.
Clitoria ternatea L. B bracteata (Poir.) DC., Prod. 2: 234.
1825.
Nauchea ternatea Desc., Mem. Soc. Linn. Par. 4: 8. 1826.

506
Nauchea terna tea Dése. var., 1.c.; desc. var., nomen nullus.
Nauchea bracteata Dupuis ex Desc., l.c. 4: 11.
Clitori a ternatea L. a coeruleo Sweet., Hort. Brit. 2nd ed.
140. 1830. nomen nudum.
Clitoria ternatea L
. 6
alba Sweet, l.c.; nomen nudum.
Clitoria ternatea L
. flore pleno alba Wall., Cat. Herb. Ind
185 no. 5344K.
1831-1832; nom. illeq.
Clitoria pilosula Wall.,
l.c. 186 no. 5347; nomen nudum.
Clitoria parviflora
Raf.
, Atl. Journ. 1(4): 147. 1832.
Clitoria ternatea L.
a
fio. alba Boj., Hort. Maurit. 91.
1837; nom. i 1 leg.
Clitoria ternatea L.
A.
fl. atrocoeruleo Hasskarl, Cat.
Hort. Bog. Alt.
275
1844; nom. ill eg.
Clitoria ternatea L.
B.
f1. coeruleo Hasskarl, l.c.; nom.
ill eg.
Clitoria ternatea L.
C.
fl. coeruleo pleno Hasskarl, l.c.;
nom. i 1 leg.
Clitoria ternatea L.
D.
f1. albo Hasskarl, l.c.; nom.
ill eg.
Clitori a ternatea L. 3 albiflora Voigt, Hort. Calcuttensis
213. 1845.
Clitori a ternatea L. var. major Burnett, PI. Util. 2: 84.
1847.
Clitoria pilosula Wall. ex Benth., Journ. Linn. Soc. 2: 37.
1858.
Clitoria zanzibarensis Vatke, Oestr. Bot. Zeitschr. 28: 261.
1878.

507
Clitori a ternatea L. var. pilosula (Wall, ex Benth.) Bak.
in Hook. FI. Brit. India 208. 1879.
Clitoria coelestris Sieb. & Voss., Vilm. Blum. 208.
1894; nomen nudum.
Clitoria tanganicensis Micheli in Dur. & DeWild, Bull. Soc.
Roy. Bot. Belg. 36(2): 60. 1897.
Clitoria albiflora Mattei, Boll. Bot. Palermo 6: 97. 1908.
Clitori a ternatea L. var. angustifolia Hochst. ex Bak. f.,
Leg. Trop. Afr. 428. 1929.
Clitoria mearnsii DeWild., Rev. Zool. Bot. Afr. 13, Supp.
Bot. 8: 8. 1925.
Clitoria tanganyicensis Micheli err, cal. Bak. f., Leg.
Trop. Afr. 429. 1929.
Clitoria ternatea L. var. alba Berhaut, FI. Senegal 47.
1954; nom. ill eg.
Clitoria ternatea L. var. angustifolia Hochst., nom. in
sched,
Clitoria ternatea L. var. bracteis ovata Schimper, nom. in
sched.
Clitoria ternatea L. var. flore albo Aiudie, nom. in sched.
Clitoria ternatea L. var. stenophylla Welw., nom. in sched.
Clitoria ternatea L. f. stenophyl!a Welw., nom. in sched.
Clitori a pinnata, nom. in sched.
Cl itoria serrulata B.G., nom. in sched.
Clitoria timoriana, nom. in sched.

508
UNVERIFIED SYNONYMS: A number of additional names have been
synonomized in the literature with this species. All names are
pre-Linnaean with abbreviated citations. The authors of these names
and the title of the publication are presently not clear. Burmann
(1737) included all of these references with a few cited in later
publications by other authors.
Clitorius flos albus, H.L. Bat. 162.
Clitorius zeylanicus, flore coeruleo, Hort. Catal.
Faba purgatriz CATHARODU Zeylonensibus Grim., Laba. Ceyl. 124.
Flos clitoridis ternatensium, Amm. Char. PI. 62,394.
Galea Phaseoloides Zeylandica, flore albo, Mus Zeyl 32.
Phaseolus adfinis, Glycyrrhizae German i cae foliis, oriental is,
f1 ore ampio coeruleo, Par. Bat. Pr. 364.
Phaseolus Indicus, Glycyrrhizae foli is, flore ampio caeruleo,
H. Amst. 47.
Phaseolus Indicus, purgans, seu Galega Phaseoloides Zeylonica,
flore coerruleo, Mus Zey. 10,31.
Phaseolus Zeylandicus, purgans, flore albis, Mus Zeyl. 61.
Perennial suffrutescent herb growing as a bush or more commonly
with a suberect base, branches trailing and intertwining to form
tangled mats, or climbing, twining, to 5 m long. Stems terete,
slender, 1-3 mm thick, weakly striate, much branched near woody base,
infrequently branched above, pubescence strigose, densely on juvenile
stems becoming scattered then glabrate with age; internodes usually
5-15 cm, pith hollow. Leaves 5- and 7-foliate, rarely with 3-foliate
leaf, membranaceous, green to silvery green, polymorphic in shape;

509
leaflets ovate, elliptic, obovate-elliptic, ovate-elliptic, oblong,
oblong-lanceolate, or oblong-linear, apex obtuse or acute, more or
less mucronate, sometimes retuse to emarginate, base cuneate to
broadly cuneate or rotund, midrib weakly raised above and prominently
raised below, primary nerves of 5-6 pairs to 7-9 pairs in elongated
leaflets, marginal nerve often strigose-pubescent, upper surface with
very short trichomes, strigose-pubescence or glabrate, and with
microscopic uncinate trichomes, usually denser on narrow, elongated
leaflets, confined more toward midrib and sometimes the major nerves
(or nearly lacking) on broader, ovate-shaped leaflets, lower surface
strigose, moderate to scattered and confined mainly to the major
nerves, trichomes short, uncinate trichomes inconspicuous and
scattered, lamina 1-5 (7) cm long, (0.4) 1-3 cm wide. Petioles
1-4 cm, shorter than the rachis, sulcate above, weakly striate;
strigose-pubescent; rachis 2-7 cm, caniculate above, pubescence and
striations similar to petiole, internode segments 1-2.5 cm.
Petiolules subquadrangular, 1-3 mm long, uncinate-pubescent and
moderately densely strigose. Stipules persistent, linear, 4-10 mm
long, 0.5-1 mm wide; pubescence uncinate and sparsely strigose,
strigose-ci1iate; stipels persistent, 1-3 mm long, acicular, reduced
to a nerve, inconspicuous, uncinate-pubescent with very sparse-
appressed trichomes. Inflorescence axillary, solitary, reduced to a
short peduncle bearing one flower at its apex, occasionally
biflowered, rarely fascicled or racemose with 4-6 flowers; peduncle
0.3-1.5 cm long, glabrous to sparsely strigose. Pedicel 3-6 mm long,
usually borne laterally from peduncle apex or occasionally terminally,
straight or often bending at swollen base to form an obtuse to right

510
angle with peduncle, moderately dense-strigose and uncinate-pubescent.
Bracts 2, apparently the outer pair, ovate to lanceolate, acuminate to
acute, spreading-ascending or weakly concave and appressed to pedicel
when pedicel is terminally located, 2-4 (5) mm long, 1 mm wide,
pubescence uncinate and more or less strigose, ciliolate. Bracteoles
very conspicuous, (4) 6-11 (15) mm long, (4) 6-11 (15) mm wide,
membranaceous, translucent, broadly ovate to nearly orbicular,
shorter than calyx tube to subequalling the calyx length and hiding
it, appressed to calyx, deciduous by fructification, apex sometimes
apiculate from central nerve, major nerves prominantly raised,
palmate, pubescence uncinate and sparse strigose, ciliolate.
Flowers showy, pale blue to azure to violaceous, or white (sometimes
fading yellowish in dried state), or white with thin edge of blue
peripherally, (3) 3.5-5.5 mm, usually papilionaceous, rarely producing
"double" actinomorphic flowers. Calyx subpellucid, more or less
10-nerved with a nerve extending to the apex of each lobe and a nerve
to each sinus where it dichotomizes sending an inconspicuous to
somewhat conspicuous branch marginally into adjacent lobes, pubescence
uncinate with sometimes appressed trichomes near the nerves
(especially ventral nerves), tube 8-14 mm long, 3-4 mm wide near base
expanding to 6-9 mm wide at the throat, shriveling and shrinking
slightly with age, lobes oblong, apex rounded with excurrent nerve of
1-2 mm, lobes 7-12 mm long, 3 mm wide at base. Vexillum pubescence
uncinate and strigose dorsally, medially white or yellow in throat
with blue to violaceous pigments distally, occasionally greenish-white
medially and white peripherally, 2-4 cm wide, claw minute, 2-4 mm.
Alae extended well beyond carina by 6-10 mm, blade white or bluish,

511
17-28 mm long, 7-13 mm wide, claw 7-11 mm. Carina falcate, 7-10 mm
long, 4-6 mm wide, claw 14-18 mm. Stamens diadelphous (cf. double
flowers below), vexillary stamen nearly free, tube nearly straight,
16-20 mm long, incurved slightly last 3-5 mm, free filaments 3-4 mm;
anthers lanceolate, 1-1.5 mm long, 0.6-0.9 mm wide. Gynophore 1-2 mm;
ovary 6-9 mm long, 1-1.5 mm wide; pubescence appressed, dense, white;
style 14-17 mm long, base pubescent like that of ovary, bearded
toward apex, densely pubescent under stigma; stigma capitate,
ca 0.7-0.8 mm in dia. ACTIN0M0RPHIC (DOUBLE) FLOWERS: Corolla azure
to dark blue, rarely white, 5-merous, all petals vexi1lary-1ike,
actinomorphic. Stamens ten, all free or infrequently with some free
and others fused in two or three fascicles (e.g., 5+3+1 +1;
3+2+5 free; 3 + 2 + 2 + 3 free). Legume subsessile, yellowish-
green to green becoming light brownish to tan in dried state, flat,
ecostate, two thickened sutures, valves linear-oblong to oblanceolate,
nearly straight to weakly curved towards apex, margins sometimes wavy,
infrequently minutely serrulate, pubescence short, strigose to
spreading, and uncinate, typically (5) 7-11 cm long, 8-11 mm wide,
rarely only ca 2.5 cm x 4-5 mm; stipe enclosed along with legume base
within persistent calyx, 1-2 mm; beak 2-6 mm; dehiscence causing
valves to twist 1-3 turns. Seeds brown turning black, smooth,
subreniform, compressed, 1.5-2 mm thick, 4-5 mm long, 5-6 mm wide,
7-10 seeds per pod; hilum broad elliptic, 1 x 0.8 mm. (Figures 63,
64, and 65).
The specific epithet of C. ternatea refers to the Moluccan Island
of Ternate from which the earliest known collection of the species was
reported by Breyne in 1678. The epithet has no relationship with the

Figure 63. C1 i tori a ternatea - I. Var. ternateay (a) habit, x i;
(IT) flower, x I; (c) calyx, x 1; (Â¥] vexillum, x 1; (e)
ala and carina, x 1; (f) gynoecium, x 1; (g) androecium,
x 1; (h) style apex and stigma, x 10; (i) anther, x 7;
(j) fruit, x 1; (k) three views of seed, x 1. (Zambrano
105, FLAS 125537: a,c-k. Lea 12, K-200: b.)

O o

514
morphological term "ternate" which denotes that the leaves are
compound with three leaflets. The Ternate Clitoria is characterized
as a suffrutescent perennial herb with twining, trailing to climbing
stems bearing 5- and 7-foliate leaves, and bearing short inflores¬
cences with solitary, azure or white flowers which quickly mature into
elongated, flat fruits.
PHENOLOGY: This species is a profuse bloomer and seed producer.
Flowers and fruits are typically collected together, both collected in
every month of the year with no apparent seasonal peaks except that
collections made in August are limited.
TYPE COLLECTION: INDIA. Suratt, Anonymous s.n. (LECTOTYPE:
LINN 902.1-microfische seen). CEYLON: Hermann 3:13 and 3:20 and 4:49
(Syntypes at BM, not seen).
Linnaeus (1753) published four species of Clitoria in his "Species
Plantarum" listing C. ternatea as the first species. Clitori a
ternatea has been the only species of these four recognized in the
literature as the type species for the genus. Britton and Brown (1913)
were the first to designate C_. ternatea as the type species, followed
by Britton and Mil 1spaugh (1920) and Hitchcock and Green (1 929).
The author of this study agrees with their selection. Only two of the
four species were cited in earlier Linnaean works prior to 1753,
C. ternatea and C. bra si liana (the latter now recognized as a species
of the genus Centrosema). Both species were included by Linnaeus
(1737a) when he originally described the genus based upon Tournefort
(1706) and Dillenius (1732). Examination of these two publications
would indicate that Linnaeus based his generic description upon
Tournefort's genus Ternatea which included three species, each

515
presently included within the species C. ternatea as subspecific taxa.
Examination of the generic description published by Linnaeus would
indicate that key characteristics such as the "tubular calyx,"
"carina shorter than the alae," and "subrenifortn seeds" agree with
C. ternatea but not with C. brasiliana which has a campanulate calyx,
carina nearly equal to the alae, and ellipsoid-cy1 indric seeds.
Evidence indicates that C_^ ternatea was the species upon which
Linnaeus described the genus Clitoria, and thus the type species.
For purposes of typification, Stearn (1957) divided Linnaean
species into four groups. Clitoria ternatea can be assigned to his
fourth group: "species based on several elements, of which one is to
be taken as the lectotype." Each of these elements would be syntypes.
Gillett, Pohill, and Verdcourt (1971) reported three Hermann
collections (3:13, 3:20, and 4:49) from Ceylon as syntypes, now
deposited at the British Museum of Natural History. Stafleu (1967)
reported that Hermann specimens were a part of the Burmann Herbarium
of Ceylon which Linnaeus helped arrange and annotate. A fourth
syntype element would be the specimen in the Linnaean Herbarium from
Surat, India. Savage (1949) reported that the handwriting on this
specimen of C. ternatea was that of Linnaeus. A fifth possible
syntype element would be the illustration (Table 31) published by
Breyne (1678) which accompanied a detailed description of Flos
clitoridis ternatensibus (=C. ternatea), the earliest known record of
the species. Tournefort (1706) based his genus Ternatea upon the
Breyne element, and in turn, Linnaeus later based his generic
description of Cl itoria upon Tournefort's genus. Linnaeus ( 1 737a,
1737b, 1748, 1753) cited Breyne and thus had studied the illustration

516
and its accompanied description, but there is no evidence as to
whether he had studied the specimen upon which the description and
illustration was based. This specimen may be deposited with the
Breyne Herbarium at Leiden, although this has not been confirmed.
Stearn (1957) indicated that the most important element in
Linnaeus's protologue was the Linnaean phrase-name. He noted that
"Linnaeus based the phrase-name nearly always on a specimen or an
illustration seen by him, for he distrusted the descriptions of
others, and the phrase-name is thus likely to give a significant
indication of his intent; certainly he would never have chosen an
element disagreeing with the phrase-name to typify his species."
Linnaeus consistently used the phrase-name "Clitori a foliis pinnatis"
for C. ternatea. In two of his earlier works, Linnaeus (1737b, 1747)
included Rivinius's name "Phaseolus foliis pinnatis," a possible
source for the Linnaean phrase name. Yet Linnaeus did not include
Rivinius's name in his Hortus Upsaliensis (1748) and Species Plantarum
(1753) while he continued to include Breyne, Tournefort, and most of
the others whose names were based upon these earlier authors. If
Linnaeus had taken his phrase-name from Rivinius's element, it is
doubtful that Linnaeus would have excluded Rivinius's reference while
retaining nearly all others. It is more probable that Linnaeus
recognized that C_^ ternatea was the only one of the four Linnaean
species that has 5- and 7-foliate leaves, whereas the other three
species are all 3-foliate, thus easily recognized by its leaflet
number. The other three species all include in their phrase-names,
the phrase "Cl i tori a foliis ternatis_ . . . ," a phrase lacking in any
element that Linnaeus included with those species. Thus in this

517
particular case, it does not seem probable that the Rivinius element
would be considered as a syntype, as there is no evidence that
Linnaeus examined anything other than a description.
In the selection of the lectotype, there seemed to be five
probable syntypes from which to choose. Three elements were from
Ceylon, one element from India, and the last element was from the
Molluccan Island of Ternate. In his protologue, Linnaeus (1737b)
reported the localities as Malabaria (Southwest India), Zeylona
(Ceylon), and Ternate. In his Hortus Upsaliensis (1748) and his
Species Plantarum (1753) protologues, Linnaeus cited only India.
Since Linnaeus liad the Indian element primarily in mind as evidenced
by the change in his treatments, the specimen in the Linnaean
Herbarium (no. 902.1) from Surat, India, was chosen as the lectotype
from among the probable syntypes.
VERNACULAR NAMES: North America: UNITED STATES: Blue Pea
(Florida), Small (1933). MEXICO: Caracolito (Chipas), Hernandez
1013: Escuintla (l.c.), Matuda 16425; Manga de nino (Jalisco), Mexia
1070; Zapotillo (Vera Cruz), Rosas 663; Zapatico de la reina,
Butterfly pea, Pigeon wings, Conchitas, and Papito (l.c.), FAO-68.
EL SALVADOR: Z a pa til 1 o de la reina, Standley 20595 and Standley
(1928, 1931, 1937). GUATEMALA: Diego, Steyermark 29160 and Standley
(1946); Concha blanca (Peten), Standley (1946). PANAMA: Campanilla,
Standley (1928, 1931).
Caribbean Islands: CUBA: Frijolito, Schott TJ4; Conchita azul,
Conchita blanca, and Manto de vieja, León (1951). BAHAMAS: Blue pea,
Britton (1920). BERMUDA: Blue pea, Britton (1918). CAYMAN ISLANDS:
Bluebell, Kings 63 and Proctor 15182. JAMAICA: Blue pea, Adams

518
(1972). PUERTO RICO: Bejuco de Conchitas, New York Academy of
Science (1924), Standley (1928), and Stahl (1936); Papito, New York
Academy of Science (1924) and Standley (1928). GUADELOUPE: Pois
savane, Pois marrón, Pois sauvage, Pois tonnelle, Pois-pois, and
Lentille sauvage, Stehle & Quentin (1948). MARTINIQUE: Pois hallier,
Stehle 3484 and 3520. BARBADOS: Blue vine, Goodwis 106. TRINIDAD:
Blue pea, White pea, Freeman (1951). CURACAO: Jerba dipalomba,
Anonymous 806 (U); Tapata di la reina, Amoldo 1809. FRENCH ANTILLES:
Pois-savane, Duss (1897).
South America: VENEZUELA: Zapatico de la reina, Domínguez s.n.,
Eliás 270 and 481, Zambrano 105, and Pittier (1944). BRAZIL:
Cunh'á, Kok et aj^ ( 1943).
Africa: CAMER0UN: Blue pea, Ogu 98. ANGLO-EGYPTIAN SUDAN:
Kordofan pea, Andrews (1952). ETHIOPIA: Geseziat Ogaden, and
Lu-kiliale isbeak, Geover s.n. SOMALIA: Ghria hele (BM), Ghria ejee
(BM), Riaegee (K), and Riahele (K), Glover & Gilliland 140; Fifiole,
Tuckard 108; Salbocole, Mattei (1908). ZAIRE: Lozanghe, Madezu,
Zinzanghi (Boma), and Laku (Mahagi), Wilczek (1954). UGANDA: Karimi
kamwe (Lungomkla), Purseglove 406. KENYA: Auk-egelai, Mathew 6389.
TANZANIA: Chugu ndo, Davis 54; Ihojo (Kisukuma), Tanner 1429.
MAURITIUS: Liane madame, and Liane Ternate, Bojer (1837).
Western Asia: IRAQ: Hab al zherate, Rustin Exp. Farm 10087.
ARABIA: Moglagidi, Schweinfurth 260. PERSIA: Bikhe-hayat, Chopra
et al. (1949). ARABIC: Bazrul-mazariyune-hindi, and Mazariyune-
hindi, Chopra et al_^ (1949).
India: Aparajita, Tiwari & Gupta (1959); Butterfly pea, Chopra
et al. (1949) and Sen (1961); Chandra kanta, Katiyar et al. (1970);

519
Mussel-shell creeper, Chopra et a_L_ (1949) and Sen (1961); Bhovera,
and Kajali, Nairne (1894); Gokarni, Maheshwari (1963) and Nairne
(1894); Neel-aparajita, and Shwet-uparajita, Voigt (1845); Sanka-gida
(Ingolis), Hohenacker 117. BENGAL: Aparajita, Chopra et al. (1949)
and Kirtikar & Basu (1918); Ni 1-aparajita, Swet-aparajita, and
Uparajita, Chopra et. a]_^ (1949). BOMBAY: Gokaran, and Kajali,
Chopra et al. (1949) and Kirtikar & Basu (1918); Gokarna, Supli,
Chopra et al. (1949). DECCAN: Ghutti, Kalizer, and Phiki, Chopra
et al. (1949). GUJERAT: Garani, Chopra et al. (1949) and Kirtikar
& Basu (1918); Koyala, Chopra et. ajL (1949). HINDI: Aparajit,
Chopra et al. (1949, 1958); Kalina, Kava-thenthi, and Visnukranti,
Chopra et al. (1949) and Kirtikar & Basu (1918); Aparajita, Kajina,
Kalizer, Khagin, Kova, Kova-theti, and Shobanjan, Chopra et al.
(1949). KANARESE: Gokarna-mul, Kirgunna, and Vishnu kantisoppu,
Chopra et al. (1949) and Kirtikar & Basu (1918); Girikarni-bal 1i, and
Sankhapushpa-bal1i, Chopra et al. (1949). KONKANI: Kazuli, Chopra
et al. (1949). MARATHI: Gokaran, Gokarni, Gokurna, Kajili, and
Sholonga, Chopra et al. (1949). PUNJAB: Dhanattar, Kirtikar & Basu
(1918). PUSHTU: Aparajita, Aparajit, Dhanattar, and Ghiria, Chopra
et al. (1949). SANSKRIT: Aparajita and Asphota, Chopra et al. (1949,
1958) and Kirtikar & Basu (1918); Gokarnamul, Khurne, Nilaghiria,
Nilaghirie kurni, and Vishnukranta, Chopra et al. (1949). TAMIL:
Kakkanam-kodi, Chopra et al. (1949) and Kirtikar & Basu (1918);
Kakkanam, Kakkattan, Kanni-kodi, Karakartan, Karisanni, Karkakartum,
Karkurattai, Karudakkovai, Karundattondai, Karuvilai, Kaurigeni,
Kavalai, Kemachi, Kevari, Kigini, Kiriganni, Kiruttini, Minni,
Taruganni, Uyavai, and Viranu, Chopra et al. (1949). TELUGU:

520
Dintana, Dintena, Mella, Nalla-ghentana, Nalla-vusinitige, Ni 1 a-
dintana, and Tel 1a-dintana, Chopra et al. (1949). TULU: Sanka,
Chopra et al. (1949). URDU: Mazeriyuni-hindi, Chopra et al. (1949).
URIYA: Onasi and Oporajita, Chopra et al. (1949).
Ceylon and Southeast Asia: SRI LANKA (CEYLON): Kata-rodu,
Deschamps 2; Katarodu-wel s, deSilva 87; Kataroduwael, Burmann (1737);
Katarodu (white), Burmann (1737); Catharodu, Burmann (1737); Katte
roddoe, Burmann 74 and Schlechtendal (1830); Nil-kata-rolu s, deSi1va
98; Siel catta kodoe, Burmann 75. BURMA: Bukiyu, Bukyu, Oung mai
phyu, and Pai noung ni, Chopra et al. (1949); Pe-nauk-ni and
Aung-mai-phya, Hundley & Koko (1961). THAILAND (SIAM): Aug chan,
Kirtikar & Basu (1935). MALAYASIA: Bonga biru, Burmann (1737) and
Rhumpf (1747); Bunga biru, Burkill (1935); Bonga calente, Rhumpf
(1747); Kachang telang, Ridley (1922) and Burkill (1935); Kakkanan,
Chopra et al. (1958); Aral, Kakkanam-koti, Sankhankuppi, Sankhapushpam,
and Sholongo-kuspi, Chopra et al. (1949).
Phillipines and Indonesia: PHILLIPINES: Balog-bálog (C. Bis.),
Giting-princésa (Bik.), Kolokánting (Tag.), Kalompági (Ilk.),
Pukinggan (Tag.), Samsampin (Pang.), and Samsanping (Ilk.), Merrill
(1923); Calocanting (T.), Colocanting, and Puquiñgang (T.), Merrill
(1903). SUMATRA: Inggir ingir, Jungle lau bakal, and Kampoeng
Singgadanik (Karo-land), Barlett & LaRue 168. CHRISTMAS ISLAND:
Bunga biru (Malayan), Powell 13. JAVA: Bonga biru, Burmann 256;
Kembang telang, Kembang telang bodas (white), Kembang télang buláu
(atro-coeruleo), Kembang telang bulau ad muda (coerulea), and Kembang
telang susun (pleno), Hasskarl (1848); Kembang tel eng and Mentéléng,
Burkill (1935). TERNATE: Bokyni Cotele de Principisse Clitoris,

521
Breyne (1678); Saja cotele, Bokyina cotele, and Fula Criqua (Port.),
Ruinpf (1747).
Australia and Pacific Islands: AUSTRALIA: Butterfly pea,
Hassell (1945) and Gardner and Bennetts (1956). FIJI ISLANDS: Kau
yalewa, Smith 1028; Thanga balangi, Parham 51. TONGA: Paipa, Yuncker
15810. POLYNESIA: Liane (Uahuka), Brown (1935). HAWAII: Butterfly
pea, Inafuku s_.nJ
General vernacular names, horticulture: Butterfly pea, Uphof
(1968); Mussel-shell creeper, Lowis (1878); Wing-leaved Clitoria,
Brown (1814).
ECONOMIC IMPORTANCE: Clitoria ternatea is the most economically
important species within the genus because of its widespread
distribution and easy availability, with reported uses in the
literature dating back to 1678. This species has been prized as an
ornamental, is reported to have several medicinal and aphrodisiacal
properties, and has recently been noted by researchers as a promising
forage crop in the tropics. It is also a useful species in some
geographical areas as a food and as a dye. Clitoria ternatea was
fully treated with specific literature references in an introductory
chapter (see chapter "Economic Importance" for details). A brief
summary of economic uses (excluding horticultural) reported by
collectors are as fol1ows:
Medicinal: Root a purgant medicine or remedy (Martinique), Hahn
234; Roots pounded and fed to children of the Sukuma tribe that have
constipation (Tanzania), Tanner 1429; Roots are chewed for lung and
throat complaints (Anglo-Egyptian Sudan), Pritchard 4. Aphrodisiac:
Before a cow is sent to the bull, a root is chewed and blown up her

522
nostrils to insure her fertility (Anglo-Egyptian Sudan), Pritchard 4.
Dyes: Used in dyeing (Mozambique), Chase 2815; Malaya kompongs,
flowers used for coloring cakes and rice (Christmas Island), Powel1
13. Forage: Legume cherished by goats (Ethiopia), Iecaina 1-81;
Young pods especially eaten by goats and gazelles (Kenya), Mwangangi
& Gwynne 1251; Bees are attracked to the flowers (Kenya), Mwangangi &
Gwynne 1251.
NOTES: There are several descriptions of C^ ternatea which
report that the leaves are rarely 9- and 11-foliate, or include
9-foliate leaves along with the more common 5- and 7-foliate leaves.
These descriptions are more common in some of the floras (e.g.,
Jamaica, Fawcett and Rendle 1920; Trinidad and Tobago, Williams,
1931: Suriname, Amshoff, 1939), probably because they based their
description of the species in part upon the earlier revision of
Clitoria by Bentham (1858) who reported the higher leaflet number,
noting that nine or eleven leaflets were rarer. Bentham's revision
included a slight change from his description of C^ ternatea in 1837
in which he reported the leaflet number as six (probably an
orthographic error as even-pinnate leaves have never been reported for
the genus) to eleven. Don (1832) was possibly the first to report
nine leaflets for this species. Examination of the specimens of this
species failed to reveal any example of nine or eleven leaflets,
although rarely 3-foliate leaves were observed at the lowest nodes of
seedlings and rarely at an upper node. The 9- and 11-foliate leaves
were on a few specimens of C^ lasciva from Madagascar which were
misidentified as C_^ ternatea. Clitoria ternatea has commonly 5- and

523
7-foliate leaves (ca 2-3:1 ratio). Reports of higher leaflet numbers
for this species are in error.
Double-flowered specimens of C. ternatea have been reported
infrequently in the literature since Commelin (1697). Most authors
reported these variants as cultivars which did not set seed. Many of
the specimens examined were cultivars, but additional specimens
indicate from the collector's data that spontaneous growth can occur.
Mature fruits were present in a few collections. Saroja (1961)
reported fertile pollen and larger size pollen grains, and Sen and
Krishnan (1961) reported the results of their genetic studies crossing
the actinomorphic flowers with the normal flowers (see chapter "Prior
Research: Morphology" for fuller details), thus, these variants are
capable of reproduction. Saroja (1961) is the only one to report
4-merous flowers with eight free stamens. Flowers of mounted
specimens are usually glued to the herbarium sheet, and because of
this and petal overlap, it was difficult to always count the petals.
A 4-petaled flower was not encountered on specimens examined for this
study.
There is considerable variation in a number of structures within
the species, which includes the bracteole size and shape, leaflet
shape, size, and pubescence of the upper surface, and flower size and
color. In viewing this variation on a world-wide basis, the variation
in these structures fits a somewhat consistent pattern throughout the
tropics except for Africa. African specimens present a problem
because many individuals are distinctly different from individuals in
other parts of the world, but within Africa there is a gradient of
variation in the structures noted that makes segregation difficult.

524
African populations need further study employing experimental
techniques to gain further insight on subspecific delineations.
Recognition of these African variants from the typical concept of
C. ternatea has historically produced a number of published names,
mostly at the species level. Cl i toria bracteata Poir. (1811) was
noted for its solitary flowers, large bracteoles, and twining stems.
A decade later, de Candolle (1825) reduced this species to a variety
of C. ternatea, segregated from the typical variety by subscabrous
leaflets, pubescent legumes, and purplish-white flowers. Berhaut's
recent flora of Senegal (1954) and Hutchinson and Dalziel's Flora of
Tropical West Africa (1966) do not recognize this variety. Clitori a
zanzibarensis Vatke (1878) was reported to have affinities to
C. ternatea but segregated by smaller flowers and leaflets, and the
linear-oblong shape of the leaflet. Clitoria tanganicensis Micheli
(1897) and C. mearnsii DeWild (1925) were two more names published for
narrow leaflet members. Baker (1929) published Hochstetter's
varietal name, C. ternatea var. angustifolia, and included Vatke's
name in synonymy. This variety was distinguished by its round
bracteoles and narrower leaflets. DeWild's and Micheli's names were
later included by other botanists in synonymy under var. angustifolia,
such as Wilczek (1954) who expanded on the characters used for
segregation, distinguishing this variety by the narrow leaflets,
rounded leaflet base, smaller white flowers, and smaller bracteoles.
Gillett, Pohill, and Verdcourt (1971) did not recognize this variety
in their flora of Tropical East Africa, and noted that the species
abounds in "minor variants." Clitoria albiflora Mattei (1908) was
noted as an intermediate between C^_ ternatea and (T._ tanqanicensis.

525
Matte i1s name is absent in floristic treatments (type from Somalia not
within geographical areas of recent floras) and has not been used for
identification of any specimen examined.
Bracteoles are typically 6-11 mm in diameter. Smaller bracteoles
are usually ovate with width up to 2 (3) mm narrower than the length,
becoming nearly orbicular as the bracteole size increases. In
northern and western Africa, some individuals appear with huge
suborbicular bracteoles, to 15 mm in diameter. These conspicuously
large bracteoles hide the calyx during the bud stage, and are lacking
in individuals from other tropical regions. Some members with narrow
leaflets exhibit smaller bracteole length of 4-7 mm. There exists a
gradient in bracteole shape and size, making this character unreliable
for segregation of African specimens.
Elongated, narrowed leaflets occur only in plants from the
eastern half of Africa. These leaflets are easily distinguished from
those of other members from different tropical areas by their linear,
oblong, or lanceolate shapes. These leaflets typically have rotund
bases whereas the broader, shorter, ovate to elliptic leaflets have
cuneate bases. However, there are gradients in leaf bases from
cuneate to rotund, and in leaflet size from narrow to broad, short to
long. Leaflet pubescence rarely has been used as a diagnostic
character. Narrow leaflets have microscopic uncinate hairs present
over most of the upper surface whereas broad leaflets have macroscopic
trichomes appressed and scattered over the upper surface with
microscopic uncinate trichomes (when present) confined to the midrib.
Both pubescent forms can become subglabrate with age. Again,
intermediate forms can occur.

526
Flowers typically fit within two size ranges. The common size is
4-5.5 cm, with smaller flower sizes, (2.5) 3-4 cm, occurring in many
African specimens. Flowers are typically pale to dark blue throughout
the tropics with occasionally a white-flowered form occurring. In
Africa, as flowers become smaller, the white color becomes more
prominent. The blue pigment conspicuous in the vexillum becomes
reduced over much of the inner surface until it forms a narrow
peripheral band or disappears entirely.
Examination of the African material reveals several possible
subspecific types in which certain combinations of characters
consistently appear within broad geographic areas. One combination
present is the typical group with the large blue flowers, common
bracteole type, and ovate to elliptic leaflets bearing a strigose
pubescence on its upper surface. This group is found scattered
throughout Africa and all tropical regions outside Africa, and
includes the type specimen, thus having a name (ternatea) automatically
associated with it. White-flowered members occasionally occur that
agree in all other characteristics and could be treated at a lower
subspecific level. A second group from Western Africa has white
flowers and a conspicuously tapering apex to the leaflet. This group
includes Poiret's type, and thus has a name (bracteata) associated
with it. A third group, also from Western Africa, has the typical
blue flowers and ovate to elliptic leaflets, but has large,
conspicuous, suborbicular bracteoles. No types are included with this
group, thus there is no name associated with it. A fourth group has
blue flowers which are intermediate in size, a calyx consistently at
the lower size range of the species, and small bracteoles, but it has

527
the common broader leaflet type. This group is commonly found in
Northeast Africa in Somalia, Eritria, and Ethiopia. A fifth group
has small whitish flowers, elongated leaflets typically subglabrate
above, and the conspicuously large bracteoles. This group is common
in northern East Africa, particularly in Anglo-Egyptian Sudan and
Ethiopia. It includes Hochstetter1s type and has the associated name
angustifolia. The last African group has small whitish flowers with
very narrow, elongated leaflets that have uncinate pubescence on the
upper surface, and usually small bracteoles. This group is common in
Eastern Africa from Mozambique to Kenya, and includes several types of
which the earliest name is zanzibarensis.
In making keys to these groups and applying it to the African
specimens, it became very difficult to segregate the collections into
the six groups as briefly described. Too many exceptions occurred
each time the keys were revised because the characters used proved
unworkable. Bracteole characteristics proved the most unreliable as
occasionally both the typical and large bracteole types were present
on the same individual, and more commonly on different individuals of
the same collection. The range of smaller bracteoles consistently
overlapped the lower end of the range for the typical bracteole size.
Finally bracteoles were eliminated as a segregating character, which
consolidated group three with group one and made segregation of group
five from group six difficult. Likewise group two was consolidated
with group one because of the consistent gradient in leaflet apices
from a tapering to an obtuse apex. Group four was united with group
one because the specimens of group four were collected with slightly
immature flowers. Comparison with other specimens having larger

528
flowers revealed that the flowers of group four agreed with the
smaller flowers found on the individual that also had the common,
larger blue flowers. With the elimination of bracteoles as a
diagnostic character, there remained only the leaflet width and
pubescence on the upper surface as characters which distinguished
groups five and six. These characters, however, proved unreliable
for segregation because of the continuous gradient present for the
characteristic. Thus the two groups were combined into one.
The two newly consolidated groups are treated as varieties noting
that there are a number of variants that can occur within each one.
These varieties are segregated by six characteristics, which include
leaflet shape, leaflet length/width ratio, leaflet pubescence on the
upper surface, leaflet width, flower length, and flower color. Most
specimens are easy to identify because of the particular combination
of the characteristics they possess. There are a number of specimens,
however, which exhibit intermediate characteristics, or the
characteristics are difficult to interpret. Specimens with
conspicuously narrowed leaflets (less than 1 cm) or elongated leaflets
(length/width ratio = 4-9) will be easily recognized as variety
angustifolia. Specimens with large blue flowers (4-5.5 cm) and
leaflets with a cuneate base are easily recognized as var. ternatea,
as well as those specimens with large white flowers. But specimens
with smaller leaves and smaller white flowers may present problems.
For difficult specimens, consult Figure 64 which illustrates several
examples of typical leaflets for each variety and Table 11 which
provides an artificial method of identification. The upper portion of
the table gives the typical characteristic found for each character

Figure 64. Cl i tori a ternatea - III. Var. ternatea: (a-g) typical
leaflets, x 1; (y) typical leaflets, x 1; (z) flower, xl,
(aa) fruit, x 1. Var. angustifolia: (i) flower, x 1;
(j-x) typical leaflets, x 1. ~~[Rao 66106, MO 2031229:
a,g. Nana 6256, K-59: b,f. Roberty 11217, G-hb.
Delessert: c,e,h. Heudelot 437, K-hb. Hooker: d.
Glover s.n., G-hb. Delessert: i,p-q,v. Schimper 1617,
MPU-31: j,m. Schimper 1736, K-hb. Hooker: k,w. Conrads
13439, K: r,x. Hi 1 debrant 1189, K-14: s-u. Vaugham
2548, BM: l,n-o. Wallich 5347, BM: y-z,aa.)

530

531
Table 11. A comparison between the two African varieties of C. ternatea. The
upper chart summarizes the morphological differences. The lower
chart gives the criteria for the assignment of numerical point
values for each characteristic.
CHARACTER
var. TERNATEA
var. ANGUSTIFOLIA
A)
Leaflet shape:
Ovate, Oval, Elliptic
Obovate
Oblong, Narrow elliptic,
Lanceolate, Linear
B)
Leaflet width:
12-40 mm
4-15 (Occasionally 20) mm
c)
Leaflet ratio
(Length/width):
1.2-2.2 (2.5)
(2.0) 2.5-9.0
D)
Leaflet pubescence
Short, scattered strigose
Conspicuously uncinate
on upper surface:
to glabrate
to glabrate
E)
Flower length:
(3.5) 4-5.5 cm
(2) 2.5-4 cm
F)
Flower color:
Blue to purple
(White form)
White to white with a
thin blue peripheral
band
Numerical Value:
6-11
13-18
TERNATEA
INTERMEDIATE
ANGUSTI FOLIA
Value=l
Value=2
Value=3
A)
Ovate, oval, obovate
Broad elliptic
(Large leaflets)
Oblong-el 1iptic
Oblong, Lanceolate
Oblong-lanceolate
Narrow el 1iptic (Smal 1
1eaflets)
B)
40-20 mm
20-12 mm
12-4 mm
c)
Ratio 1-2
Ratio 2-2.6
Ratio 2.6-9.0
D)
Strigose, to
Glabrate, uncinate
Uncinate, conspicuous to
scattered;
on midrib & nerves:
inconspicuous, sparse;
Uncinate absent
Uncinate with strigose
sparse, marginally
Strigose absent
E)
5.5-4.0 cm
4.0-3.5 cm
3.5-2.5 (2) cm
F)
Blue to purple
White with thin blue
White with greenish-white
with yellow or
white medially;
(White if flower
greater than 4 cm)
peripheral band;
Flower color unknown
medially

532
used in segregating the two varieties. The lower portion gives an
alternative means of identification to that of the key. A value of
one is assigned to each characteristic typical of var. ternatea and
a value of three for each characteristic typical of var. angustifolia.
Intermediate characteristics and size ranges are assigned the
intermediate value of two. In using the chart, a sum of 6-11 would
indicate var. ternatea and a sum of 13-18 would indicate var.
angustifolia. White-flowered forms of var. ternatea (flowers large,
4-6 cm) typically totaled to a value of 9-11, although this value
would drop if a value of one was assigned for the flower color when
the flower size was greater than 4 cm. In applying this artificial
system to the specimens examined, African collections fell into two
distinct groups. The data are summarized in Table 12. All
collections outside of the African tropics obtained value totals of
6-11, with few collections having a total above nine.
It is obvious from examination of the specimens that some gene
flow occurs between these two groups. Additional experimental studies
can provide more data upon which new taxonomic decisions may be made.
With the available data, it seems that it is best at this time to
recognize two African varieties as Wilczek (1954) has done because
they can be morphologically distinguished, rather than treat the
African members as one large polymorphic group as Gillett, Pohill, and
Verdcourt (1971) have done.
DISTRIBUTION (Figures 65 to 69 and 72): Clitori a ternatea is an
anthrogenic pantropical species. Originally grown as a cultivar, the
species has escaped and become naturalized throughout the lower
latitudes. The origin of the species is unknown, but examination of

533
Table 12. A comparison of the two African varieties of C. ternatea using the
artificial numerical value system. The chart summarizes the number
of collections per total point value assigned per country.
var. TERNATEA
var.
ANGUSTIFOLIA
COUNTRY
6 7 8 9 10 1 1
12 13 14
15 16 17 18
Northern Africa:
Senegal
0
2
1
3
2
0
0
0
0
0
0
0
0
Gambia
1
0
0
0
0
0
0
0
0
0
0
0
0
Portuguese Guinea
0
1
0
0
0
0
0
0
0
0
0
0
0
Mali
1
0
0
0
0
0
0
0
0
0
0
0
0
Sierre Leone
0
1
1
0
1
0
0
0
0
0
0
0
0
Ivory Coast
2
1
0
0
0
0
0
0
0
0
0
0
0
Ghana
0
3
2
0
0
0
0
0
0
0
0
0
0
Togo
1
0
0
0
0
0
0
0
0
0
0
0
0
Dahomey
0
1
0
0
0
0
0
0
0
0
0
0
0
Nigeria
1
3
3
0
0
0
0
0
0
0
0
0
0
Cameroun
1
3
1
0
0
0
0
0
0
0
0
0
0
Sudan
0
1
13
3
0
0
0
0
0
1
2
0
0
Ethiopia
0
3
3
9
1
0
0
3
0
1
1
2
0
Somalia
0
1
2
2
1
2
0
0
0
0
0
0
0
Southern Africa:
Sao Tome
0
1
0
0
0
0
0
0
0
0
0
0
0
Gabon
0
1
0
0
0
0
0
0
0
0
0
0
0
Cabinda
1
0
0
â–  0
0
0
0
0
0
0
0
0
0
Angola
0
1
0
1
0
1
0
1
1
1
3
1
0
Southwest Africa
0
1
0
0
0
0
0
0
0
0
0
0
0
Zai re
3
2
0
1
0
0
0
0
0
0
3
3
1
Uganda
0
0
0
1
1
1
0
0
1
2
2
3
3
Kenya
1
5
4
5
3
1
0
1
2
4
3
3
2
Tanzania
2
9
4
8
8
1
0
1
1
5
3
10
9
Burandi
0
0
0
0
0
0
0
0
0
0
1
2
0
Maiawai
0
0
0
1
2
0
0
1
1
3
0
0
1
Zambia
1
0
0
0
0
2
0
0
0
0
0
0
0
South Africa
1
1
2
1
0
0
0
0
0
0
0
0
0
Mozambique
1
2
1
1
2
0
0
1
0
1
1
0
0

534
the distribution maps can provide some insight. The species is
reported from the neotropics, occurring mainly along coastlines on
sandy soils, or along the shores of large inland lakes, with the
same pattern observed in much of Indonesia, Australia, and the Pacific
Islands. In Africa and Southern India, the pattern is different.
The species is commonly collected inland on sandy or clay soil,
usually in swamps, marshes, or dry lake beds. Plants outside of
Eastern Africa are somewhat consistent in their variation, whereas
within this large area, there is a great deal of variation with
gradients occurring for most of the characteristics. Probably the
species was native to the eastern half of Africa (and possibly
Southern India?) and then was spread by man to various tropical ports,
where the plant became naturalized. Altitudes are rarely given for
this species, but specimens have been collected from sea level to
occasionally altitudes of 1500 m.
KEYS TO VARIETIES AND FORMS:
1. Flowers papilionaceous; stamens diadelphous.
2. Leaflets broad, not elongate to only slightly elongate, ratio
of length/width is 1.2-2.5:1; leaflet shape ovate, oval,
elliptic, ovate-elliptic, obovate, or sometimes oblong-ovate
to oblong-elliptic; leaflet pubescence on upper surface
moderate to sparse of very short appressed hairs becoming
glabrate, uncinate trichomes lacking or sparse, confined near
major nerves; leaflet base cuneate to rotund; flowers large,
(3.5) 4-5.5 cm, typically pale blue to purple, occasionally
white (pantropics) 31a. var. ternatea

535
3.Flowers blue to purple with vexillum yellow to white
medially.
4.Peduncle 0.5-1 (1.5) cm long, rigid; flowers solitary
per axil or rarely biflowered at apex of peduncle . .
31aa. f. ternatea
4. Peduncle 1-4 cm long, base rigid and upper portion
lax, slender; flowers several per axil.
5.Flowers 3-6, racemose near apex of solitary
peduncle 31 ab. f. pauciflora
5. Flowers 1 or rarely 2 at end of peduncle, 1-4
peduncles per axil 31ac. f. fasciculata
3. Flowers white with vexillum greenish to greenish white
medially 31ad. f. a 1bif1 ora
2. Leaflets narrow and elongated, ratio of leaflet length/width
is (2) 2.6-9.0:1; leaflet shape linear, lanceolate, oblong,
lanceolate-oblong, or narrow elliptic (leaflet minute size);
leaflet pubescence on upper surface moderately dense to
scattered, uncinate, macroscopic trichomes absent; leaflet
base rotund; flowers small, (2.5) 3-4 cm, typically white or
vexillum white with thin peripheral band of blue (Africa) . .
31b. var. angustí'fo 1 ia
Flowers double, actinomorphic, (4) 5 vexillary-1ike petals;
stamens all free or some filaments free with others connate below
in 2-3 fascicles (Common cultivar, can become naturalized) . . .
31d. var. pieniflora
7. Stamens all free.
8. Flowers blue to purple
3Ida. f. pieniflora

536
8. Flowers white 31db. f. 1eucopetala
7. Stamens subpolyadelphous, some free with remainder connate
below in 2-3 bundles of 2-6 stamens . . .
31dc. f. subpolyadelpha
31a. C. ternatea L. var. terna tea
Leaflets ovate, oval, elliptic, obovate, ovate-elliptic,
sometimes oblong-ovate or oblong-elliptic, 1-4 (5) cm wide, apex
broadly acute to obtuse, sometimes emarginate, base cuneate to rotund,
upper surface pubescence strigose. Flowers blue or occasionally
white, papilionaceous, 4-5.5 cm. Legume 6-11 cm long, 8-11 mm wide.
DISTRIBUTION (Figures 65 to 69): This variety is pantropical.
31aa. ternatea L. var. ternatea f. ternatea
Flowers blue to purplish, solitary or rarely biflowered, borne at
the apex of rigid, short (0.5-1.5 cm), solitary peduncles.
DISTRIBUTION (Figures 65 to 69): This common form is widely
known as a cultivar and is naturalized in the pantropics. An
extensive number of specimens was examined, thus subheadings are used
with the countries arranged in a geographical sequence. Distribution
will begin in the neotropics and proceed eastward. Some specimens
lack flowers and may belong to f. al ba, but are cited with the typical
variety, which is more common, and marked with this symbol (*).
LOCALITY UNKNOWN. Anonymous s.n. (E-82 & 83, HAL
36874, NY,MPU-24,S-5 sh.); 28 Jul 1821, Anonymous s.n. (MPU-16);
10 Jun 1825, -oot. [?] 837 (W-mixed); hb. Delessert, Burmann s.n.

Figure 65. Caribbean American distribution of Clitori a ternatea,
subgenus Cl i tori a. Var. ternatea f. ternatea (>); var.
ternatea f. albiflora (★!; var. pleniflora f. pleniflora
wr.

538

539
(G-2 sh.); hb. Reichenbach f. (W 191795 & 359353); Kat—ng (?),
Tierpfloure (?), Anonymous s.n. (W 6834); 1821, Cottah (?) s.n.
(E-74); Cuming 616 (CGE); hb. Faculte de Medécine (MPU); cult.,
hb. Moricand s.n. (G); hortus botanicus, 10/1/1918, Mrysmann (?) 381
(U); Muhlenberg 975 (PH); 29 Aug 10, Ohlwei1er MB9 (MO); hort., hb.
Rofsl s.n. (PR); Roth s.n. (HAL-mixed); Jard. Bot. de Cho-b-ah (?),
1839, Anonymous 67 (MPU); Fran—bar Joe, unit., Frat., Anonymous 2568
(G); Korealam, 18 Jan 82, Luriakose s.n. (NY); Malines canefields,
S. Andrews, Jan 1846, McNab s.n. (E-mixed); Ostindia (?), hb. Rudlof
s.n. (*S); Achimota Gardens, Feb 1930, Irvine 1407 (E); Wullschagel
s.n. (W); spring garden, Wullschlagel 126 (W); Wullschlagel 937 (W),
hort. bot., hb. Erlangensis 1773 (M); hb. Moricand, Anonymous s.n.
(G-436, no. 1).
NORTH AMERICA
UNITED STATES. 1825, Leman I (G). FLORIDA:
ALACHUA: Gainesville, 1935, Ritchey s.n. (FLAS); l.c., 7 Jun 1935,
Ritchey 15532 (FLAS); l.c., Ritchey 117930 (FLAS); l.c., Plant
Introduction Garden, 17 Oct 1934, Ritchey s.n. (FLAS). DADE: between
Coconut Grove & Cutler, Nov 1904, Small 2225 (NY); Homestead, 5 Nov
1934, McFarlin 9855 (FLAS,NY); Exotic Garden, Miami, 26 Jul 40,
Crevasse s.n. (FLAS); Buena Vista, Miami, 18 Dec 1929, Moldenke 5461
(NY). MONROE: Key West Cemetery, Oct 1884, Curtiss s.n. (GH); Key
West, 30 Nov-3 Dec 1913, Small & Small 4876 (FLAS,G,GH,MICH,MO,NY,
PENN-2 sh.,S,U) and 16 Dec, 4971 (NY,S); Big Pine Key, 30 Jan 1940,
Martin 1311 (NY,UC). ST. LUCIE: cult., seed from Costa Rica, Legume
Plots, Indian River Field Lab W of Fort Pierce, 8 Nov 1967, Beckner

540
2184 (FLAS). ILLINOIS: COOK: Chicago Garfield Park Conservatory,
3 Aug 1939, Steyermark s.n. (F). KENTUCKY: FAYETTE: cult.,
Lexington, Short s.n. (PH). MISSOURI: ST. LOUIS: cult., Shaw School
Botany, 23 Sep 1886, B.G. s.n. (MO). TEXAS: Home, 19 Oct 1890, Joor
s.n. (MO). ATASCOSA: cult., 4.2 mi NW of Poteet, 9 Oct 1935, Cory
17159 (A). HARRIS: cult, seed from Panama, Houston, 14 Oct 1915,
Teas s.n. (MO). TRAVIS: cult. Austin, 6/9/27, Wall 383 (S); cult,
and established, Austin, 26 Aug 1943, Pruett 10^ (F,W).
MEXICO. Anonymous s.n. (M 12464); Habanna, Knechtel 729
(W). CAMPECHE: cult., Chan Laguna, 4 Dec 1931, Lundell 1022 (F,GH,
MICH,MO,NY). COLIMA: Manzanillo, 20 Oct 1910, Orcutt 4486 (F); vie.
of 1.c., 28 Nov 1925, Ferris 6043 (GH) & 6051 (GH). JALISCO: Juxpán,
Palapar Redondo, 20 m, 6 Nov 1926, Mexia 1070 (UC). NAYARIT: cult.,
San Bias, Jan 1889, Wright 1344 (F). SINOLA: 1934, Ortega 7319 (F);
cliffs, Signal Hill, Mazatlan, 30 ft, 16 Sep 1925, Mexia 34 (UC).
TABASCO: Chiltepec, 15 km ENE of Paraíso, 26 Apr 1963, Barlow 12/1
(MICHjWIS). VERA CRUZ: cult., Boquerón carretera Jalapa, 80 m, 31
Jul 1967, Rosas 663 (GH); Ciudad Aleman, 30 m, 18°41 ' N, 21 Jul 1969,
Calderon 1939 (A,MICH,MO); 16.5 km N of Tempoal, 20 Jul 1972, FAO-L 68
(UMO). YUCATAN: Johnston s.n. (F,NY); Izamal, 1888, Gaumer s.n. (F).
BELIZE (BR. HONDURAS). Corozal, 1933, Gentle 32 (F) and
Jul, 4783 (F,MICH,NY); Little Cocquericot, Belize River, 31 May 1933,
Lundell 4412 (F,MICH); Stann Creek, Robertson 121 (BM).
GUATEMALA. St. Tomas, Friedrichsthal 144 (W).
GUATEMALA: 1942, Aguilar 546 (F). IZABAL: cult., vie. Quirgua,
75-225 m, 15-31 May 1922, Standley 24325 (GH). SAN MARCOS: near
Ayutla, 45 m, 14-15 Mar 1939, Standley 68832 (F). ZACAPA: vie.

541
Zacapa, Dec 1906, Pittier 1738 (BM,NY); l.c., 200 m, 7-16 Oct 1940,
Standley 73598 (F); cult. Teculutan, 200 m, 6 Oct 1939, Steyermark
29160 (F).
E L SALVADOR, vie. San Salvador, 190- , Benson 303 (NY);
l.c., 650-850 m, 2-7 Feb 1922, Standley 20595 (GH).
HONDURAS, near Roatón on Island of Roatón, 16 Aug 1970,
Harmon & Dwyer 3929 (UMO). ATLANDIDA: vie. Tela, 8 Mar 1926,
Mitchell 15 (F,GH); cult. Lancetilla Valley, near Tela, 20-600 m,
6 Dec.-20 Mar 1928, Standley 53437 (F). CORTES: vie. La Lima, 30 m,
11-20 Apr 1947, Standley & Chacon 7273 (F). VALLE: Amapala, 10m,
11 Sep 1945, Rodriquez 3335 (F,GH,UC). YORO: Near Progreso, 30 m,
24 Jan 1928, Standley 54974 (F).
NICARAGUA. CHINANDEGA: Corinto, 15 Apr 28, Wall 383
(S). MANAGUA: Managua, 18 Dec 1940-9 Feb 1941, Grant 1077 (F).
COSTA RICA. Puntarenas, 28 Jul 1882, Lehmann 1734
(BM,G); l.c., Rowlee & Rowlee 63 (NY); cult. Limón, La Zona, 12 Dec
1945, Iruios (?) 1309 (F).
PANAMA. Apr. 1852, Andersson s.n. (S); Breyas, 8 Feb 1926,
Mel 1 (NY). BOCAS DEL TORO: Water Valley, vie. Chiriqui Lagoon, 26
Nov 1940, Wedel 1771 (GH,M0); Little Bocas, vie. Chiriqui Lagoon, 12
Jul 1941, Wedel 2511 (GH,M0). CANAL ZONE: along La Cruces Trail,
75m, 25 Feb 1935, Hunter & Allen 740 (GH,M0); Balboa, 14 Jul 1936,
Wall 383 (S). PANAMA: cerca de la Playa de Pta. Paitilla, 12 Aug
1970, Taylor & Correa s.n. (MO).

542
WEST INDIES
LOCAL I T Y UNKNOWN. Ind. Occ., Anonymous s.n. (S);
1.c., Forsstrom s.n. (S); 1.c., Swartz s.n. (S); W. Indies, 1874,
Kuntze s.n. (F); l.c., 1885, Eggers 191 (MICH); l.c., Jan 1817, hb.
Sarrdey (?) s.n. (W 40138); Ins. Anti 11., Schomburgk s.n. (HAL).
BAHAMAS. INAGUA: 3 Dec 1890, Hitchcock s.n. (MO). GREAT
INAGUA: Mathew Town, 6 Nov 1904, Nash & Taylor 1467 (*NY); l.c.,
15 Dec 1963, Dunbar 227 (A). NEW PROVIDENCE: Nassau, 29 Jan 1903,
Curtiss 54 (BM,E,F,G-3 sh.,GH,M,MO,NY,PR); l.c., 17 Feb 1905, Wight
110 (F,GH,NY).
CUBA. Anonymous s.n, (PH,U 37643A); Auber s.n. (PH); hb.
Short s.n. (MO); Read s.n. (*PH). HAVANA: prope Havana, Sep 1862,
Schatela s.n. (F); Havana, Jun 1916, Meredith s.n. (PENN,PH); Habana,
3 Nov 1864, Schott 114 (F-2 sh.); La Havana, May 1908, Terre s.n. (P);
vie. Santiago de las Vegas, 25 Sep 1904, Baker & wm Hermann 1926 (F);
l.c., cult, at station, 8 May 1905, van Hermann 910 (BM,F,NY); Vedado,
12 Apr 1914, Ekman 20 (S); Vento, 3 Jul 1904, Wilson 604 (F). ORIENTE
1856-7, Wright s.n. (GH); Santiago de Cuba near Morro Fortress,
3/5/1902, Hamilton 5 (NY); Santiago de la Cuba between Marimon and the
town, 17 Oct 1916, Ekman 7950 (S); Sevilla Estate near Santiago, trail
Chirigus to Guama R., 13 Sep 1906, Taylor 360 (PH-mixed); vie. San
Luis, 15-18 Sep 1902, Pollard & Palmer 301 (F,GH,MO,NY,PH,UC); Cobre,
28 Jan 1860, Wright 1592 (G,GH,MO,NY,PH). SANTA CLARA: Laguna de
aura, Manigua, 22 Sep 1921, Fernando 632 (GH-mixed); Rd between
Soledad & San Antonio, 18 Jul 1936, Smith & Hodgdon 3165 (A,F).

543
CAYMAN ISLANDS. CAYMAN BRAC: East end Spot Bay,
10-30 ft, 24 May 38, Kings 63_ (BM,M0). GRAND CAYMAN: Jan 1821,
Hitchcock s.n. (MO); Bodden Bay Rd, 13 Feb 1899, Mi 11spaugh 1318
(F-2 sh.); center of island, 14 Feb 1899, Mil 1spaugh 1383 (F,PH);
Georgetown, 26 Apr 38, Lewis 43 (BM.MO); l.c., Crewe Rd, 29 Jul 1965,
Brunt 1727 (BM); cliff off Jackson Point SE of Georgetown, 15-20 ft,
23 Apr 38, Kings 186 (BM,M0) arid 227 (BM,MO); Newlands Barcadere Rd,
16 Jul 1965, Brunt 1645 (BM); along track between Old Isaacs &
Winterland, 22 Apr 1956, Proctor 15182 (BM).
JAMAICA. Apr 1818, Caley s.n. (BM); Smith s.n. (PH);
Swartz s.n. (S-2 sh.); Albion, 14 Apr 1927, Orcutt 1391 (UC); E of
Airport Sta. 4, Palisadoes, 25 ft, 4 Jul 1952, West & Arnold 432
(FLAS). PORTLAND: Port Antonio near shore, 16 Jan 1906, Wight 63
(F,NY). ST. ANDREW: Univ. of West Indies campus, 600 ft, 27 Jun
1963, Crosby, Hespenheide, & Anderson 346 (F,GH,MICH,M0,NY,UC);
Kingston Streets, 12-9-90, Hitchcock s.n. (M0-2 sh., mixed); RR tracks
near Kingston, Feb-Mar 1916, Kill ip 74 (W); 7 mi E of Kingston,
Seven-Mile Beach, 29 Jun 1958, Sauer 1904 (WIS); between' Kingston &
Gregory Park along RR, sea level, 22 Feb 1920, Maxon & Kill ip 301
(F,GH,NY). ST. ANN: Llandovery to Runaway Bay, 29 Mar 1908, Harris
10370 (BM,F,NY); W end of Cardiff Hall Beach, Runaway Bay, 21 Jul
1961, Powel1 969 (MICH,U). ST. MARYS: Gray's Inn, Dec 1927, Orcutt
4294 (UC); Mouth of Rio Nuevo, ca 10 km E of Ocho Rios, sea level,
27 Jun 1958, Sauer 1876 (WIS). ST. THOMAS: mile 13.5 on Windward Rd,
300 ft, 21 Nov 1959, Adams 5489 (M); Yallahs, 30 m, 16 Jun 1952,
Araque-Molina & Barkley 22J449 (MICH).

544
HAITI. Massif de la Pelle, Port-au-Prince, Fort Bolosse,
50 in, 2 Oct 1927, Ekman 9102 (S); Pont Bludet, 11 Aug 39, Pride s.n.
(WIS); Damiln, 13 Jul 39, Pride s.n. (WIS); near Artibenite River 2 mi
W of Petite River, Mar 1925, Sweet s.n. (NY); vie. Gras Morue Road,
Port de Paix, 9 Jan 1929, Leonard & Leonard 1 1786 (GH,UC-mixed);
cult., Jeremie, 5 Mar 1942, Seibert 1800 (MO). ILE DE LA TORTUE:
slopes above Basse Terre, 21-29 Mar 1929, Leonard & Leonard 13983
(MICH,NY); Haut de Paluriste Rd, vie of La Vallée, 28 Dec 1928-9 Jan
1929, Leonard & Leonard 11591 (MO).
DPMI N I CAN REPUBLIC. Johnston s.n. (E); Najayo,
Hispaniola, Dec 1962, Angusto 449 (NY). BARAHONA: Am Duvergé Fluss,
75 m, Aug 1910, Fuertes 375 (E,G-4 sh.,GH,NY,U); prope Barahona,
26 Nov 1910, Fuertes 375 (W). SANTIAGO: Quinigua near Santiago,
29 Oct 1946, Howard & Howard 9695 (GH,MICH,MY,S); Hato del Yaque,
10 km on road to Los Matos, 200 m, 23 Mar 1969, Liogier 14534 (NY).
PUERTO RICO. Rd to El Conquistador Hotel, 22 Apr 1963,
Hagner 222 (U); Joanna-dior pres de Jerrano, 23 Feb 1865, hb.
Tuczkievicz s.n. (MPU); Montalva, 2-4 Mar 1915, Britton, Cowel1, &
Brown 4875 (NY). BAYAM0N: Santurce, 2 Jan 1938, Otero 456 (A,F,M0).
FAJARDO: ad Fajardo versus mare, 20 May 1885, Sin ten 1639 (G-2 sh.,
GH,M-2 sh. ,S). GUANICA: 22 Jan 1899, Millspaugh 728 (F,NY); Santa
Rica, 8 Sep 1914, Stevenson 2244 (F). GUAYAMA: 16 Mar 1874, hb.
Kuntze (NY). MAYAGUEZ: ad Mayaguez versus Guanajilo, 4 Jan 1885,
Sinten 825 (M,PH). PONCE: RR 2 mi W of Ponce, 6 Dec 1902, Heller
6224 (E,F,G,GH,M0,NY,PH). RIO PIEDRAS: Fuica, Buen Consejo, May
1914, Hioram (F,NY). SALINAS: Paso Seco, 28 Feb 1932, Britton &

545
Britton 9933 (NY). VIEQUES: Isabel Sequenda to Campo Cielo, 24 Jan
1914, Shafer 2358 (NY).
VIRGIN ISLANDS. ST. CROIX: Browne s.n. (*PH-frag.);
Feb 1955, Hunnewell 20115 (GH); Read s.n. (PH); Dec 1895, Ricksecker
1_52 (F-2 sh.,GH,M0,NY,UC); Bassin, Jan 1897, Ricksecker 7 (E,F,M0);
Grauge, 1800, Hausen s.n. (S); Frederiksted, 1 Feb 1913, Rose, Fitch,
& Russell 3216 (NY); River Gut, 10 Feb 1924, Thompson 688 (GH);
Armas Hope, 27 Apr 1925, Thompson 826 (PH) and 8 Oct, 859 (S-2 sh.).
ST THOMAS: B1auner 60 (G); hb. End!. s.n. (W); Nov 1880, Eggers (NY);
1885, Eggers s.n. (NY); Jan 1887, Eggers s.n. (F); Jan 1887, Eggers 87
(GH); Nov 1880, Eggers 123 (GH); ad Sugar Estate, Jan 1887, Eggers 52
(M-2 sh.,MICH-mixed,S); Nisky (or Nistry), Dec 1880, Eggers 221 (G-2
sh.,M-2 sh.,MPU,UC,W); Ehrenberg s.n. (HAL); Ehrenberg Jr. 150 (HAL);
Gutthnick 60 (BM); Westindien, Harinarson s.n. (S); Placide
Duchaissamy, Holton s.n. (NY); 28 Feb 14, Kuntze 200 (NY-mixed);
Charlotte Amalie, 17-18 Jan, Mi 1lspaugh 373 (F) and 3!T7 (F) and 409
(F) and 411 (F); 17 Nov 1905, Raunkaier s.n. (U); Schomburgk s.n.
(NY); pr. Mt. Joubert, Apr 1827, Wydler 79 (G). TORTOLA: St.
Bernard, Jun 1832, hb. Lindley, Anonymous 587 (CGE); Lower Estates,
10 m, 8 Nov 1965, D'Arcy 299 (A) and 15 Feb 1966, 722b (A) and 722C
(FLAS); Fish Bay to Road Tower, 13-17 Feb 1913, Britton & Schafer 920
(NY); Copses Experiment Station, 3 Oct 1918, Fishlock 195 (GH,PH).
VIRGIN GORDA: 13 Nov 1918, Fishlock 50 (NY).
LEEWARD ISLANDS. ANTIGUA: 20 Apr 1913, Wheeler 23
(BM); 1849, Wul1schlaegel 126 (M); Gunthorpes, 8 Sep 1937, Box 1044
(BM,F,MICH); near St. Johns, 16-17 Jan 1907, Shafer 11 (NY). BARBUDA:
Codington Village, 14 Mar 1932, Fairchild 3824 (*UC). GUADELOUPE:

546
1839, Beaupertuis s.n. (A); Duchassninq s.n. (NY,W-3 sh.); 10-4-1892,
Puss 3232 (NY); Read s.n. (PH); pres du Prestoytere, Jan 1933, Quentin
626 (A). MONTSERRAT: Windward Rd near Bethal, 9 Feb 1907, Shafer 503
(F,NY). ST. BARTHELEMY: Forrstrom s.n. (S-3 sh.); Von Goes s.n. (S).
ST. EUSTATIUS: Van Goes s.n. (U). ST. KITTS: Cauada Estate, 8 Sep-
5 Oct 1901, Britton & Cowell 232 (NY). ST. MARTIN: Bergbrant, 50 m,
2 Feb 1958, Flummel s.n. (S).
WINDWARD ISLANDS. BARBADOS: Dodds, St. Phillip,
Anonymous s.n. (F,NY); St. George, Apr 1940, Goodwis 106 (BM);
Speightstown, 23 Oct 1950, Florschutz & Florschutz 67 (U). BEQUIA:
500 ft, Jun, Joseph B253 (BM). MARTINIQUE: Brongniart 17 (S);
Flee 677 (A); Sieber 238 (W); 30 m, 11 Jan 1939, Stehle & Stehle 3484
(NY); La Vauchin, 5m, 15 Feb 1939, Stehle & Stehle 3520 (NY-mixed);
St. Anne, 3 Jul 1939, Egler 39-48 (NY); Case-Pilote, Feb 1868, Hahn
234 (BM-mixed,G-6 sh.,M,P,NY,W); Schoelcher 9 May 1934, Rodriquez 3699
(P). ST. VINCENTS: Smith & Smith 823 (GH.NY).
TRINIDAD AND TOBAGO. TOBAGO: Nov 1889, Eggers
5915 (NY); Mockley Vale, Apr 1910, Broadway 3570 (F-mixed). TRINIDAD:
Usine St. Madeleine, San Fernando, 17 Mar 1925, Broadway 5627 (F,M0,S,
U,UC).
ARUBA. Cult. Savoneta, 26 Jan 1953, Stoffers 1900 (U).
CURACAO. Realino s.n. (NY); Beehenburg, 14 Oct 1907,
Anonymous 806 (U); Ronde Klip, 28 Dec 1951, Amoldo 1809 (U); Santa
Cruz, 20-27 Mar 1913, Britton & Shafer 3009 (NY,U); Gr. Piscadera,
Hofje van der Mark, 10 Jan 1949, Burgers 100 (U); Hofje near Weis
Afo., 22 Feb 1969, Nagelkerken 115 (A,U); Klein Sta Martha, 27 Sep

Figure 66. South American distribution of C1itoria ternatea, subgenus
Cl i tori a. Var. ternatea f. ternatea (#); var. ternatea f.
albifiora (★); var. plenifiora f. pleniflora

548

549
1952, Stoffers 84 (U) and 109 (U); Cas Cora, 7 Oct 1952, Stoffers 258
(U); Aschenberg s.n. (U).
SOUTH AMERICA
ECUADOR. 1964, Delgado 17 (MO). GUAYAS: Guayaquil, 1852,
Andersson 183 (S); cult, l.c., Jul 1927, Mille s.n, (F).
COLOMBIA. Cacagual ito, 1500 ft, 15 Nov 1898, Smith 291
(BM,E,F,G-3 sh.,GH-mixed,MICH,M0,MPU,NY,PENN,PH,S-2 sh.,U,UC,WIS);
vie. Bananquilla, Jul 1927, Elias 233 (U); Cundinamarca, Tocaima,
carretera de Tocaima a Pubenza, 350-400 m, 14 Apr 1952, García-Barriga
14186 (NY); cult. Tocaima, 435 m, 20 Feb 1876, Andre 1824 (NY). Cali,
valle del Cauca, 1000m, Jaramillo 1027 (F). CALDAS: Chinchiná,
1350-1400 m, 22 Nov 1946, Cuatrecasas 23077 (F). MAGDALENA: Ciénaga,
6-13 Sep 1898, Smith 291 (NY).
VENEZUELA. COJEDES: cult. M.A.C. Station, San Carlos,
9 May 1944, Rudd 513 (VEN). MARGARITA ISLE: El Valle, 17 Jul 1901,
Miller & Johnston 55 (BM,F,GH,M0-2 sh.,NY); l.c., 15 Jul 1903,
Johnston 36 (GH). MIRANDA: Coastal Rd between Airport & Las Caracas,
5 Jun 1967, Robertson & Austin 114 (MO). SUCRE: Curnana, 50m, 29 Dec
1953, Dominquez s.n. (VEN). ZULIA: Maracaibo, 1893-94, Mocquerys 809
(*NY,S); Granja Facultad de Agronomia L.U.Z., Maracaibo, 3 May 1974,
Zambrano 105 (FLAS,LA).
SURINAM. 1841, Berthoud-Coulon 11.112 (BM) and 11.115 (BM);
Hostman 43a (S,U) and 41_6 (BM,CGE,NY,W-2 sh.) and Mar 1842, 4U (G);
Jun 1838, Splitqerber s.n. (W). SARAMACCA: Militair Hospital,
Paramaibo, Oct 1901, Went 489 (U).

550
FRENCH GUYANA, cult. Karouany, Jun 1855, Sagot s.n.
(BM.S); Konakry [?], Feb 1910, Boué 29 (G).
BRAZIL. Anonymous s.n. (MPU-18); Exped. Novara, Anonymous
s.n. (W); Blanchet s.n. (BM) and 39_ (F); Montiba, Blanchet s.n. (W);
1841, Martius 1118 (BM.G-2 sh.,HAL,M,NY,S); Widgren 1333 (S). BAHIA:
Blanchet 5 (G-2 sh.); hb. Salzmann (MPU); 1830, Salzmann 144 (G).
GUANABARA: cult. Jard. Bot. do Rio de Janeiro, Anonymous s.n. (RB
11853); prope Rio de Janeiro, Daellinger s.n. (M); cult. Rio de
Janeiro, Donvilie s.n. (G); 1.c., Tweedie 1 276 (U). MINAS GERAIS:
cult. mun. Belo Horizonte, 27 Apr 1935, Barreto 5707 (F). PERNANBUCO:
cult. Tapera, Oct 1930, Pickel 112 (F) and Dec 1930, 1J2_ (BM); 1 .c.,
Aug 1931, Pickel s.n. (F,GH,MICH,NY,PH); Pesqueira, Fazenda Peixe,
22 Feb 1962, Santos 1105, Fromm 1073, & Sacco s.n. (M). SAO PAULO:
Fazenda Santa Elisa, Campinas, 15 May 1942, Santoro 6754 (UC).
BOLIVIA. TARIJA: Caingua, 15 km N de Villa Montes, 3 Jun
1971, Krapovickas, Mroginski, & Fernandez 19448 (WIS).
PARAGUAY, cult. Jardin Botánico Asuncion, 25-26°S -
56-58°W, 20 Sep 1945, Teague 117 (BM).
URUGUAY: This author has not examined any specimens of
C. ternatea from Uruguay. Herter (1954, fas. 14, p. 453, no. 1729)
included an illustration of the species in his "Flora Ilustrada del
Uruguay," which is C. ternatea var. ternatea. The specimen used for
the illustration was probably a cultivar.
EUROPE
FEDERA L REPUBLIC OF GERMANY, cult.
Univ. Bot. Gard., Muchen, 1958, Anonymous s.n. (M,NY); hort. Bot.

Figure 67. African distribution of Clitori a ternatea,
Cl itoria. Var. ternatea f. ternatea (#J;
f. a 1 bif 1 ora (^‘JT var. ternatea f. paucif
angusti fol ia (★).
subgenus
var. ternatea
ora (•&); var.

552

553
Monacensis, 18.., Anonymous s.n. (M); hort. ad Moers, hb. Grimm, 1806,
Salzwedel s_. n. (M).
NORTHERN AFRICA
CAPE VERDE ISLANDS. 1895, Cardoso 146 (*K).
SENEGAL. Sparrman s.n. (S-4 sh.); St. Louis, 1837,
Heudelot 538 (G); Dakar, 22 Apr 1829, Perrotet 223 (BM,W).
GAMBIA. Kuntaur, 1948, Ruxton 20 (K).
PORTUGUESE G U I N E A. Arredores de Bissau, 4 Jan
1963, Guerra 3980 (K).
MALI (FRENCH SUDAN). Nioro, 15 Oct 1954, Davey 106 (K).
SIERRA LEONE. 1854, Daniels s.n. (BM); Rakupr
Station, Jul 1950, Macluskie 16_ (E).
IVORY COAST. Adiopodume, 26 Sep 50, Roberty 12217 (G);
Ourne, 5 Mar 51, Roberty 14099 (G,M0); Abidjan, 8 Dec 54, Roberty 15794
(G-2 sh.); de Turois, Abidjan, 29 Dec 64, Toilliez 234 (*G).
GHANA (GOLD COAST). Received from W Africa Cacao Research
Institute in 1947-48, Anonymous 67^ (E); Kwadaso Agrie. Station, 29 Apr
1963, Obeng-Darko 5050 (K). GOLD COAST: S. Gold Coast, 11 Jun 1924,
Cheeseman s.n. (BM); Keta Jelokofe Road, 15 Nov 1957, Enti 609 (K).
NORTHERN: Gambaga, Mamprise Dist., 22 Mar 1958, Hepper & Morton
A3135 (K).
TOGO: Derriere Lome, rive de la lagune asséchée, 15 Jun 1931,
Duffour 6607 (G-3 sh.,MPU).
DAHOMEY: Cercle de Zagnanado, pays de Hollis, entre Abbo
& Massi, 6/2.1910, Chevalier 22990 (K).

554
NIGERIA. NORTHERN: 20 mi S of Gombe, 10 Oct 1921, Lely
664 (K); Nupe, 1906-07, Yates s.n. (BM); River Benue, Sep 1910, Talbot
s.n. (BM,M0); within 50 mi of Maifoni, Bornu, 6 Jul 1909, Parsons s.n.
(K). WESTERN: Abbeokute, 1859, Barter 3394 (K); Forest Hill, near
Forest Hgrs. Office Compound, Oyo, 20 Apr 45, Onochie 7750 (BM);
Lokofa Dist., Parsons s.n. (K).
CAMEROUN. CAMEROON: cult. Victoria Bot. Gardens, 22 Jan
60, Ogu 98 (E); N. Cameroons, Oct 1910, Talbot s.n. (BM). MAROUA:
pres Doumrou, 8 km SSE de Kaele, 31 Aug 1964, Letouzey 6579 (K);
collines de Mogazang, 10 km N de Maroua, 23 Sep 1964, Letouzey 7038
(K); 20 mi von Mora, Aug 54, Vogel 42 (BM).
CHAD. Baguirmi & region du lac Fittri, Ardebe, 31 Aug 1903,
Chevalier 9662 (K).
ANGLO-EGYPTIAN SUDAN. Nubia, 1839, Paul s.n.
(M) and 280 (*M). AL ISTIWA'I YAH: cult. Juba, Bahr el Gebel, 11 Jul
1929, Simpson 7354 (BM); Logire, Torit Dist., 33°1'E-4°l1N, 950 m,
6 Jun 49, Jackson 800 (BM). AN NIL AL AZRAQ: Tozi, Blue Nile prov.,
2-10-1952, Lea 12 (K); Jebeleire, White Nile, 1909, Broun 910 (BM).
AS SUDD A'ALI AN NIL: Jongols Post, 33°20'E-l0°8 'N, 21 Sep 1951,
Sherif A4012 (K). BAHR AL GHAZAL: Wau, 13 Feb 1937, Myers 6297 (K).
DARFUR: Darfour, Purdy 34 (K). KASSALA: S of Tokar Delta, Aqiq,
50 ft, 12 Apr 49, Bally B7009 (G,K); Has Tauranil, Red Sea, Feb 69,
Lord s.n. (K); Diokirhten, Nubische Kiiste, 1865, Schweinfurth 1752
(BM); maritan plain between Port Sudan and Suakin, Red Sea, 29 Nov
1927, Newberry 366 (BM); flats near Port Sudan, Red Sea, 25 Jan 1912,
MacDoregal & Sykes 19 (BM). KORDOFAN: Nuba Mts., 1500 ft, 23 Nov
1921, Lugard 22 (BM); Abou Gureis, Nurschagga, Kordofan, Aug 1875,

555
Pfund 141 (F); ad pagum Cordofanum Milbes, inter Nubicum, 4 Dec 1839,
Kotschy 287 (BM,E,G-3 sh.,GH,K,M-2 sh.,MO-mixed,MPU,NY-2 sh.,S,W.
non: HAL,MO).
ETHIOPIA (ABYSSINIA). Abyssinia, Salt s.n. (BM); l.c.,
Sep 1853, Schimper 387 (K,MPU); l.c., Schimper & Kotschy 6635 (E);
in valle Aguar, 8 Apr 1839, Schimper 1009 (BM-2 sh.,G-2 sh.,HAL,K,M,
NY-2 sh.,S,W); Donver et Rar el Fil, Wurtenburg s.n. (W-mixed).
ERITRIA: Massawa, 17 Jan 1963, Hagos 144 (K); Massaua, Lamphar,
Dec 1872, Hildebrandt 728 (BM,PH,W-2 sh.); Massaua pres qu'ile d Abd
el Rader, 10 Feb 1892, Schweinfurth & Riva 207 (G); Ghinda, hauteur
de Donkollo, 950 m, 14 May 1892, Schweinfurth & Riva 2135 (G,K);
Ghinda-Baresa, 1000 m, 17 Feb 1893, Terracciano & Pappi 37 (G);
Amasen, Lungo il terrente Fil-Fil, 23 May 1902, Pappi 5440 (BM-mixed;
probably belonged to 5441 collection) and 5441 (BM); Keren, Boyas,
15 Aug 1861, Heudner 41 (K,S). GAMU-G0FA: Woito River, ca 5 mi from
junction with Sagan, 12 Aug 1969, Gilbert & Gilbert 1553 (K). HARAR:
Hardegger s.n. (W); Kallafo, 23 Aug 1960, Anonymous, IECAMA no. 1-81
(K). SHOA: Ghibbie Valley, 37°35'E-8°l51N, 1225 m, 10 Apr 1961,
Brehme in Mooney hb. 9063 (K).
S 0 M A L I A. lungo d'Daua, horti Romoni, 24 Apr 1893, Riva
1081 (G); val lata del Neb Rusfidi, 19 Feb 1890, Riva 924 (G). LOWER
GIUBA: Basso Giuba a Gel ib, Mobilen, Feb 1940, Agrario 23 (W).
NORTHEAST: Las Anod, 7-2-1942, Peck Y124 (K); between Bihen & Las
Anod, 27 Oct 44, Glover & Gilliland 140 (BM,K). UPPER GIUBA:
immediate vicinity of Juba River, Bardera, 500 ft, 30 Sep 1953,
Bailey B9042 (K,UC).

556
SOUTHERN AFRICA
SAO TOMÉ, cult. Cidade, 24 Feb 1949, Santo 212 (BM).
GABON. 5 Nov 1909, Thollon 131 (K).
CABINDA. Cabinda, 1915, T.T. 6442 (BM).
ANGOLA. LUANDA: Ambriz, cult, from seed sent from Gaboon,
May 1873, Monteiro ^ Monteiro s.n. (K).
SOUTHWEST AFRICA (NAMBIA). Runtu, garden escape,
Okavango native territory, 1-12-55, Winter 3773 (K,M).
ZAIRE (BELGIUM CONGO). EQUATEUR: Eala, Coqui1hatvi11e terr.,
Nov 1923, Goossens 4469 (K). LEOPOLDVILLE: Bas Congo, 1921,
Claessens s.n. (M); I.N.E.A.C. Luki, Bas Congo, L'ukula terr., 1959,
Flamigni 10731 (UC); Boma, 4 Jul 1895, DeWe'vre 94 (K). ORIENTAL:
Brousse Sta., Doruma, Oct 1936, Graer 674 (BM,K).
UGANDA. EASTERN: Mjanji, Sarnia Bugwe, 3300 ft, 10 Nov 52,
Harker 12 (K). NORTHERN: Kangole, Karanoja terr., 4000 ft, 22 May
40, Thomas 3446 (K).
KENYA. Kiangamini Island 55 mi NE of Lamu, 5 ft, 26 Jul 61,
Gillespie 54 (K); Tana River, 1600 ft, 5 Apr 1910, Battiscombe 267
(K); Malindi, 28 Jul 58, Hacker C15 (K); Kibarani, 21 Apr 45, Jeffery
K2_(K); Mombassa, 1909, Horn s.n. (W); Jilore, Giryana, Mombassa, 23
Nov 1923, Gore 22 (BM); Shimba Hills, Kwale dist., 450 ft, 10-5-1968,
Magogo & Glover 1037 (K); Voi, 2000-3000 ft, May 1931, Napier 958 (K);
Teita Hills, 2000 ft, 20 Mar 1906, Grenfeld s.n. (K); Tsavo Nat. Park,
Apr 1965, Hucks 325 (K). CENTRAL: Embene, near Embu, 3600 ft, Graham
1736 (K). NORTHERN: Marsabit, 4000 ft, May 1959, Adamson 5 (K); ca
12 mi NNE of Kangetet, 1900 ft, 25 May 1970, Mathew 6389 (K); Lake

557
Rudolf, 14 Sep 1899, Wei 1 by s.n. (K); Della, N.F.D., 23 Jun 1951,
Kirrika 74 (K). SOUTHERN: Kibwesi, Ukambani, 1000 m, 4 Dec 1905,
Scheffler 36 (BM,E-2 sh.,G,K,PR,S,W).
TANZANIA (TANGANYIKA). CENTRAL: plains between Babati &
Dodona, 1 Mar 38, Martunes 114 (BM); Ikowa Dam, c 60 km E of Dodona,
900 m, 29 Jul 1970, Thulin & Mharo 540 (*K); road from Chunya to
Itigi, Itigi dist., 1500m, 26 Jan 1968, Richards 22959 (K); Mpwapwa,
31 Dec 1946, Rensberg & Musomi 582 (K). EASTERN: Bagamoyo, sea
level, 9 Dec 1952, Curle 446 (BM); Kilosa, 8 Mar 1926, Burtt 131 (BM);
osticher Mahenge Bezirkzwischen Mahenge-Plateau und Zusammenfluss von
Kilombero (Ulanga) Luwegu zum Rufiji, 400 m, 8 Jun 1932, Schlieben
2287 (BM,G; non K,M,S). SOUTHERN: 160 km W Lindi, Bezirk Lindi,
450 m, 22 Apr 1935, Schlieben 6358 (BM,G,M,S). SOUTHERN HIGHLANDS:
Mwayembe Swamp, Iringa Dist., 2700 ft, 28 Oct 1970, Richards &
Arasululu 26337 (K); Ruaha Nat. Park, 9 km NE of Msembe, 790 m,
26 Jan 1972, Bj^rnstad 1271 (K); Ruaha N. Park, track to Msonya,
840 m, 12 Jan 1966, Richards 20978 (K,NY,S): near Igawa, Mbeya
Dist., 1350m, 18 Jan 1957, Richards 7882 (K); Igawa-Mberga road,
1800 m, 4 Mar 1961, Richards 14228 (K). TANGA: Maganga, 25 Jan 1967,
Rykebasch M371 (K); Gereza East, Mnyusi, Korogwe Dist., Sep 1954,
Semsei 1797 (K). WESTERN: East Rukwa, 3000 ft, Apr 1938, Martunes
263 (BM). ZANZIBAR: in Garden gezogen, Nov 1873, Hildebrandt 1188
(BM,K,W-2 sh.).
ZAMBIA (NORTHERN RHODESIA). SOUTHERN: Victoria Falls
Station, Livingston Dist., 30 Mar 1956, Gilger 610. CENTRAL: Feira,
15 Dec 1970, Fanshawe 11025 (K).

558
RHODESIA (SOUTHERN RHODESIA). Sabi Valley Experiment
Station, Chipinga Dist., Feb 1960, Soane 263 (K); cult., Marandellas
Dist., 19 Mar 1961, Crosby 968 (K).
SOUTH AFRICA. TRANSVAAL: 24 mi ESE of Skukuza,
Nelspriut Dist., 1600 ft, 17 Jan 1953, Acocks 16735 (K); Grasskop,
28 Nov 1917, Pole-Evans 16915 (K); Kruger Nat. Park, Numbi gate,
2000 ft, 14 Nov 1951, Schijff 124 (K); Kruger Nat. Park, 17 Jan 1953,
Schijff 1765 (K); Komati Poort, 1000 ft, 18 Dec 1897, Schlechter 11825
(BM,E,G-5 sh.,K,PR-2 sh.).
MOZAMBIQUE (PORTUGUESE EAST AFRICA). LOURENCO MARQUES:
Lourenco Marques encosta da Ponta Vermelha, 18 Jan 1960, Lemos &
Balsinhas 4 (BM,K-2 sh.). SOFALA: Sofala, sea level, 24 Sep 1961,
Methuen 244 (K). SUL DO SAVE: Villa Inhambane, Dec 1936, Sousa 1943
(K,MO); Limpopo Valley, Sousa 701 (K). ZAMBEZIA: Mocuba, posto
Agricola, 16 May 1949, Barbosa & Carvalho 2645 (K).
MALAWI (NYASALAND). CENTRAL: Dist. Lilongwe, 18 Aug 1964,
Salubeni 339 (BM.MO). NORTHERN: Rhumpi, Dec 1952, Craske 1 (BM);
Rhumpi Dist., St. Patrick's Parish, 3000 ft, 9 Feb 1969, Pawek 1702
(K).
WESTERN INDIAN OCEAN
A L D A B R A ISLANDS. WEST ISLAND: 200 m S of
manager's house, 10 Jan 1972, Wood 1634 (K).
SEYCHELLES ISLANDS. 1841, Pervillé 124 (K).
MAHE ISLAND: Port Victoria, 1867, Neville s.n. (K); Northwest Bay,
1 May 1963, Sauer 3678 (WIS).

Figure 68. Asian distribution of Clitori a ternatea, subgenus
Cl i tori a. Var. ternatea f. ternatea (♦); var. ternatea
f. albiflora (^*77 var. ternatea f. fasciculata 7^)-

560

561
MASCARENE ISLANDS. MAURITIUS (ISLE DE FRANCE):
hb. Cambessedes s.n. (MPU); Commerson s.n. (RB; *G); cult., Bouton
s.n. (CGE); 1864, Bouton s.n. (K); Sieber 189 (HAL,M,PR,W); 1826,
Sieber 238 (G-2 sh. & mixed,HAL,W); Hilsenberg 1352 (PR). REUNION
(ILE DE BOURBON): 1846-52, hb. Delessert, Voyage Boivin 1463 (G);
Commerson s.n. (G); 1833, Goudot s.n. (G-mixed); 1826, Perrottet s.n.
(G).
RODRIGUEZ ISLAND. Rodrigues, Venus Expedition,
Aug-Dec 1874, Balfour 1123 (BM,K,PH).
ASIA
IRAQ. Ghozi Garden, Baghdad, Anonymous, Rustin Exp. Farm
10087 (K).
SAUDI ARABIA. Arabia, Carter s.n. (E). ADEN:
Grounds of Federal Regular Army Officer's Mess, Khormaksar, sea level,
12-3-1967, Smith & Lavranos 5_ (K); Fort Morbut, Waring 145 (K); A1
Huseini, N of Lahej, escapee, 450 ft, 25-30 Nov 1937, Scott & Britton
273 (BM). YEMEN: Hille, Gebel Bura, 31 Dec 1888, Schweinfurth 260
(G-3 sh.); Horjeilah, Wadi Chaba, 500 m, 1887, Defiers 145 (G).
PAKISTAN. SIND: Karachi University Botany Department,
9-4-1969, Laidi 60_ (MO); 1 .c., 5-4-67, Husain s.n. (NY); cult.
Karachi, Oct 51, Jafri s.n. (BM).
INDIA. Anonymous s.n. (MPU-17); Banks s.n. (W); Boquert s.n.
(G); Sharp s.n. (W); Thunberg s.n. (S); Indes oriental is, Dec 1830,
Dalhouise s.n. (G); 1. c., Koenig s.n. (BM); 1. c., Heyne s.n. (BM);
1 .c., hb_^ Porsen..berg 1352 (W); 1 .c. , Oct 1 796 , Rattler 50 (M); NW
India, Stewart s.n. (G) 6 mi on the Dhurmarroa (?), E India, Aug,

562
Ritchie U2_ (E); Bejwapalli [?], 26 Nov 1913, Haines 5244 (*K);
Central India, King s.n. (*W); W. Himalaya, Duthie s.n. (K);
Tittaghur, 16 Jan 1912, Millspaugh 3116 (F). ANDHRA PRADESH:
Sirkakulam, 150 m, 8-9-62, Balatuishnan 1010 (*FLAS); Ellora Hills,
Deccou, Ralph 250 (G). ASSAM: Cachar, Sep 1873, Keenan s.n. (K).
DELHI: Delhi, Aug 1952, Lata s.n. (UC); l.c., 23°38'N-77°12'E,
Johri s.n. (FLAS); l.c., Roshnara Gardens, 7 Nov 1954, Kunar s.n.
(NY); l.c., Dec 1954, Dewan s.n. (U); University campus, Delhi, 28
Oct 1955, Sawhney 89 (WIS); University ridge, Delhi, 7 Aug 1955,
Ma1 i 1 e 41 (F). GUJARAT: Ahmadabad Dist., Bombay, 1922, Lascton (?)
3316 (CGE). KARNATAKA: Ramagiri, Closepet, 25 Oct 1944, Venkatesh
s.n. (S); prope urben Mangalor, terr. Cañara, 1847, Hohenacker 117
(BM,G,K,M,W): Yalrigi, Dharwar Dist., 2000 ft, Feb 1919, Sedwick 5351
(K); Darwar jungles, Aug, Wight 870 (mixed with f. albiflora: BM,CGE,
E-mixed,G-hb. de Candolle, NY. non: E,G,HAL,S). KERALA: Cannanore,
1833, Campbell s.n. (E); Trivandrum, 18 Feb 1934, Erlansen 5404A (W).
MAHARASHTRA: 1-11-1959, Balapure 66149 (*M); Chanda Dist., 26 Nov
1913, Haines 5900 (*K); Raburi, Ahmadnagar Dist., Aug 1919, Nana 6256
(K); cult. Bibi, Khed taluka, Poona Dist., 17 Sep 1960, Rao 66106 (BM);
Poona, College Bot. Garden, Nov 1888, Anonymous s.n. (E). ORISSA:
Palaspal, Keonjhar State, 2 Oct 1939, Mooney 1158 (K). PUNJAB:
Drummond 24631 (K). TAMIL NADU: Maisor & Carnatic, 1860, Thompson
s.n. (BM,CGE,G,M,NY,S,U,W); Palar, South Arcóte, 7 May 1907, Barber
8307 (K): Hosur Taluk, Salem Dist., 3000 ft, 9-6-32, Kanothyeshoda
229 (A,NY): Periakulam, Pulney Hills, Madura Dist., 1000 ft, Nov
1925, Anglade s.n. (G,UC); Pulneys, Evershed s.n. (BM); Paruanadussa
(?), Madras (annotated), Jan 1836, Wight 733 (K,M-2 sh.,NY,S,W-2 sh.);

563
Madras, 21 Mar 1918, Rich s.n. (K). UTTAR PRADESH: Dehra Dun, 2200
ft, Aug 1898, Gamble 27155 (K); Moradabad, Jul 1843, Thomson 209 (BM,
E,K); Dudhwa Range, upper Gangetic Plain, Dist. Kheri (Oudh),
11-4-1898, Inayat 21469 (K). WEST BENGAL: Bengal, Anonymous s.n.
(F-2 sh.); l.c., Hooker & Thomson s.n. (CGE,F-2 sh.,M,NY-mixed,
U-mixed, W141756 & 141757; non: NY,S,U,W18672); Calcutta, 1830,
C.B.D. s.n. (E-79); l.c., Dalhousie s.n. (GH); l.c., Mar 1878,
Gamble 5836A (K); cult., hort. bot. Calcutta, 1865, hb. de Candolle
s.n. (*G); l.c., Griffith s.n. (CGE); l.c., Wallich 5344H (BM,CGE-2
sh.,E,G,not K); Bengal, Gangetic Plain, Dist. Houghly, 9 Feb 1877,
Watt s.n. (E) and 14 Nov 1878, Watt 300 (E).
BANGLADESH. DACCA: 25 Mar 1863, Clarke 6691 (BM).
S R I LANKA (CEYLON). 1773, Burmann 74 (G) and 75;
Burmann 109 (G); 1841, Kelaarb 61 (G-mixed); hb. Schmiedel s.n. (*M);
Thomson s.n. (K); Ambagamawa, ca 3000 ft, 30 Sep 1930, deSilva 87 (NY);
Taldena, Mar 1890, Deschamps 2 (G); Akuramhadda, Aug 1890, Deschamps 2
(G); Ruhuna Nat. Park, Block I, Patanagala, 28 Oct 1968, Wirawan 699
(MO-2 sh.). CENTRAL: Peradeniya, 1572 ft, 9 Dec 1930, deSilva 98
(NY); Galagedera, Kandy Dist., 19 Mar 1970, Rudd 3316; Kandy, near
hospital, 3 Feb 1970, Rudd & Balakrishnan 3064 (MO). EASTERN:
Trincomalee, Aug 1860, Anonymous s.n. (E-81); l.c., Aug 1860, Dunbar
s.n. (E). NORTHERN: Jaffna, 1941, Bates s.n. (GH). SOUTHERN: ca
1.5 mi beyond Tissa resthouse on road to Kataragama, Hambantota Dist.,
9 Feb 1970, Rudd 3106 (MO,NY). WESTERN: Kochikade, 10 Apr 1931,
Simpson 7959 (BM); Puttalam Dist., shrub W of Pomparippa in sanctuary,
2 Jun 1972, Maxwel1 & Jayasuriya 804 (MO).

564
ANDAMAN ISLANDS. PORLOB ISLAND: Feb-Mar 1934,
Ram 3742 (E).
BURMA. Brandis 1307 (CGE-mixed). PEGU: Rangoon, 18 Jul
33, Dickason 5696 (W); Victoria Lake, Ragoon, 26 Sep 1925, Valentin
s.n. (S); Thandaung, Rangoon, Oct 1934, Dickason 5202 (W); Oukkan,
Pegu, McLel1 and s.n. (*K). UPPER BURMA: MAGWE: Minbu District,
Nov 1902, Mokim 546 (G); 1.c., 10 Aug 1909, Lace 4919 (E,K).
MANDALAY: Kyaukmagyi, Meiktila District, 28 Oct 36, Smith 13750
(K); Mandalay, 10-11-50, White 42 (UC).
THAI LAND (SIAM). 1872, Bradley 99 (GH-mixed,PH-mixed,UC);
Anonymous s.n. (W). CENTRAL: Authaya, 23 Nov 56, Bolinger s.n.
(NY,U); Bagkok, 1929, Wall s.n. (S); Bangkok, 1899, Zimmermann 5
(M,U-mixed; non BM,G,PR,U); Tha Luang, 75 mi N of Bankok, 1924,
Smith s.n. (GH). NORTH: Chieng mai, 1000 ft, 16 Jun 1912, Kerr 1530
(BM) and 300 m, 1530B (BM,E); Prae, 27 Oct 1929, Franck 103 (S); Den
Chai, Prae, 1 Oct 1929, Franck 113 (GH).
MALAYS I A. Pulo Penang, Beck s.n. (W); Pulo-Pinang, 1835,
Delessert s.n. (G); Malacca, Griffith s.n. (CGE) and 1728 (K);
Singapore, 1886, Beck s.n. (W); Singapore, Penang, 1830, Walker 117
(G); Combang Talong, Burmann s.n. (G).
LAOS. Bassac, 1866-1868, Thorel s.n. (F).
SOUTH VIETNAM. COCHINCHINE: 1862-1866, Thorel s.n.
(A,F); cult., Germain s.n. (*F); f1uv. Mekong, Dec 1867, Pierre s.n.
(E,F) and Dec 1868, Pierre s.n. (G); hort. Bot. Garden Saigon,
Anonymous s.n. (GH); l.c., Dec 1868, Pierre s.n. (A); ad Thu Dau Mot,
Aug 1867, Pierre s.n. (F,G); Long-Than, 1862-1866, Thorel 37 (BM,E);

565
Tourcham, Phanrang prov., May 1918, Kloss s.n. (BM); village So Phusc
Duong, prov. So Ninh Thuan, Aug 1935, Petelot 2168 (GH).
CHINA. Henry s.n. (*NY). HONG KONG: 13 Sep 1893, Bodinier
668. (E); pres Bethanie, 13 Sep 1893, Bodinier 89988 (MPU); 6 Jul 1891,
Delavay s.n. (K-2 sh.); Forbes s.n. (BM). KWANGTUNG: cult. Canton,
13 Oct 52, Anonymous 3552 (U 64741B).
T A I W A N (FORMOSA). Apr 1934, Yamada s.n. (S); Takao, 11 Sep
1930, Tanaka 5440 (BM,NY); 1.c., Henry s.n. (NY); Takao Gau, Apes
Hill, 5-6-1912, Price 588 (K); Boryo to Kurara, Hoshun prov., 15 Nov
1918, Wilson 10995 (A).
PHILIPPINES AND INDONESIA
PHILIPPINES, de Samboangeon (?), 1819, Perrottet s.n.
(G). LUZON: Cuming 616 (BM,CGE,E,G-2 sh.,NY,W); Pangasinon, Dec
1907, Ramos 4951 (GH); Manila, Blanco 394 (GH,NY,W; non: A,BM,F,M0,
W); l.c., Oct-Dec 1904, McGregor 46 (NY); l.c., Oct-Nov 1903, Merri11
3439 (NY); l.c., 5 Nov 1911, Mi 11spaugh 2814 (F); l.c., 1879,
Rothdauscher s.n. (M-2 sh.); l.c., Carborn Swamp, May 1907, Da^ 42^
(MO); Manila, Malate, Oct-Dec 1925, Buensuceso s.n. (PH-mixed); prov.
Cagayan, Mar 1909, Curran 16743 (E); Los Banos, prov. Laguna, 15 Jan
1905, Williams 2057 (GH,NY-2 sh.) and 3071 (NY); Pansol, Laguna,
19 Apr 1913, Gates 7194 (F); prov. Laguna, Jul 1964, Quisumbing 1996
(BM,G,GH); Bauang, prov. Union, Feb 1904, Elmer 5576 (G-2 sh.,NY,PR).
CEBU: Maelan, Jan-Feb 1875, Moseley s.n. (BM). MINDANAO: Davao,
Dist. Davao, Mar 1904, Copeland 449 (NY); Mati, Dawan streams, Davao
prov., Mar-Apr 1927, Ramos & Edano 49125 (S,UC-mixed).

566
BORNEO. 1884, Grabowsky s.n. (BM). BRITISH NORTH BORNEO:
Kudat, Mount Kinabalu, 20 Dec 1915, Clemens 9553 (BM,GH); Banquey
Island, 20 ft, Jul-Sep 1973, Castro & Meleqrito 1587 (G,UC).
CELEBES. Bonerate, 30 m, 8-5-1913, Leeuwen 1434 (U);
Lamala, Tongke, Sep 1919, Kaudern 371 (S); Boulaauy-Mongondon,
Motoukat, N. Celebes, 23 May 1917, Kaudern 116 (S); Boeton, 50 m,
12-10-1929, Kjel1 berg 143 (S); Bouthain, Jan 1930, Kjel 1 berg 3092 (S).
SUMATRA. Korthals s.n. (K,U,W). EAST COAST: Asahan, 20
Jun 1918, Bartlett & LaRue 168 (GH); Ria na Paso, between Djoema
Tombak & Taratak, Tanah Djawa, Simeloengoen, 7 Jun 1923, Bartlett 8278
(NY); Goenoeng Soeasah, near Rantau Paraput, Bila, 27 May 1932, Toroes
2343 (GH,UC).
CHRISTMAS ISLAND, escapee at drumsite, South
Point, 50-800 ft, 28 May 1968, Powell 13 (K).
JAVA, signatures illegible (W 359350, 2 collections); Burmann
256 (G) 1768-71, Cook's 1st Voyage, Anonymous s.n. (BM); Hoffmansegg
s.n. (HAL); Horsfield 12L (CGE); 1 759, K1in ting s.n. (G); Reinwartz
s.n. (S); hb_^ Ventenant s.n. (G); 1838, Webb s.n. (G); Zol 1 inger 45
(BM,CGE,G-4 sh.,S); SE Java, 1880, Forbes 1230C (BM); W Java, Apr
1818, hb. Tandry s.n. (W); hort. bot. Bogor, 5 Apr 1893, Hal 1ier D340
(G); l.c., 1872-3, Wawra 1331 (W); Batavia, 27 Mar 1839, Anonymous 72
(E); l.c., Horsfield s.n. (K); ..elknede (?), 1902, Backer s.n. (U);
Buitenrorg, Tjidepet, 28 Apr 1922, Brink 1056 (U); Semarang, 20 Nov
1882, Cooper s.n. (U); furn culturuuschool, Malang, 400 m, Groenhart
70 (U); Turrchen ruigto Waterval Baong by Lawang, 0. Java, Groenhart
112 (U); Madoera pris dans de..(?)... cultive in Buitensorg, 31 Jan

Figure 69. Australasian distribution of Clitori a ternatea, subgenus
Cl i tori a. Var. ternatea f. ternatea (#~}T~

568

569
1905, Hochreutiner 2803 (G); Lengi teleng, 1837, Hoffmansegq 125 (G);
Suro kerto, Horsfield s.n. (BM).
TIMOR. Anonymous s.n. (G-184; K-137); hb. Baner 88 (W);
1830, Gaudichaud s.n. (G); hb. Moricand s.n. (G); Kollarro seaside,
Wahby (?) 254 (BM); 1789, Nelson s.n. (BM).
W E T A R ISLAND. Nusa Tenggara, Meta Lakela to Ilwaki,
1-100 m, 8-9 Apr 1939, Bloembergen 3672 (NY).
MOLUCCAN ISLANDS, cult., Anonymous s.n. (W).
TANIBAR ISLANDS: Jamdena, Laurang, 11 Apr 1956, Waalkes 3346 (A).
AUSTRALIA AND PACIFIC ISLANDS
AUSTRALIA. New Holland, H.H.J. s.n. (CGE); Rositer's,
Mar 1910, Morrison s.n. (E). NORTHERN: Humpty Doo Experimental Rice
Farm, Darwin, 23 Dec 1971, Redhead B192 (K); Arnhem Land Aboriginal
Reserve, Nightcliff, Darwin, 12°22'S-130°53'E, 22 Mar 1948, Specht 31
(A,K). QUEENSLAND: Gaynah below Duke & Duchess Hills, Burnett
District, 10-5-1956, Fox s.n. (*NY); Townsville, N Kennedy Dist.,
13 Jan 1968, Henderson H324 (K). WESTERN AUSTRALIA: Kalumburu
Mission, Kimberleys, 10-12-52, Broadbent 52 (BM).
MELANES I A. NEW GUINEA: Northeast New Guinea: Kiapit,
Morobe Dist., hb. Forest Dept., Anonymous 2646 (K); Markam River, Lae
Subdistrict, Morobe District, sea level, 26 Nov 1967, Coode 32579 (E);
Erap, Morobe Dist., Sep 1953, Womersley 5383 (A,K). Papua: Basima,
Cape Vogel, Baniara Subdistrict, Eastern Dist., 100 ft, 11 Sep 1954,
Saunders 125 (A,K); ca 1 mi N of Rigo, Central Dist., 20 ft, 13 Aug
1962, Schodde 2774 (A,E,G,K); Port Moresby, 50 ft, 10-4-35, Carr 11812
(BM,K); ca 0.5 mi E of Delena Village, Kairuku Subdistrict, Central

570
Dist., 50 ft, 9 Aug 1962, Darbyshire 790 (A,K); Daru Island, Western
Division, 9 Mar 1936, Brass 6294 (A.BM). SOLOMON ISLANDS: Tenasn,
coconut plantation, Tenasn, Guadalcanal, sea level, 13 Nov 1954, Brown
W117 (BM). NEW CALEDONIA: in garden, Noumee, Aug 1884, Grunow s.n.
(W); Noumie, 4 m, 23 Feb 1950, Guillaumin 8546 (A); Noumea, Mar 1914,
Franc 820A (G,GH,K,NY); Caricaté St. Vincent, 13 Aug 1924, Daniker 764
(S). FIJI ISLANDS: 15 Jan 45, Parham 5158 (A); E shope of main
ridge, Karo, 100 m, 29 Jan-5 Feb 1934, Smith 1028 (NY); Lautoka, 10
Oct 20, Greenwood 145 (K); Sawaieke, shore of Herald Bay, Ngau, 0-30 m,
24 Jun-5. Jul 1953, Smith 7894 (NY,S,UC); Fidschi, Suva, Aug 1932,
Meebold 16504 (M); Water Bay, Lausi R. near Suva, prov. of Rewa, Viti
Levu, Apr 32, Parham 51 (BM); Hot Springs, Tavua, Nadarivatu, Viti
Levu, alt. 100, Oct 1907, Gibbs 878 (BM).
MICRONESIA. MARIANA ISLANDS: Guam: Asan Point,
Civilian Camp, 1 m, 23 Aug 1949, Anderson 4 (MY); 0.5 mi N of Asan,
27 Jan 1945-46, Moore 252 (UC); Guam Experimental Station, 191-,
Thompson 252 (NY). Rota: 50 ft, 27 Jul 1946, Grether 4447 (K,UC).
Saipan: S of Puntan Magpi, NW coast, 30 m, 5 Feb 1950, Fosberg 31329
(NY). Tinian: Spring 1945, Knight 112 (FLAS); 2nd terrace E of Mt.
Lasso, East Side Island, 100 m, 13 Jun 1946, Fosberg 24923 (NY).
CAROLINE ISLANDS: Palau Islands: old seaplane base on W side of
Ngarakabesang (Arakabeson) Island, 2-5 m, 23 Mar 1950, Fosberg 32283
(NY). GILBERT ISLANDS: Tarawa Island: Betio, 4 Oct 1967, Adaire 44
(K).
POLYNESIA. HAWAIIAN ISLANDS: Oahu: cult. Honolulu, May
1943, Meebold s.n. (GH,M); l.c., 1 Feb 1931, Inafuku s.n. (NY-mixed);
l.c., University of Hawaii, 11 Dec 1929, Degener 17958 (NY); college

571
grounds, Rock s.n. (A-2 sh.). SAMOA: Reinecke 148 (*G); Upolu, 1905,
Rechinger 1256 (NY,W) and 1578 (W). TONGA: Nomuka: near the sea on
road to village, 30 Apr 1953, Yuncker 15810 (BM,G,GH,U,W). Haapai:
Uiha, sea level, 30 May 59, Soakar 541 (K). SOCIETY ISLANDS:
Tahaiti: 27 May 1922, Setchell & Parks 168 (GH.UC.W); Sep 1852,
Andersson s.n. (S-2 sh.); Papeete ach Fatuahua, 1852, Andersson s.n.
(S).
31ab. Cl itoria ternatea L. var. ternatea f. pauciflora Fantz,
f. nov.
Inflorescence racemose, 3-6 flowered. Peduncles solitary,
axillary, 1-4 cm long, base rigid with upper portion lax, slender.
Flowers blue.
TYPE COLLECTION: SENEGAL, ca 8 km N of Dakar near Ngor, 28 Aug
1962, Broadbent 14 (HOLOTYPE: K-159.).
NOTES: This form is conspicuous and easily recognized by the
more primitive type of inflorescence, a racemose axis with several
flowers.
DISTRIBUTION (Figure 67): Known only from the type locality.
3Tac. Clitoria ternatea L. var. ternatea f. fasciculata Fantz,
f. nov.
Peduncles 1-4 per axil, fascicled, typically (0.5) 1-4 cm long,
base rigid with upper portion lax, slender. Flowers blue, solitary at
apex of peduncle, rarely biflowered.

Figure 70. Cl i tori a terna tea - II. Var. ternatea f. paucif
"(a) node with leaf and inflorescence, x 1. Var.
f. fasciculata: (b-c) two nodes with fascicled
inflorescences, x 1. (Broadbent 14, K-159: a.
17313, MO 1717338: b-c7T
ora:
ternatea
Yuncker

573

574
TYPE COLLECTION: JAMAICA. St. Andrew Parish. Thicket along
Mona Road, 550 ft, 8 Nov 1957, Yuncker 17313 (HOLOTYPE: MO 1717333.
ISOTYPES: F. 1478717, G-335,MICH,S).
Yuncker 17313 had better material and was not mounted with other
collections on one herbarium sheet as are the other collections cited.
Nodes with only one inflorescence frequently have immature branchlets
present, as do the Field Museum and Stockholm specimens which are
poorer representatives of the fasciculate condition.
DISTRIBUTION: Known only from a few widely scattered localities,
either as a cultivar or found naturalized (?) in localities where the
species had been introduced.
INDIA, pr. Madura [?], Griffith 208 (CGE). WEST BENGAL:
cult. Calcutta, Wall ich 5344H (K; not BM,CGE,E,G).
FIJI ISLANDS: Lantoka, 10 Oct 1920, Greenwood 145
(K).
31 ad. Clitori a ternatea L. var. ternatea f. a 1 biflora (Voigt)
Fantz, comb, nov.
Ternatea flore simplici albido Tourn., Mem. Math. Phys.
Acad. Roy. Sci. Amsterdam 105. 1 706; nom. ill eg.
Flos clitori us, flore albo Burm., Zey. 100. 1737;
nom. i11 eg.
Clitoria foliis pinnatis L. var. a, Hort. Cliff. 360.
1 737 ; nom. i11 eg.
Clitori a fo 1 i is pinnata L. 6 Flos cl itorius, flore alba
(Burm.) L., FI. Zeyl. 130. 1747; nom. illeg.

575
Ternatea flore simp!ici albido Tourn. ex Mill., Card.
Diet. ed. 4. 3: TERNATEA. 1 754 ; nom. ill eg.
Clitoria ternatea L. ¡3 eadam foliiolus obtusioribus, flore
albido Lain., Ency. Bot. 2: 50. 1786; nom. i 1 leg.
Clitoria bracteata Poir., in Lam. Ency. Bot. Supp. 2:
301. 1811.
Clitoria ternatea L. 8 bracteata (Poir.) DC., Prod. 2:
234. 1825.
Nauchea ternatea Desc. var., Mem. Soc. Linn. Papers 4: 8.
1826; nomen null us.
Nauchea bracteata Dupuis ex Desc., Mem. Soc. Linn. Papers
4: 11. 1826.
Clitoria ternatea L. B alba Sweet, Hort. Brit. 2nd ed. 140.
1830; nomen nudum.
Clitoria pilosula Wall., Cat. Herb. Ind. 186 no. 5347.
1831-1832; nomen nudum.
Clitoria ternatea L. a ternatea fio, alba Boj., Hort.
Maurit. 91. 1837; nom. ill eg.
Clitoria ternatea L. D. fl. albo Hasskarl, Cat. Hort.
Bog. Alt. 275. 1844; nom. illeg.
Clitoria ternatea L. B al biflora Voigt, Hort. Calcuttensis
213. 1845.
Cl itoria pilosula Wall. ex Benth., Journ. Linn. Soc. 2:
37. 1858.
Clitoria ternatea L. var. pilosula (Wall, ex Benth.)
Bak. in Hook. Fl. Brit. Ind. 208. 1879.

576
Cl i tori a al biflora Mattel', Boll. Bot. Palermo 6: 97.
1908.
Clitoria ternatea L. var. alba Berhaut, FI. Senegal 47.
1954; nom. 111 eg.
Cl i tori a ternatea L. var. flore al bo Aiudie, nom. iji
sched.
Cl 1 tori a timoriana, nom. in sched.
Flowers white with vexillum greenish or greenish-white medially,
solitary or rarely biflowered, borne at the apex of rigid, short (0.5-
1.5 cm), solitary peduncles.
TYPE COLLECTION: INDIA. Bengal, Capt. Jenkins s.n. (NEOTYPE:
Hb. John Ball, E-57).
White-flowered individuals have been reported in the literature
since 1706, often noted as a variety. Names applied to this white-
flowered group are generally illegitimate in accordance with the
International Code of Botanical Nomenclature (1972). Those names
published prior to 1 753 were invalidly published, lienee illegitimate.
Most names published after 1753 were either polynominal names, or the
published name lacked a description, hence illegitimate. Descourtilz
(1826) noted a variety and described it, but he did not provide a name
for it:
The only distinction between this group and that of the typical
ternatea is the flower color, i.e. white flowers instead of blue
flowers. This is a minor variation, thus better treated as a form
instead of a variety. No forma names have been published for this
group. Two epithets previously published are available for use.

577
The epithet bracteata was published at the species level by Poiret
(1811) and reduced to varietal status by de Candolle (1825). The
epithet albiflora was published by Voigt at the varietal level.
Voigt's name was chosen for the form name to reflect the mor