LIBRARY OF THE

UNIVERSITY OF ILLINOIS

AT URBANA-CHAMPAIGN

560.5

.22)

BIOLOGY

B

• 29

TROPICAL AMERICAN PLANTS, II

LOUIS 0. WILLIAMS

FIELDIANA: BOTANY
VOLUME 29, NUMBER 6

Published by

CHICAGO NATURAL HISTORY MUSEUM
SEPTEMBER 14, 1961

TROPICAL AMERICAN PLANTS, II

LOUIS 0. WILLIAMS

Curator, Central American Botany

FIELDIANA: BOTANY

VOLUME 29, NUMBER 6

Published by

CHICAGO NATURAL HISTORY MUSEUM
SEPTEMBER 14, 1961

Library of Congress Catalog Card Number: 61-1 77SO

PRINTED IN THE UNITED STATES OF AMERICA
BY CHICAGO NATURAL HISTORY MUSEUM PRESS

13

Tropical American Plants, II

This is the second paper in a projected series which was planned
to contain information and descriptions of tropical American plants,
especially those of Central America. The first number of the series
was published in Ceiba 4: 270-275. 1955.

This paper contains notes and descriptions resulting from the
study of the Central American material used in the revisions for the
Flora of Guatemala, which is again under way. There are, in addi-
tion, notes on or descriptions of specimens which have been received
for determination. Most of the orchid material has been sent by
Mr. A. H. Heller, resident in Nicaragua, and Mr. Henry P. Butcher
of Panama, both discriminating collectors of the Orchidaceae.

Carludovica costaricensis (Harling) L. Wms., comb. nov. As-
plundia costaricensis Harling, Acta Horti Bergiani 18: 186. 1958.

This Central American economic plant has usually gone under the
name of Carludovica utilis (Oerst.) Benth. & Hook. Harling points
out that this is based on a mixed type and requires a new name. —
Leaves of this species have been and perhaps still are used in weav-
ing hats in Costa Rica. The long, flexible and strong stems are used
in making furniture and in tying large bundles. Common names of
the plant are cuajiote, palma, chidra, cola de gallo, coligallo and tucuso.

In his monograph Harling has distributed the Central American
species of Carludovica among six not very convincing segregate genera.

Cryptophoranthus Endresianus Kranzl. Fedde Repert. 17:
437. 1921.

Nicaragua: Tuma Grade, province of Matagalpa, alt. 1000 m.,
A. H. Heller 2672, flowered in cultivation 1960.

Cryptophoranthus is a small genus with but five species reported
from Costa Rica and Panama, and all of these rare. This is the first
collection of this highland genus in the mountains north of the low
isthmus of Nicaragua, which has been a barrier to the migration of the
montane flora of southern Central America. Mr. Heller's Nicaraguan
collections are adding many plants to the known flora of the country.

345

346 FIELDIANA: BOTANY, VOLUME 29

Epidendrum lamella turn Sesse" & Mocifio, Fl. Mex. ed. 2: 206.
1894, non Westc.

Specimens of this species collected in Mexico by Sesse", Mocifio
and Maldonado, numbers 1+321 (type in Madrid) and 4327, in the
herbarium at Madrid and in Chicago, indicate that this species is a
synonym ofLeochilus carinatus (Knowles & Westc.) Lindl., as Ames,
Hubbard & Schweinfurth have suggested that it might be (The Genus
Epidendrum 207. 1936).

Lockhartia Integra A. & S. Sched. Orch. 10: 108. 1930.

Nicaragua: Pendent on mossy branches, in deep shade, 3 miles
east of La Libertad, Chontales district, alt. 2200 feet, January 1960,
Heller 5821*.

A rare species and the determination here made from the charac-
ters. New to Nicaragua and another of the species collected for the
first time north of the ecological break at the San Juan River.

Maxillaria fulgens (Reichb. f.) L. Wms. Ann. Mo. Bot. Card.
28: 425. 1941.

Nicaragua: From J. Molina finca, Cerro Diablo, department of
Jinotega, alt. 4000 feet, September, Heller 1316; mossy rocks in pines
below Milwaukee finca, department of Matagalpa, alt. 3500 feet,
October, Heller 2361.

Another of Mr. Heller's finds which extends the range of the
species from Panama and Costa Rica northward beyond the bio-
logical barrier at the San Juan River.

Ornithocephalus Lankesteri Ames, Sched. Orch. 3: 24. 1923.

Nicaragua: Near tops of secondary growth trees, near base of
Pina Blance, department of Jinotega, alt. 3000 feet, October 1960,
Heller 1664.

New to Nicaragua and another mountain species which has
crossed the biological barrier at the San Juan River.

Pleurothallis Butcher! L. Wms., sp. nov.

Herba caespitosa epiphytica usque ad 15 cm. alta. Folia lineari-lanceolata,
acuta vel acuminata, purpurea, carnosa. Sepalum dorsale lineari-lanceolatum, uni-
nervium. Sepala lateralia connata, cymbiformia, ovalia, trinervia. Petala lan-
ceolata obliqua, acuta vel acuminata, uninervia. Labellum oblongo-ovatum vel
oblongo-subcordatum, acutum, trinervium. Columna ca. 5 mm. longa, arcuata.

WILLIAMS: TROPICAL AMERICAN PLANTS, II 347

Small caespitose epiphytic herbs to about 15 cm. tall. Secondary stems 2-
4.5 cm. long, covered with 2-3 tight chartaceous sheaths which disintegrate with
age. Leaves 10-11 cm. long and 1-1.2 cm. broad, linear-lanceolate, acute or acu-
minate, sessile on the secondary stems, somewhat fleshy and deep purple when
fresh, or the upper surface greenish. Inflorescence a few-flowered (±6) raceme,
about as long as or longer than the subtending leaf, bracts infundibuliform, char-
taceous, acute or acuminate, 3-4 mm. long, pedicels about 1.5 mm. long. Flowers
purple, the lip motile; sepals hairy on the inner surface; dorsal sepal linear-lanceo-
late, acute, 1-nerved, about 13.5 mm. long and 2 mm. broad, pubescent inside
above with coarse hairs, keeled dorsally along the mid-nerve; lateral sepals con-
nate to near the apex, cymbiform when fresh, oval, about 13 mm. long and 9 mm.
broad when spread, the free tips acute, each sepal 3-nerved, pubescent on the inside
along the outer nerve with coarse hairs and also along the sinus between the tips
of the sepals, glabrous outside, somewhat keeled along the mid-nerve of each sepal;
petals obliquely lanceolate, acute or short acuminate, 1-nerved, about 5 mm. long
and 1 mm. broad; lip when spread oblong-ovate or oblong subcordate, acute,
3-nerved, the lateral nerves short, basal angles of the lip rounded, thin, obscurely
lacerate, disc of the lip provided with 2 subparallel fleshy calli near the margins
and at and below the middle of the lip, disc obscurely muricate, tip of the lip ciliate;
column about 5 mm. long, arcuate, the clinandrium thin and the margin lacerate.

Panama: Epiphyte in mossy cloud forest, along a branch of
Rio Machomonte, Cordillera, vicinity of Volcan, slopes of Chiriqui
volcano, Chiriqui province, alt. 4000 feet, January 19, 1961, Henry P.
Butcher 651 (type in Chicago Natural History Museum) .

Among the nearly 200 species that are recognized from Central
America (more than 50 in Panama) I do not know any that are
closely allied to the present one. There are several that resemble
it superficially; the widespread and variable P. segoviensis Reichb. f.
is one of these. The coarse pubescence of the inner surfaces of the
sepals, the lip with obscurely muricate disc, the deeply cymbiform
and connate lateral sepals are characters by which the species may
be known.

Pleurothallis Liebmanniana Kranzl. Saertryk. af Vidensk.
Medd. Foren. 71: 170. 1920.

Nicaragua: Cerro Diablo, J. Molina finca, alt. 4000 feet, 1960,
A. H. Heller 2266.

New to Central America; previously known from the Mexican
states of Oaxaca and Chiapas.

Trichopilia turialvae Reichb. f. Hamb. Gartenz. 19: 11. 1863.

Nicaragua: Santa Maria Ostumas, department of Matagalpa,
alt. 4500 feet, rare, Heller 1104.

348 FIELDIANA: BOTANY, VOLUME 29

Previously known from Panama and Costa Rica; now recorded
from the mountains of Nicaragua.

Ribes leptostachyum Benth. PI. Hartw. 186. 1839.

Costa Rica: Paramo, Sierra de la Muerte, province of (?)Limon,
alt. 11,000 feet, October 10, 1943,Bar6owr 1025; Cerro de la Muerte, alt.
11,000 feet, July 6, 1932, Stork 3129; flowers reddish, shrub to 3 m.,
in paramo in moist places, summit of Cerro de la Muerte above the
cloud forest area, Cordillera de Talamanca, province of Cartago, alt.
3300 m., April 10, 1949, Williams 16296.

There is little question that the Ribes of the Costa Rican paramos
is closely related to several species found in the paramos of the Andes
rather than to Ribes ciliatum Humb. & Bonpl. of Mexico and Guate-
mala. Janczewski (Mem. Soc. Phy. Hist. Nat. Geneve 35: 199-517.
1907) has divided the Andean species rather too much, and had he
had available material from Costa Rica would probably have sep-
arated it as another species.

Tetr orchid him Molinae L. Wms., sp. nov.

Frutex vel arbor dioecia usque ad 10 m. alta. Folia elliptica vel late oblan-
ceolata, acuminata; inflorescentia spicata, mascula vel feminea; capsula bicellula,
subdidyma.

Shrubs or usually trees 2-10 m. or more tall, glabrous. Branches slender, with
several terminal leaves and prominent leaf -scars below, finally becoming smooth;
leaves short-petiolate, the petiole 4-10 mm. long, with a pair of small glands at the
junction of the petiole and the blade; leaf blade elliptic to broadly oblanceolate,
acuminate, often abruptly so with the apex somewhat obtuse or subacute, 6-13 cm.
long and 2-4.5 cm. broad, the margin entire or slightly repand, nerves 4-7 sub-
opposite pairs. Male inflorescence one or more interrupted spikes from the apex
of the branches, 4-6 cm. long, shorter than the subtending leaves; flowers sessile,
usually 2-3 in glomerules scattered along the spikes; sepals 3, ovate-triangular,
acute, about 2 mm. long; stamens 3, sessile, each anther 4-celled. Female inflo-
rescence spicate, short, few-flowered, 2-3 cm. long; capsule 2-celled, subdidymous,
dehiscent parallel to the long axis, about 7-8 mm. high and about 10 mm. broad
parallel to the line of dehiscence, 6-7 mm. broad on other axis; style flat, thick,
obscurely bifid. Seeds lenticular, somewhat rugose, red, about 7 mm. broad and
4 mm. thick.

Honduras: Shrub 2 m. tall in cloud forest above San Juancito in
San Juancito mountains, Dept. Morazan, alt. 2000 m., November 6,
1947, Williams & Molina 13341; flowers pale yellow, tree 4-6 m. in
cloud forest area, in mountains above San Juancito, Dept. Morazan,
alt. 2000 m., March 25, 1948, Williams & Molina 13980; arbol 6 m.,
comun el la floresta de nubes de la Montana de la Tigra, suroeste de
San Juancito, Depto. Morazan, alt. 2000 m., febrero 5, 1950, Wil-

WILLIAMS: TROPICAL AMERICAN PLANTS, II 349

Hams & Molina 17086 (type in Chicago Natural History Museum;
EAP, US) ; same data, Williams & Molina 17093; tree 10 m., in cloud
forest near San Juancito, Dept. Morazan, Williams 17^13. — These
collections all bear male inflorescences except number 17086.

Only three species of Tetrorchidium were known in Central Amer-
ica, and the present species is not closely allied to any of these. It is
easily distinguished by the much smaller leaves. Tetrorchidium rubi-
venium P. & E. of South America appears to be the most closely
allied species.

This species, like many others which we discovered, seems to be
endemic in the cloud forest of only one of the isolated peaks which
occur on the upper peneplain of Honduras. This species was abun-
dant in the cloud forest in the area of Montana de la Tigra at about
2000 meters but was not collected from any of the neighboring moun-
tains. No other species of Tetrorchidium has been collected from
these high mountains.

It is with pleasure that I name this species for my companion on
innumerable collecting trips into all the Central American republics,
Antonio Molina R.

DAVILLA Vand.

Plants of the genus Davilla are not uncommon in Central America;
however, the genus has not been critically studied here or in its range
into South America or the West Indies. Monographic treatment
is needed.

In revising the manuscript of the genus for the Flora of Guatemala
it became necessary to look into the many species to which those in
Central America might belong. The two species known to occur
in Central America are reviewed here.

Davilla aspera (Aubl.) Benoist var. aspera. Tigarea aspera
Aubl. PI. Guian. Fr. 2: 918, t. 350, 1775. Davilla aspera Benoist,
Bull. Soc. Bot. France 60: 392. 1913.

This is the oldest name for plants of the Central American group.
The two species of the genus found here are closely allied to D. aspera,
which was described by Aublet from material originating in French
Guiana. The West Indian and Central American representatives
are hardly more than varieties of this. Furthermore, there are sev-
eral species extending southward into Brazil that look to be very
closely allied to D. aspera. I have not seen specimens from Central

350 FIELDIANA: BOTANY, VOLUME 29

America which seem to belong to the typical variety. The com-
monest Davilla of Central America is the following variety:

Davilla aspera var. Matudae (Lundell) L. Wms., comb. nov.
Davilla Matudae Lundell, Phytologia 1: 371. 1940.

The representative of the species in Mexico and Central America
and perhaps elsewhere differs from the typical variety in minor de-
tails. The two large calyx lobes are glabrous or nearly so. The
indument of the upper surfaces of the leaves is sparse and often re-
duced to a few rough hairs with pustulate bases; that of the lower
surface may be on all of the veins or confined to the main nerves.
The narrow wings of the petioles are reflexed.

The West Indian Davilla Sagraeana A. Rich, has often been cred-
ited to Central America. The material so determined belongs mostly
to the present variety. Leon and Alain (Fl. Cuba 3: 292. 1953) place
D. Sagraeana as a synonym of D. rugosa Poir., which may well be a
synonym of D. aspera.

Lanjouw and Heerdt (in Pulle, Fl. Suriname 3: 294. 1941) include
Costa Rica in the range of Davilla multiflora (DC.) St. Hil. The
specimen that they saw is probably var. Matudae.

Tetracera volubilis L. and T. sessiliflora Tr. & PL have been con-
fused by many botanists with Davilla aspera var. Matudae.

Davilla Kunthii St. Hil. PL Usuel. Bras. 6. t. 22, 1825. D. lu-
cida Presl, Rel. Haenk. 2: 73. 1835. D. ovata Presl, I.e.

Closely related to Davilla aspera and its var. Matudae but dis-
tinguished by the more abundant and often fulvous pubescence of
the leaves, inflorescences and calyx lobes, and by the canaliculate
petioles with narrow upright or more commonly inrolled wings. The
species is found from Mexico to Panama and is relatively homogene-
ous. Almost certainly it will be found in South America. Tetracera
mollis Standl. is very similar superficially to this species and occa-
sionally may be confused with it.

Doliocarpus hispidus Standl. & L. Wms., sp. nov.

Scandens, ramulis usque ad 8 mm. diametro. Folia late elliptica vel elliptico-
obovata, obscure denticulata, hispida, basi cuneato; petioli hispides, canaliculates;
inflorescentia unipauciflora; fructus subglobosus ca. 1 cm. diametro, dense pubescens.

Scandent vines, the stem up to 8 mm. or more in diameter, thickened at the
nodes. Fertile branches short, to about 15 mm. long, bearing the leaves and fas-
ciculate inflorescences, somewhat thickened; leaves broadly elliptic to elliptic-

WILLIAMS: TROPICAL AMERICAN PLANTS, II 351

obovate, obtuse or apiculate-acuminate at the apex, obscurely denticulate, base
cuneate, with 15-20 nearly parallel nerves on each side of the mid-nerve, impressed
above, prominent below, spreading hispid-pubescent, especially on the nerves be-
low, sparsely so above or almost glabrous except the canal of the mid-nerve, densely
pubescent toward the base, 10-21 cm. long and 4-10 cm. broad; petiole short,
1.5 cm. long, canaliculate, fleshy, hispid; inflorescences fasciculate, few, one- or
possibly few-flowered; peduncles about 1 cm. long, hispid-pubescent, with 2 coria-
ceous bracts toward the apex, nearly joined and rotate, subtending the fruit; the
fruit with a short (1.5 mm.), thick (2.5 mm.) pedicel; calyx lobes 4, in 2 series, in
fruit broadly ovate to suborbicular, densely short hispid-pubescent outside, gla-
brous within; stamens many; petals not seen; fruit subglobose, about 1 cm. in
diameter, densely erect-spreading pubescent.

Costa Rica: Woody liana, fruits dull red, Esquinas forest, region
between Rio Esquinas and Palmar Sur de Osa, province of Punta-
renas, alt. 75 m., January 31, 1951, Allen 5822 (type in Chicago
Natural History Museum; also in Herb. Esc. Agr. Panam., US).

Not closely related to any of the species of Central America and
easily distinguished from all of those which we have seen by the short
fertile branches which apparently do not continue to grow. The
densely pubescent fruit is also distinctive. — Our specimen is not good.
It had but one attached fruit and no flowers.

Tetracera belizensis Lundell, Contr. Univ. Mich. Herb. 6: 44.
1941.

The type, from British Honduras, Gentle 2794, was originally dis-
tributed as Davilla aspera (Aubl.) Ben. Steyermark's collection 1+1787
from Rio Tameja, department of Izabal, Guatemala, also distributed
as Davilla aspera, is this species and the first specimens collected in
Guatemala.

Ouratea crassinervia Engler in Mart. Fl. Bras. 12, (2): 337.
1878. 0. costaricensis Standl. Field Mus. Bot. 18: 604. 1937.

Costa Rica: Vicinity of Gudpiles, prov. of Limon, March 12, 13,
1924, Standley 87083.

Panama: Cut-over secondary forest, Santa Rosa Hills, January 4,
1958, Smith & Smith 3431; Isla de Colon, Bocas del Toro, August
1938, Woodson, Allen & Siebert 1948.

Brazil: Itapacora, Pohl?, Sellow?

In his account of Ouratea Dwyer credited the Brazilian 0. curvata
(St. Hil.) Engler to Central America (Lloydia 7: 128. 1944) stating
that "Examination of the type material of 0. crassinervia leaves no
doubt that it is conspecific with 0. curvata."

352 FIELDIANA: BOTANY, VOLUME 29

We have a photograph of the type of Gomphia curvata St. Hil.
[Ouratea curvata (St. Hil.) Engler ex Dwyer] as well as a photograph
and a duplicate specimen of the type collection of 0. crassinervia
Engler, cited above. Dwyer has considered the type of 0. crassi-
nervia, the specimen collected at Itapacora, to be the cotype collec-
tion for 0. curvata St. Hil. Saint Hilaire, however, quite definitely
cites one of his own collections for his species (Saint Hilaire 338),
which is not closely related to 0. crassinervia Engler.

Ouratea lucens (HBK.) Engler in Mart. Fl. Bras. 12, (2) : 350.
1876. Gomphia lucens HBK. Nov. Gen. & Sp. 7: 219. 1825. G. Jur-
gensenii Planch, in Hook. Journ. Bot. 6: 11. 1847. 0. Jurgensenii
Engler, I.e. 351. Stenouratea Wrightii v. Tiegh. Ann. Sci. Nat. se>. 8,
16: 219. 1902. 0. isthmica Riley, Kew Bull. 1924: 108. 1924. 0. steno-
botrys Riley, I.e. 109. 0. Wrightii Riley, I.e. 110. 0. ligans Dwyer,
Lloydia 7: 126. 1944. 0. prominens Dwyer, I.e. 143.

Ouratea lucens is the common and widespread species of the genus
in Central America. It is distributed from Mexico to Panama and
adjacent Colombia, most commonly along the Atlantic side of the
area. There is considerable variation in the material now available
for study but I find no characters by which the five species main-
tained by Dwyer may be kept apart (Lloydia 7: 121-145. 1944).

Ouratea lucens var. podogyna (Donn.-Sm.) L. Wms., comb,
nov. 0. podogyna Donn.-Sm. Bot. Gaz. 18: 2. 1893. 0. insulae
Riley, Kew Bull. 1924: 106. 1924. 0. pyramidalis Riley, I.e. 107.

This variety is very close to the species, but based on the material
presently available it would seem to be distinguishable. The leaves
of the variety are usually larger and more coriaceous, usually broad-
est above the middle, and often more prominently serrate. The in-
florescence is frequently, if not always, branched, and the panicle is
large and spreading in some specimens.

The variety is known from Mexico, Guatemala, British Honduras
and Honduras. The type, von Turckheim 1034, lent by the U. S.
National Herbarium, has a rather small but branched inflorescence,
much smaller than is often the case in the variety. The following
specimens are in the Chicago Natural History Museum herbarium:

Mexico: Pinuela, Escuintla, Chiapas, 700 m., February 19, 1945,
Matuda 5009, 5010; Acacoyagua, Chiapas, February 29, 1948, Ma-

WILLIAMS: TROPICAL AMERICAN PLANTS, II 353

tuda 17510; Corsego, Pueblo Nuevo Com., Chiapas, April 15, 1948,
Matuda 17660.

Guatemala: Montana del Mico, Dept. Izabal, alt. 65-600 m.,
April 1, 1949, Steyermark 38512; shrub 10 feet in jungle, alt. 2000 feet,
Chama, Dept. Alta Verapaz, July 26, 1920, Johnson 399.

British Honduras: Valentin, El Cayo Dist., June-July 1936,
Lundell 6296; on limestone, without definite locality, Conservator of
Forests 113; water hole near Vaca, El Cayo District, April 6, 1938,
Gentle 2^59; tree 70 feet, Temash, April 22, 1934, Schipp 652; Camp
33, British Honduras-Guatemala Survey, alt. 2900 feet, April 29,
1934, Schipp 1239; Tikal, April 12-15, 1931, Bartlett 12631.

Honduras: Tree 75 feet, Wispernini Camp, March 1938, von
Hagen & von Hagen 13^8, 1397.

Cleyera theaeoides (Sw.) Choisy, Me"m. Soc. Phys. Hist. Nat.
Geneve 14, 1: 110. 1855. Eroteum theaeoides Sw. Prodr. 85. 1788.
Cleyera serrulata Choisy, I.e. Eurya panamensis Standl. Ann. Mo.
Bot. Gard. 25 : 829. 1938. Cleyera Matudai Kobuski, Journ. Am. Arb.
22: 403. 1941. C. revoluta Kobuski, I.e. 405. C. tacanensis Kobuski,
I.e. 406. C. Skutchii Kobuski, I.e. 408. C. costaricensis Kobuski,
I.e. 410. C. panamensis Kobuski, I.e. 412.

In the abundant material now available from Mexico, Central
America and Jamaica I find no convincing characters that distin-
guish one from another the several species of Cleyera from Central
America maintained or proposed by Dr. Kobuski in his monograph
of the American species. Characters distinguishing two or more of
these species, and characters used in Dr. Kobuski's key, may occa-
sionally be found on the same sheet: there may be variation in pubes-
cence of twigs, pedicels, sepals and petioles; in leaf size, shape, mar-
gins (may be plain or revolute), and surface; leaf may be coriaceous or
membranaceous (none are if mature); in length of pedicels; and in
inflorescences, which may be single flowers or fascicles.

Every specimen of Cleyera from Jamaica which I have seen can
be exactly duplicated by a specimen from Central America, except
one which has petals a little larger than any which we have from
Central America. Furthermore, it is not uncommon to find a species
from the mountains of Jamaica turning up in the mountains of Cen-
tral America.

Thanks are due to the Science Museum, The Institute of Jamaica,
Kingston, for the loan of their material of Cleyera theaeoides which
was used in this comparative study.

354 FIELDIANA: BOTANY, VOLUME 29

Freziera Grisebachii Krug & Urban, Bot. Jahrb. 21 : 542. 1896.
Eurya lancifolia Standl. Field Mus. Bot. 8: 317. 1931. Freziera lan-
cifolia Kobuski, Ann. Mo. Bot. Gard. 25: 354. 1938.

Known from Jamaica, British Honduras, Honduras and Mexico (?).

A specimen collected by Paul Allen, Allen 6408, near Agua Azul,
Department of Cortes, Honduras, at 600 m. elevation, February 9,
1952, is an almost exact duplicate of Harris 6067, typical material
from Jamaica of F. Grisebachii. The type of F. lancifolia, Schipp 455,
and another collection, Schipp 710, both from British Honduras,
differ from the first two specimens mentioned only in having the
upper side of the mid-rib pubescent.

A specimen collected in Mexico by Liebmann (no. 307) and anno-
tated by Kobuski as F. macrophylla Tul. probably also belongs here.

Freziera macrophylla Tulasne, Ann. Sci. Nat. Bot. ser. 3, 8:
330. 1842. Eurya macrophylla Szy. in Engler & Prantl, Nat. Pflan-
zenf. III. 6: 190. 1893.

Guatemala: Barranco Emenencia, above San Rafael Pie de la
Cuesta, Dept. San Marcos, alt. 2100-2400 m., March 14-15, 1939,
Standley 68690; El Pocito, south of San Martin Chile Verde, Dept.
Quezaltenango, alt. about 2200 m., January 27, 1941, Standley 85074;
between Todos Santos and Finca El Porvenir, Dept. San Marcos,
alt. 1300-3000 m., March 1, 1940, Steyermark 37049; above Finca
El Porvenir, alt. 1400-1700 m., March 8, 1940, Steyermark 37272.

Standley (in manuscript) has written: "The specimens from Gua-
temala, in part referred by Kobuski to F. guatemalensis, seem to agree
perfectly with authentic material of F. macrophylla, and to be cer-
tainly different from F. guatemalensis"

This species is closely related to F. Grisebachii, mentioned above,
and to F. candicans Tulasne, the latter known from Costa Rica,
Panama and South America.

Laplacea coriacea L. Wms., sp. nov.

Arbor (?). Folia coriacea, glabra, late ovata vel late obovata, obtusa, integra
vel undulata; inflorescentia axillaris, uniflora; sepala imbricata, carnosa vel coria-
cea, ciliolata; stigma stelliformis, 5-lobata; capsula ovoidea, 5-loculata.

Trees or shrubs of unknown size; branchlets many, glabrous. Leaves toward
the apex of the branches and branchlets, stiff-coriaceous, completely glabrous,
symmetrical or slightly asymmetrical, broadly oval to broadly obovate, obtuse,
the margins entire or slightly undulate, the blade 4-8 cm. long and 2.5-5 cm.
broad, often decurrent into the petiole; nerves 7-12 pairs, nearly plane above,

WILLIAMS: TROPICAL AMERICAN PLANTS, II 355

slightly canaliculate beneath, the mid-rib plane above, elevated beneath; petiole
to 1.5 cm. long, subglandular, sometimes winged; inflorescences axillary, the flowers
solitary; peduncle terete, 1-2.5 cm. long; sepals 5, imbricated, about 3 mm. long
and twice as broad, fleshy and coriaceous, ciliolate, otherwise glabrous; petals
unknown; stamens unknown; ovary elliptic-ovoid, densely sericeous; stigma stel-
late, 5-lobed; capsule ovoid, about 1.5 cm. long and 0.8 cm. thick, 5-loculate.

Guatemala: Leaves stiff -coriaceous, rich shining green above, pale
green beneath, between Cerro de Monos and upper slopes of Monte
Virgen, Sierra de las Minas, Dept. Zacapa, alt. 2000-2600 m.,
January 17, 1942, Steyermark 42880 (type in Chicago Natural History
Museum) .

Somewhat allied to Laplacea grandis, from which it is easily dis-
tinguished by its broad, obtuse, coriaceous and completely glabrous
leaves. This species, along with a number of sterile specimens that
apparently belong in the Sabiaceae and in another family (which I
do not recognize) having opposite leaves, has been confused by both
Standley and Kobuski with Laplacea grandis, a species of relatively
low elevations. Laplacea grandis is now known from Mexico, Hon-
duras, Costa Rica and Panama and may be expected in British
Honduras and Guatemala. It is unlikely that a species of relatively
low elevations would be found also at rather high elevations. Data
on specimens of Laplacea grandis indicate that it is found at some
90 meters in Mexico and up to about 700 m. in Honduras. In Costa
Rica and Panama the range is from some 300 to 1050 meters. The
type of Laplacea coriacea was found above 2000 meters.

Hypericum caracasanum Willd. Sp. PI. 3: 1458. 1803.

Costa Rica: Above timberline near summit of Dos Burros peak,
Cerro de la Muerte, province of San Jose", alt. 11,000 feet, May 14,
1943, Dayton 3069; common, flowers yellow, strict shrub 1 m. tall,
shrub-paramo, Cerro de la Muerte, Pan-American Highway, 5 km.
above Millsville, Cordillera de Talamanca, province of Cartago, alt.
3400-3500 m., July 22, 1949, Holm & Iltis 454.

On the paramos of Cerro de la Muerte, doubtless on adjacent
paramos, and on the Volcan de Irazu, there are three known species
of Hypericum, at least two of which are found also in the paramos
of the Andes of South America. In addition to the above, new to
the flora of North America, H. irazuense Kuntze and H. strictum
HBK. are not uncommon there.

Hypericum irazuense Kuntze, Rev. Gen. 1: 61. 1891. Hyperi-
cum silenoides of authors, hardly of Jussieu.

356 FIELDIANA: BOTANY, VOLUME 29

A common species of the paramos of Costa Rica and also known
from Irazu volcano. This may be one of the species found also in
the paramos of the Andes but I am not able to prove that it is. Hy-
pericum stenopetalum Turcz. (1858) from Venezuela is very similar
if not identical. Jussieu's illustration of his H. struthiolaefolium
(1804) seems to be that of a plant very like ours, and there are per-
haps others.

Keller has determined Costa Rican material of this species as
H. silenoides Juss., and was followed in this by Standley. Jussieu,
however, described that species as herbaceous and the illustration
provided indicates that it was — not like this shrubby species of
Costa Rica.

TOVOMITA Aubl.

My review of the species of Tovomita Aubl. and of Tovomitopsis
Planch. & Triana seems to indicate that Tovomitopsis is based on
rather poor characters and might best be referred to Tovomita.

The plants of the genus are dioecious or polygamodioecious,
although among our Central American specimens there are relatively
few that are staminate only. Most of the pistillate plants have more
or less well-developed stamens. The pistillate or perfect flowers
are usually larger than the staminate ones.

The number of sepals in the genus is variable. There may be
two, or four of which the outer two are smaller, or sometimes three
with the lowest one smallest.

Tovomita nicaraguensis (Oerst.) L. Wms. Tropical Woods
111 : 18. 1959. Tovomitopsis nicaraguensis Oerst. ex Planch. & Triana,
Ann. Sci. Nat. se'r. 4, 14: 266. 1860. Tovomitopsis psychotriaefolia
Oerst. ex Planch. & Triana, I.e. 263. Tovomitopsis costaricana Oerst.
ex Planch. & Triana, I.e. 266. Chrysochlamys guatemaltecana Donn.-
Sm. Bot. Gaz. 13: 26. 1888. Chrysochlamys costaricana Hemsl. in
Godm. & Salv., Biol. Cent. Am. Bot. 1: 87. 1879. Chrysochlamys
psychotriaefolia Hemsl. I.e. Chrysochlamys nicaraguensis Hemsl. I.e.
Tovomitopsis guatemaltecana Standl. & Steyerm. Field Mus. Bot.
23: 66. 1944.

The species occurs from Guatemala to Panama and perhaps
farther south, where it is usually found at rather low elevations.

The original collections of the three species described by Oersted
were made by him near Turrialba in Costa Rica and along the
San Juan River in Nicaragua. The types are preserved in Copen-

WILLIAMS: TROPICAL AMERICAN PLANTS, II 357

hagen and we have photographs of them. The type of Tovomitopsis
costaricana is an exceedingly poor specimen, but those of T. psy-
chotriaefolia and T. nicaraguensis are quite good. These species have
been maintained by subsequent botanists who have studied the
group, although I find no way to separate them. The type of Tovo-
mitopsis costaricana of Oersted is essentially undeterminable; how-
ever, the leaf of the type seems to match almost exactly those on
the types of T. psychotriaefolia and T. nicaraguensis. The type speci-
mens of the first two were possibly collected at the same place and
time.

Tovomita glauca (Oerst.) L. Wms., comb. nov. Tovomitopsis
glauca Oerst. ex Planch. & Triana, Ann. Sci. Nat. se>. 4, 14: 264.
1860. Chrysochlamys glauca Hemsl. in Godm. & Salv. Biol. Cent.
Am. Bot. 1:87. 1879.

All of the specimens (about 25) which we have of this species are
from the highlands of Costa Rica. The species is allied to the pre-
ceding one.

Tovomita gracilis L. Wms., sp. nov.

Frutex usque ad 4-5 m. Folia ovato-lanceolata, acuta vel acuminata. Sepala 4,
exteriores minores, cochleata. Petala 5, cochleata, oblongo-ovata, obtusa, carnosa.
Capsula baccata, 5-locularis.

Small dioecious or polygamodioecious shrubs to 4-5 m. tall, sparsely branched,
the branches very slender, covered with appressed canescent hairs, becoming gla-
brous with age. Leaves ovate-lanceolate, acute or acuminate, obtuse or somewhat
acute to the base, with 5-7 pairs of nerves, the mid-nerve with appressed canescent
hairs or glabrous, thin, 8-11 cm. long and 3.5-5 cm. broad; petioles slender, short,
pubescent with appressed canescent hairs, becoming glabrous, 6-8 mm. long. Inflo-
rescence a diffuse paniculate cyme to about 15 cm. long, borne on leafless branches
or twigs. Flowers small, the buds about 2.5-3 mm. long and 1.5-2 mm. in diam-
eter; sepals 4, in 2 pairs, about 15-nerved; petals 5, oblong-oval, obtuse, cochleate,
fleshy, 2-3 mm. long and 1-2 mm. broad; staminate flowers not seen; stamens of
polygamodioecious flowers many, of various lengths, the longest about 2 mm.;
anthers apparently sterile, filaments rather broad; stigmas 5, sessile; fruits capsu-
lar-baccate, obovate-conic, 2.5-3 cm. long and 1.2-2 cm. in diameter, 5-loculate,
5-seeded, dehiscing with force, red.

Costa Rica: Arbuste 4-5 m., bois, collines de San Pedro de
San Ramon, alt. 1050 m., 3-V-1925, Brenes 4217 (type in Chicago
Natural History Museum).

Not closely allied to any Central American species of this or
allied genera. It is easily distinguished by leaf shape; by the very
slender branches; by the inflorescence borne on leafless twigs; and

358 FIELDIANA: BOTANY, VOLUME 29

by the appressed canescent pubescence on the younger twigs, pet-
ioles and bases of the mid-rib. The very young leaves, those emerging
from buds, are covered with white hairs.

Bixa Orellana L. var. leiocarpa (0. Kuntze) Standl. & L. Wms.,
comb. nov. Orellana americana var. leiocarpa 0. Kuntze, Rev. Gen.
45. 1891. Bixa Orellana f . leiocarpa Macbr. Field Mus. Bot. 13 (4) :
11. 1941.

Widely distributed with the species but easily distinguished by
the fruits, which are smooth and without bristles. — The specific
name honors Francisco Orellana, discoverer of the Amazon River
and associate of Pizarro in the conquest of Peru.

Hybanthus Galeottii (Turcz.) Morton, comb. nov. lonidium
Galeottii Turcz. Bull. Soc. Nat. Moscou 27 (2) : 339. 1854. lonidium
occultum Polak. Linnaea 41 : 548. 1877. Calceolaria nigricans Dowell,
Bull. Torr. Bot. Club 33: 554. 1906. Hybanthus Purpusii Standl.
Field Mus. Bot. 8: 141. 1930. Hybanthus occultus Standl. I.e. 18:
714. 1937.

In the account of Mexican Hybanthus, Morton (Contr. U. S.
Nat. Herb. 29: 77-81. 1944) placed lonidium Galeottii among the
dubious species. An examination of a photograph of an isotype
(CNHM Neg. no. 24006, of a specimen in the Delessert Herbarium,
Geneva, Galeottii 7085, from Jalapa, Veracruz, Mexico) shows that
the species is identical with H. Purpusii Standl. (type from Zacuapan,
Veracruz, Mexico), which in turn seems to be indistinguishable
from H. occultus (Polak.) Standl. (type from Cerro La Carpintera,
Costa Rica). The epithet Galeottii has priority.

VIOLA L.

There will be shown in the Flora of Guatemala seven species of
Viola, two introduced and cultivated and five native. Among the
native species Viola scandens Willd. is amply distinct. Viola Nannei
Polak., V. guatemalensis Becker, V. Schaffneri Becker, and V. Seleri-
ana Becker are all very closely related and I doubt that they really
are distinct, one from another, or from Viola Hookeriana HBK.,
which is presently restricted to Mexico.

Casearia Bartlettii Lundell, Lloydia 2: 104. 1939. C. elegans
Standl. Field Mus. Bot. 9: 311. 1940. Samyda Bartlettii Lundell,
Wrightia2:104. 1960.

WILLIAMS: TROPICAL AMERICAN PLANTS, II 359

Dr. Lundell has recently changed his C. Bartlettii to the genus
Samyda, commenting that "the large white flowers with the staminal
tube 6-7 mm. long are more typical of Samyda." — I find flower size
not too unusual in this large genus and in our specimens of this spe-
cies the staminal tube is very much shorter than that indicated. The
sepals do not form a tube as is the case in Samyda.

The species is known from Mexico, Guatemala and Honduras.

Casearia javitensis HBK. var. myriantha (Turcz.) L. Wms.,
comb. nov. C. myriantha Turcz. Bull. Soc. Nat. Mosc. 36: 609. 1863;
Johnston, Sargentia 8: 212. 1949. C. Brighamii S. Wats. Proc. Am.
Acad. 21: 459. 1886. C. laetioides Warb. in Engler & Prantl, Pflan-
zenf. 3, Abt. 6a: 52. 1893.

Widely distributed from Mexico through Central America (ex-
cept not reported in El Salvador and Nicaragua) to Panama.

This Central American variation of C. javitensis differs from all
the South American material which we have in possessing glabrous
or nearly glabrous capsules which are much heavier in texture and
somewhat larger than those of the typical variety. The geographical
range of the two varieties apparently does not overlap. In addition
to the pubescent or tomentose capsules of the typical variety, John-
ston also mentioned the ascending sepals which embraced the capsule.
The sepals, however, are as often reflexed as ascending in our mate-
rial of both varieties.

Casearia nitida Jacq. Enum. PL Carib. 21. 1760; Sel. Stirp. Am.
Hist. 132. 1763. Samydafruticosafoliis nitidis cordatis — P. Browne,
Civ. Nat. Hist. Jam. 217. 1756. Samyda nitida L. Syst. Nat. ed. 10:
1025. 1759. C. corymbosa HBK. Nov. Gen. & Sp. 5: 366. 1821.
?C. dubia DC. Prodr. 2: 51. 1825. ?C. banquitana Krause, Beih.
Bot. Centralbl. 32, Abt. 2: 345. 1914; Johnston, Sargentia 8: 211.
1949. C. laevis Standl. Contr. U. S. Nat. Herb. 23: 845. 1923.
C. dolichophylla Standl. I.e. 846. C. Gentlei Lundell, Field & Lab. 13:
9. 1945. C. banquitana var. laevis Johnston, Sargentia 8: 211. 1949.

A common and occasionally abundant species in Mexico, the
West Indies, all of Central America and extending into South Amer-
ica. It is fairly easily distinguished from the other species of Casearia
in Central America by the combination of short paniculate inflores-
cences, and the fruits which usually have three rather sharp vertical
ridges along which the fruit splits at maturity.

360 FIELDIANA: BOTANY, VOLUME 29

The synonymy above covers both the glabrous and pubescent
phases of the species, which I have not wished to separate by varietal
names. Johnston, who thought C. nitida to be restricted to Jamaica,
separated these two phases. The varietal name which he made be-
longs to the typical phase of the species as I understand it. It seems
to me that the names given above certainly belong to one not too
polymorphic species, considering the rather great geographical range
and the genus and family to which it belongs.

Jacquin based his Casearia nitida on Patrick Browne's reference.
Linnaeus based his Samyda nitida on the same reference a year earlier.
Jacquin apparently worked independently of Linnaeus' prior work
and Linnaeus should not be used as parenthetical authority in C.
nitida Jacq.

I have not seen authentic material or phototypes of C. banquitana
Krause, although I hardly think that there is a question that it be-
longs here. Casearia dubia DC. is based only on a drawing of which
we have a photographic copy. It is almost certainly the glabrous
form of this species.

Gasearia sericea Williams & Molina, sp. nov.

Arbores vel frutices usque ad 6 m. altae. Folia oblongo-obovata vel obovata,
breviter acuminata, obtusa, subtus piloso-sericea. Inflorescentia fasciculata. Se-
pala oblonga vel oblongo-obovata, obtusa. Stamina et staminodia 8. Stigma
capitata.

Small unarmed trees or shrubs 2-6 m. tall. Branches slender, with prominent,
white, elongate lenticels, densely pilose, the hairs mostly erect. Leaves oblong-
obovate to obovate, short-acuminate and obtuse, acute or cuneate to the often
unequal base, densely and softly pilose-sericeous below, especially along the nerves,
obscurely sinuate-dentate or essentially entire; blades 6-10 cm. long and 3-4.5 cm.
broad; petioles about 1 cm. long, pilose; principal veins and nerves prominulous on
both surfaces. Inflorescences fasciculate, borne in the leaf axils, few-many-flow-
ered. Flowers each borne in the axil of a small, reddish, scarious bract; pedicels
slender, pilose, about 3-6 mm. long; calyx lobes 5, oblong to narrowly oblong,
obtuse, somewhat unequal in size, two usually smaller, connate for a short dis-
tance at the bases, pilose-sericeous outside but less so inside, 4-5 mm. long, 2-
2.5 mm. broad; petals none; stamens about 8, the filaments adnate to the alter-
nating staminodes and with them forming a corona 2 mm. or less high, staminodes
as many as the stamens, about 3 mm. long, oblong, obtuse, the free portion densely
pilose; ovary (flowers in anthesis) pilose, style slender, stigma capitate, reaching
to the anthers; fruit unknown.

Honduras: Flores verdes, arbol 5 m., Rio Guarabuqui, terrenos
de los indios Xicaques de la Montana de la Flor, Departamento de
Morazan, alt. 1800 m., Antonio Molina R. 30^7; arbol 2-6 m., mar-
gines del Rio Jalan, 3 km. al norte de Guaimaca, Departamento de

WILLIAMS: TROPICAL AMERICAN PLANTS, II 361

Morazan, alt. 800 m., Junio 13, 1950, Antonio Molina R. 3085 (type
in Chicago Natural History Museum; dupl. Escuela Agricola Pan-
americana) .

Casearia sericea belongs to a group of species which have the
flowers borne in fascicles. It is not closely allied to any of the spe-
cies of Central America but of these is perhaps closest to C. aculeata.
It is distinguished from the species which we know by the combina-
tion of free or but slightly connate sepals with the corona free from
these; fairly large flowers; form and size of the corona; leaf shape,
margin, size, and placement of pubescence.

Casearia tacanensis Lundell, Lloydia 4: 54. 1941.

Mexico: Small tree 6-7 m. high, 15 cm. diameter, Volcan de Ta-
cana, Matuda 2441; tree 10 m. high in virgin forest, Santa Rosa near
Escuintla, 1600 m., Matuda 4245.

In a large and difficult genus this species is one that is easily dis-
tinguished— so distinct in fact that I thought that our fruiting speci-
men must belong to another genus. A comparison with the type,
which is in flower, kindly lent by the Herbarium, University of
Michigan, leaves no doubt that the two Matuda collections cited
are conspecific and that they belong in the genus Casearia. The
flowers stand apart from most in that the staminal tube is strongly
connate up to about the middle of the sepals. The sepals are sharply
reflexed at the point of attachment of the summit of the staminal
tube. The fruit is a 3-4-valvate capsule, globose, densely, coarsely
and irregularly obtuse- tuberculate and to 2 cm. in diameter or per-
haps more.

Hasseltia micrantha L. Wms., sp. nov.

Arbor parva, usque ad 5-6 m. alta. Folia elliptica vel elliptico-oblonga, breviter
acuminata, acuta vel obtusa, triplinervia, subtus sparce piloso-pubescentia. In-
florescentia cymosa. Sepala 4, oblongo-ovata, obtusa. Petala 4, late oblanceolata,
obtusa. Stamina 20-30.

Small trees 5-6 m. tall. Branchlets pubescent, soon becoming glabrous. Leaves
elliptic to elliptic-oblong, short-acuminate at the tip, acute or obtuse, triplinerved
from the base and with two small crateriform glands just above the junction with
the petiole, glabrous above, sparsely pilose-pubescent below, especially on the
nerves; blades 7-15 cm. long and 3-6 cm. broad; petioles slender, sparsely pubes-
cent, 1-2.5 cm. long. Inflorescence a multiflorous, terminal, compound cyme
shorter than the leaves, the peduncles increasingly pubescent upward, the pedi-
cels and calyces canescent. Calyx valvate, sepals 4, oblong-ovate, obtuse, 1-nerved,
1.5-1.7 mm. long and 0.8-1.2 mm. broad, canescent outside, glabrous or nearly so
within. Petals broadly oblanceolate, obtuse, 1.5-1.7 mm. long and about 1 mm.

362 FIELDIANA: BOTANY, VOLUME 29

broad above the middle, pubescent and ciliolate. Stamens 20-30, about as long as
the perianth, inserted on a glandular or interrupted staminodial disc below the
ovary, filaments hirtellous-pilose. Ovary pubescent but soon glabrous, style 1.2-
1.4 mm. long, stigmas 3; fruit unknown.

Nicaragua: Flores blancas, arbol 5-6 m., bosque lluvioso de mon-
tanas de Esquipulas y Aleman, drenaje de Rio Aleman, departamento
Zelaya, alt. 150 m., noviembre 27-29, 1951, Paul J. Shank & Antonio
Molina R. 4806 (type in Chicago Natural History Museum; dupli-
cate, Herb. Escuela Agricola Panamericana) .

Related to the widely distributed H. floribunda HBK., from
which it is easily distinguished by the very small flowers and the
sparsely pubescent under surface of the leaves.

HOMALIUM Jacq.

A study of the Central American material of Homalium now
available — and it is still not much — seems to indicate that there are
fewer species in the region than those indicated by Dr. Blake in his
revision of the group, which was prepared more than forty years ago.
Since the time of Blake's revision one additional species has been
described for the Central American flora and the range of still an-
other one has been extended to the area. I believe that the known
species of Homalium in Central America are only two. One of these
is widespread and must take an older name from the West Indies.

Stamens 5 or more in each fascicle; filaments pilose, at least toward the base.

1. H. trichostemon
Stamens 4 or fewer (usually 3) in each fascicle; filaments not pilose.

2. H. racemosum

1. Homalium trichostemon Blake, Contr. Gray Herb., n. ser.
53: 60. 1918; Contr. U. S. Nat. Herb. 20: 233. 1918.

The species is known from several collections made in Mexico, and
it has been collected four times in Guatemala, in the department of
Escuintla, by Standley. Blake comments that Homalium senarium
Moc. & Sesse" (DC. Prodr. 2: 54. 1825), which was insufficiently
known to him, might prove to be either this species or H. mollicellum
Blake, the only species known in Mexico. A photograph of the sketch
upon which the Mocino and Sesse" name is based is available. Six
stamens are indicated in each fascicle, which would apply only to
H. trichostemon among species known from our region. There is in
the Sesse" and Mocino herbarium a single collection of Homalium
(No. 2276; Chicago Nat. Hist. Mus. neg. 43,770). This collection,

WILLIAMS: TROPICAL AMERICAN PLANTS, II 363

which I have seen, has four or more stamens in each bundle and
seems to me to be indistinguishable from the West Indian H. pleian-
drum Blake. Whether or not H. pleiandrum and H. trichostemon are
the same may be decided when more material is available.

2. Homalium racemosum Jacq. Enum. PI. Carib. 24. 1760.
H. hondurense Donn.-Sm. Bot. Gaz. 20: 4. 1895. H. nicaraguense
Blake, Contr. U. S. Nat. Herb. 20: 225. 1918. H, stenosepalum Blake,
I.e. H. eurypetalum Blake, I.e. 234. H. riparium Standl. Carnegie
Inst. Wash. Publ. 461: 74. 1935, non Gilg, 1908. H. Schippii Standl.
Tropical Woods 52: 27. 1937.

Widely distributed, from Mexico through Central America, the
West Indies and northern South America to Brazil and Colombia,
perhaps to Peru.

Several names have been applied to Central American material
of this species, based primarily on the length of the pedicels, on the
size and occasionally on the shape of the sepals and petals. The
length of the pedicels is subject to some variation, mostly 1-8 mm.
in length. We have two sheets of the isotype of H. nicaraguense
Blake, which is described as having pedicels 0.5 mm. long or less;
on ours the pedicels are 1-3 mm. long. The type of H. Schippii
Standl., with pedicels said to be to 14 mm. long, has none so long.
Peduncles of 2- or more-flowered branches were measured. There is
variation in the size of flowers and consequently of the petals and
sepals, but the series is so complete that it seems impossible to use
these measurements in the separation of the species proposed or
maintained by Dr. Blake. Leaf size and shape do not seem to be
sufficiently consistent to be used in specific separation, as might be
expected. The presence or absence of tufts of pubescence in the axils
of the veins on the lower surface of leaves seems not at all consistent,
even on leaves of a single specimen.

Macrohasseltia, A New Genus of the Flacourtiaceae

MACROHASSELTIA L. Wms., gen. nov.

Arbores. Folia 3-plinervia, supra basi 1-2-glandulosa, vel glandulae nullae.
Flores hermaphroditi. Sepala 5-6, valvata. Petala 0. Stamina numerosa, libera,
disco inserta; anthera parva, subglobosa; filamenta pilosa, sepala aequalia. Ovar-
ium unilocularis. Stylus perbrevis. Stigma 4-5-lobata, stellata. Capsula ovata,
indehiscens. Inflorescentia paniculata vel subcymosa, pauciflora.

Trees. Leaves alternate, petiolate, with 3 prominent nerves near the base
(subtriplinerved), the base of the leaf with glands or eglandular, the petiole at its

364 FIELDIANA: BOTANY, VOLUME 29

junction with the blade much swollen and perhaps glandular on older leaves. In-
florescence paniculate or subcorymbose, relatively few-flowered. Flowers perfect.
Sepals 5-6, valvate, coriaceous, soon deciduous. Petals none. Stamens many
(±100), free, borne at the base of the sepals and disc, not in bundles, with stami-
nodes or undeveloped stamens mixed in at the base of the filaments; anthers small,
subglobose; filaments about as long as the sepals, pilose. Style very short or none,
the stigma 4-5-lobed, stellate. Ovary 1-celled. Fruit probably an indehiscent
capsule.

Type species: Hasseltia macroterantha Standl. & L. Wms.

Macrohasseltia macroterantha (Standl. & L. Wms.) L. Wms.,
comb. nov. Hasseltia macroterantha Standl. & L. Wms. Ceiba 3:
53. 1952.

Costa Rica: Flowers whitish, tree 30 m., vicinity of El General,
province of San Jose", alt. 670 m., June 1939, Skuteh J$37 (US, F, G) ;
"chancho bianco," tree 35 m., flowers greenish, basin of El General,
province of San Jose", alt. 675-900 m., May 1940, Skuteh 4911 (US).

The genus belongs in the Scolopieae-Prockiinae of Gilg's treat-
ment of the Flacourtiaceae in Engler & Prantl, Nat. Pflanzenfam.
ed. 2, 21: 421. 1925. It is related to Pleuranthodendron and Prockia,
and also to Hasseltia. It may be distinguished from these genera by
the star-shaped, 4-5-lobed stigma that is essentially sessile on the
ovary, by the lack of petals (some species of Prockia also without
petals), by the deciduous sepals and by the very large number of
stamens which are slightly shorter than the expanded sepals, and
among these may be staminodes or undeveloped stamens.

The leaf form and nervation especially recall those of Pleurantho-
dendron and Hasseltia, with main nerves arising very near the base
of the blade. The glands are not prominent at the base of the leaves
as in the genera mentioned but exist on some of the leaves. The
petiole at its juncture with the leaf blade, on old leaves, is much
swollen or perhaps glandular; this is not so with younger, smaller
leaves of the type specimen.

The generic name Macrohasseltia is selected to indicate relation-
ship to Hasseltia. The tree is perhaps larger than any of those known
to occur in the allied genera.

Pleuranthodendron, A New Genus of the Flacourtiaceae

Sleumer (Notizbl. 14: 49. 1938) described a new genus of the
Flacourtiaceae to which he applied the name Hasseltiopsis. The
type species of this genus is Banara dioica Benth. (Journ. Linn. Soc.
5, suppl. 2: 94. 1861). At the same time Sleumer described two new

WILLIAMS: TROPICAL AMERICAN PLANTS, II 365

species of the genus, H. albomicans, from Colombia, and H. leuco-
thyrsa, from Peru.

Sleumer considered Banara dioica Benth. and B. mexicana Gray
(Proc. Am. Acad. 5: 174. 1862) to be synonymous, which is not the
case. The two species do not belong to the same genus. The first
belongs to the genus Banara and the second to the genus described
here. It was obviously Sleumer's intent to base his genus Hassel-
tiopsis on the common Mexican plant which Gray had had, for
Sleumer's description fits it very well and the two new species which
he described in his genus I believe to be synonyms of Gray's species.
It is unfortunate that Sleumer took the name Banara dioica Benth.
as the type species of his genus, which, for that reason, becomes a
synonym of Banara, for I believe that B. dioica of Bentham is a
Banara.

Banara dioica Benth. is apparently a rare species which is known
to me only by specimens from the state of Veracruz in Mexico and
from the department of Santa Barbara in Honduras (Mexico: Gale-
otti 7019; Liebmann 15031, 15032. Honduras: Allen 6372}.

PLEURANTHODENDRON L. Wms., gen. nov.

Arbores. Folia alterna, serrata vel Integra, petiolata, basi laminarum bigland-
ulosa. Inflorescentia panicula diffusa. Flores hermaphroditi; sepala 4-6, valvata;
petala 4-6; stamina indefinita; antherae parvae; ovarium superum, sessile, tomen-
tosum; capsula subcoriacea.

Small trees. Leaves alternate, petiolate, biglandular at the junction of the
blade and petiole, the glands sometimes obscure, tri-quinqueplinerved at the base,
the margins entire to serrate, the teeth sometimes gland-tipped. Inflorescence
often a diffuse panicle or panicles originating in the axils of the upper leaves or
terminal, the bracteoles early deciduous and the scar often glandular. Flowers
perfect; sepals 4-6, valvate; petals 4-6, alternate with the sepals, soon falling.
Stamens numerous, small, bilocular, the cells opening longitudinally. Ovary supe-
rior, sessile, tomentose, 1-celled, placentae 3-4. Fruit a subcoriaceous capsule.

Type species: Banara mexicana Gray.

The generic name is suggested by the large number of flowers on
this tree, which collectors have commented on. The genus is allied
to Hasseltia, Prockia, and Macrohasseltia.

Pleuranthodendron mexicana (Gray) L. Wms., comb. nov.
Banara mexicana Gray, Proc. Am. Acad. 5: 174. 1862. Hasseltia
pyramidalis Hemsl. Diag. PL Mex. 4. 1878. Hasseltia mexicana
Standl. Contr. U. S. Nat. Herb. 23: 840. 1923. Hasseltiopsis leuco-
thyrsa Sleumer, Notizbl. 14: 51. 1938. Hasseltiopsis albomicans

366 FIELDIANA: BOTANY, VOLUME 29

Sleumer, I.e. Banara leucothyrsa Macbr. Field Mus. Bot. 13 (4):
27. 1941. Hasseltia quinquenervia Standl. & L. Wms. Ceiba 3: 54.
1952. Hasseltiopsis mucronata Cuatr. Tropical Woods 101: 17. 1955.

A widely distributed species occurring from Mexico through Cen-
tral America to Colombia, Peru and Brazil. There is some variation
in the species, as might be expected in one occurring over such a wide
area. However, there seems to be nothing constant on which the
several species proposed may be maintained. Much of the material
from the northern part of the range of the species seems to have the
leaves somewhat smaller and more tapering to the petiole than is
the case in the specimens collected from Costa Rica southward.
Cuatrecasas calls attention to the mucronate anthers of Hasseltiop-
sis mucronata Cuatr. The type specimen has several flowers opened
and the anthers visible; in these the anthers seem to be more often
lacking a mucro than not. Specimens vary in the quantity of the
pubescence; those from the southern part of the range usually have
the inflorescence and the twigs a bit more pubescent than is the case
in those from the north. The leaves when young may be obscurely
puberulent below.

TETRATHYLACIUM Poepp. & Endl.

The genus Tetrathylacium is one of the most distinctive of the
American genera of the Flacourtiaceae, and one about which there
is relatively little information. There was but one species known
for many years; then it was found that Edmonstonia pacificum Seem,
was of the same genus (and quite possibly the species is the same).
In 1925 Standley described two species from Central America and
more recently Sleumer described one from Ecuador.

In Bentham & Hooker's Genera Plantarum and in Gilg's account
of the Flacourtiaceae (Nat. Pflanzenf. ed. 2, vol. 21) there are some
discrepancies of observation, even though the authors of these two
works knew only the original species of the genus. In Bentham &
Hooker's account the sepals are said to be none and the petals 4.
Gilg gives the petals as none.

In T. macrophyllum each flower is subtended by a bract that pro-
jects out from beneath what I believe to be the calyx. This structure
is firmly attached to the rachis and is persistent; it is rotate and ob-
scurely 3-4-1 obed. The corolla is urceolate-globose and exactly as
described by Bentham & Hooker. The corolla is persistent on the
bases of the fruits and falls away with them.

WILLIAMS: TROPICAL AMERICAN PLANTS, II 367

In T. Johansenii Standl. the structure which I believe to be the
calyx is very much more reduced than in T. macrophyllum and con-
sists of very thin membranaceous lobes. These are not persistent on
the rachis. The flowers in this species are crowded densely on the
secondary rachises, much as in some species of Peperomia or in some
aroids, and each flower is a part of a "mosaic," the visible portion
rhombic in outline.

There are now recorded from Central America three species of
Tetrathylacium: T. macrophyllum, mentioned below; T. costaricense
Standl., recorded from Costa Rica and dubiously distinct from
T. macrophyllum; and T. Johansenii Standl., which is known from
Panama and is very distinct from the type species of this genus.
It is reported also from Colombia but our specimen is too poor to
verify the determination.

Tetrathylacium macrophyllum Poepp. & Endl. Nov. Gen. &
Sp. 3:34, t. 240. 1842.

Panama: Fruit red, tree 10 m. tall, hills north of El Valle de An-
ton, alt. 1000 m., June 23, 1940, Allen 2147.

The Panamanian material seems to differ from typical material
of T. macrophyllum Poepp. & Endl. only in having the leaves broader
in relation to their length. It is the first report of the species for
North America.

In the herbarium of Chicago Natural History Museum there is a
specimen labeled "2096 (Tetrandria) ? Maynas, 1831." The leaf on
this specimen is so nearly like that shown on the Poeppig and End-
licher plate, cited above, that it may well be the leaf that was used
by the artist for the illustration.

Erblichia odorata Seem. Bot. Voy. Herald 310, t. 27. 1854.
Piriquetia xylocarpa Sprague & Riley, Kew Bull. 1923: 373. 1923.
Erblichia xylocarpa Standl. & Steyerm. Field Mus. Bot. 22: 353.
1940. E. Standleyi Steyerm. Field Mus. Bot. 22: 356. 1940.

This is the type species of the genus, now known from Mexico,
Guatemala, British Honduras, El Salvador, Honduras, Costa Rica
and Panama.

Standley and Steyermark (Field Mus. Bot. 22: 351-357. 1940) in
their "Notes on the American species of Erblichia" indicate that the
ovaries of this species are velutinous. Seemann's plate indicates that
the ovary and the styles are somewhat pubescent. In the descrip-
tion, however, indument of the ovary is not mentioned and it well

368 FIELDIANA: BOTANY, VOLUME 29

might have been had it been velutinous. I find that the young cap-
sule has not a vestige of pubescence in a specimen collected by Paul
Allen in Code" province, Panama (number 2468) . In other specimens
which I would place in this species the variation in pubescence on the
ovary is from none to velutinous. The same variation is to be found
in pubescence of twigs and petioles. Even in the limited number of
specimens of this species which I have at my disposal a fairly com-
plete series is to be found.

Erblichia odorata var. mollis (Standl. & Steyerm.) L. Wms.,
comb. nov. E. xylocarpa var. mollis Standl. & Steyerm. Field Mus.
Bot. 22: 355. 1940.

Known only from southern Mexico and adjacent Guatemala.

The variety differs from the species in having the under surface
of the leaves velutinous or almost tomentose. The specimens avail-
able seem to indicate that this variety occurs at greater elevations
than does the typical one.

Carica mexicana (A. DC.) L. Wms., comb. nov. Jacaratia
mexicana A. DC. Prodr. 15, pt. 1: 420. 1864. Jacaratia conica Ker-
ber in Eichl. Jahrb. Bot. Gart. Berl. 2: 282, t. 9. 1883. Carica hepta-
phylla Sesse" & Mocifio, PI. N. Hisp. 172. 1887, non Velloso, 1827.
Carica dolichaula Donn.-Sm. Bot. Gaz. 23: 247. 1897. Pileus hepta-
phyllus Ramirez, Anal. Inst. Med. Nac. Mexico 28: t. 1-4- 1901.
Leucopremna mexicana Standl. Contr. U. S. Nat. Herb 23: 850. 1924.
Pileus mexicanus Johnston, Contr. Gray Herb. n.s. 70: 79. 1924.
Jacaratia dolichaula Woodson, Ann. Mo. Bot. Gard. 37: 404. 1950;
45:23,/. 1958.

Carica mexicana is distributed from Mexico to Panama. It is
recorded from all the Central American countries except Honduras,
and I believe that I have seen it on the "North Coast" there but was
unable to collect it. The variation in the leaves of the species is
rather great. Most of our specimens have 3-5-digitate leaves. The
drawing upon which the species is based shows 5-7-digitate leaves.
One specimen available (Steyermark 45489} has entire and 3-digitate
leaves, while still another (Gaumer 321) has a detached 3-lobate leaf.
The single detached leaf on Gaumer 321 bis is 3-digitate. Steyer-
mark's number 45112 has digitate-lobate leaves.

The related South American Carica heterophylla P. & E. shows
even greater variation in the leaves than does C. mexicana. The
leaves may be simple, lobate, digitate, or digitate-lobate.

WILLIAMS: TROPICAL AMERICAN PLANTS, II 369

Cuphea cyanea DC. Prodr. 3: 85. 1828. C. cyanea var. hirtella
Koehne in Engler, Bot. Jahrb. 2: 417. 1882.

A relatively uncommon species of Cuphea which occurs from cen-
tral Mexico to Guatemala. It has been divided by Koehne into two
varieties, each with two formae. The original was based by de Can-
dolle on a Sesse" and Mocino drawing, of which we have a photograph.
I believe that the original must be the same as Koehne's var. hirtella
which, if it must have a varietal name, would be C. cyanea var.
cyanea. Guatemalan material has usually been determined as var.
hirtella.

Cuphea micrantha HBK. Nov. Gen. & Sp. 6: 196. 1823.
Cuphea Valerii Standl. & L. Wms. Ceiba 3: 125. 1952.

In reviewing Cuphea for the Flora of Guatemala it became appar-
ent that C. Valerii, described by Standley and myself from Honduras,
was very close to, if not identical with, C. micrantha, originally de-
scribed from Venezuela. The species was reported first from Guate-
mala by Koehne. The problem is further complicated in that the
relationship of Cuphea micrantha, C. Wrightii A. Gray and C. cartha-
genensis (Jacq.) Macbr. is apparently very close. It is possible that
the three species mentioned are not distinct but should be referred
to C. carthagenensis, perhaps as varieties.

Cuphea mimuloides Cham. & Schlecht. Linnaea 5: 570. 1830.
Maja hypericoides Klotzsch in Schomb. Fl. & Faun. Guian. 1191.
1848, without diagnosis. Cuphea mimuloides var. guianensis Koehne
in Engler, Bot. Jahrb. 1: 446. 1881. Cuphea mimuloides var. hyperi-
coides Koehne in Engler, Pflanzenr. IV. 216: 101. 1903.

Originally described from Mexico and reported by Koehne from
Cuba and also from British Guiana. We have several collections
now from Central America.

Guatemala: Between Agua Blanca and Amatillo, Dept. Jutiapa,
Steyermark 301+11.

British Honduras: Maskall Pine Ridge, Gentle 1042.

Honduras: San Juan del Rancho, Dept. Morazan, Molina 1718;
same locality, Standley 14.277, 15143, 23858, 25036; same locality,
Williams 17034; Mt. Uyuca, Dept. Morazan, Williams & Molina
11035; San Marcos, Dept. Choluteca, Williams & Molina 10896.

Cassipourea guianensis Aubl. Hist. PI. Guian. Fr. 1 : 529, t.211.
1775. Legnotis elliptica Sw. Prodr. 84. 1788. Cassipourea elliptica

370 FIELDIANA: BOTANY, VOLUME 29

Poir. Encycl. Suppl. 2, 131. 1811. C. podantha Standl. Field Mus.
Bot. 4: 241. 1929. C. macrodonta Standl. I.e. 242. C. belizensis Lun-
dell, Bull. Torr. Bot. Club 66: 598. 1939.

Range: Mexico, Guatemala, British Honduras, Honduras, Nica-
ragua, Costa Rica, and Cocos Island, Panama; northern South Amer-
ica through the Amazon Basin and from Colombia to Peru.

A study of the rather abundant material of Central American
Cassipourea seems to indicate that there is but one species repre-
sented and that this can not be separated from C. guianensis Aubl.
The range of variation of the Central American material can be
matched by that from the West Indies and South America. The
dentate-leaved C. macrodonta is matched in South America by C. den-
tata Benth. The forms with dentate leaves are not constant and some
specimens have both entire and dentate leaves.

Cassipourea guianensis and C. elliptica have been maintained as
distinct species, the first having sessile or subsessile flowers while
the second has pedicellate flowers. In our collections there is a series
of specimens with all gradations, from those having no pedicels or
very short ones to those having pedicels as long as or longer than the
flower buds. There are minor differences of pubescence — of the
ovary, of the exterior surface of the calyx, of the pedicels, and of
the petiole — but I find nothing constant.

Combretum fruticosum (Loefl.) Stuntz, U. S. Dept. Agri. Bui*.
PL Ind. Seed & PL Imp. No. 31: 86. 1914. Gaura fruticosa Loefl.
Inter. Hispan. 248. 1758. Combretum secundum Jacq. Enum. 19.
1760. C. farinosum HBK. Nov. Gen. & Sp. 6: 110. 1823. C. Wars-
zewiczianum Eichler in Mart. Fl. Bras. 12 (2) : 110. 1867. C. Benth-
amianum Van Heurck & Muell. Arg. in Van Heurck, Obs. Bot. 220.
1871. C. farinosum var. phaenopetalum Donn.-Sm. Bot. Gaz. 23: 7.
1897. C. superbum Pittier, Contr. U. S. Nat. Herb. 18: 242. 1917.
C. polystachyum Pittier, I.e. 243. C. phaenopetalum Pittier, I.e. 243.
C. lepidopetalum Pittier, I.e. 245.

In his revision of the species of Combretum in the New World
(Journ. Linn. Soc. Lond. 55: 116-121, ff. 1953) Exell maintains as
distinct Combretum farinosum and C. fruticosum, which occur from
Mexico to Argentina and are to be found in all of the Central Amer-
ican countries. When I was revising the material in our herbarium,
some 200 collections, it seemed easy at first to separate the collections
into the two species maintained by Exell. However, a closer exami-
nation of the material to find stable characters revealed that the

WILLIAMS: TROPICAL AMERICAN PLANTS, II 371

characters used to separate the two species are not consistent. The
coloration and the size of the scales in the inflorescence and on the
leaves seem to intergrade and perhaps also are affected to some ex-
tent by the age of the plant. The apparent differences in the calyx
and in the petals seem to break down under examination.

Mitracarpus rhadinophyllus (Rob.) L. Wms., comb. nov.
Borreria rhadinophylla Rob. Proc. Am. Acad. 45: 409. 1910. Mitra-
carpus villosus var. glabrescens Griseb. Cat. PL Cub. 143. 1866.
Mitracarpus glabrescens Urban, Symb. Antill. 7: 551. 1913.

Apparently most abundant in British Honduras, which was the
source of the specimen described by Robinson, but found also in
Mexico (Baja California, Tabasco, Mexico, perhaps other states),
Nicaragua and Cuba.

Allenanthus erythrocarpus Standl. Ann. Mo. Bot. Card. 27:
344. 1940. Chimarris decurrens Steyerm. Ceiba 3 : 18. 1952.

The details of the types of these two proposed species from Pan-
ama are identical and there can be no question that they represent
only one species.

Microspermum repens (Blake) L. Wms., comb. nov. Iltisia
repens Blake, Journ. Wash. Acad. Sci. 47: 409. 1959.

Costa Rica: Cerro de la Muerte, alt. 3400-3500 m., Holm &
Iltis 594-

The genus Microspermum previously contained two species, M.
numulariaefolia Lag. and M. debile Benth., both known only from
Mexico. These two species differ from one another, among other
characters, in that the summit of the achenes in M. numulariaefolia
bears a pappus of two (or three) bristles, in addition to an obscure
corona. The achenes of M. debile lack the pappus bristles.

The outer cycle of flowers in both Mexican species is usually dif-
ferentiated from the inner flowers by having the corolla enlarged and
somewhat ray-like. The three exterior lobes of these marginal flowers
are variously enlarged, while the two inner lobes are but slightly en-
larged or are about the same size as the lobes on disc flowers. A few
heads of M. numulariaefolia seem not to have "ray flowers."

The generic characters for the group would seem to include achenes
with 2 (-3) pappus bristles, or none, the heads pseudoheterogamous
or homogamous. The flowers on the only known collection of Iltisia

372 FIELDIANA: BOTANY, VOLUME 29

are immature, and heterogamous heads do not occur or were not
noted. The achenes are without a pappus of bristles. The Costa
Rican plant seems to fit into Microspermum and seems to be most
closely related to M. debile Benth.

Publication 926