•••

I/I B RARY

OF THE

UNIVLRSITY
OF ILLINOIS

5&O.5

8101061

The person charging this material is re-
sponsible for its return on or before the
Latest Date stamped below.

Theft, mutilation, and underlining of books
are reasons for disciplinary action and may
result in dismissal from the University.

University of Illinois Library

DEC 211! 70

L161— O-1096

FIELD MUSEUM OF NATURAL HISTORY

FOUNDED BY MARSHALL FIELD, 1893

PUBLICATION 291

BOTANICAL SERIES

VOL. XI, No. 2

SPERMATOPHYTES, MOSTLY PERUVIAN-IV

BY

J. FRANCIS MACBRIDE

ASSISTANT CURATOR OF TAXONOMY

T"'

THE

AUG 26 1931

B.E.DAHLGREN UNH

ACTING CURATOR, DEPARTMENT OF BOTANY
EDITOR

CHICAGO, U. S. A.
JULY 29, 1931

FIELD MUSEUM OF NATURAL HISTORY

FOUNDED BY MARSHALL FIELD, 1893

PUBLICATION 291
BOTANICAL SERIES VOL. XI, No. 2

SPERMATOPHYTES, MOSTLY PERUVIAN-IV

BY

J. FRANCIS MACBRIDE

ASSISTANT CURATOR OF TAXONOMY 1M|£ i inpffW ' 7 Tlir

AUG 26 1931
UNIVERSITV b? ILLINOIS

B. E. DAHLGREN

ACTING CURATOR, DEPARTMENT OP BOTANY
EDITOR

CHICAGO, U. S. A.
JULY 29, 1931

PRINTED IN THE UNITED STATES OF AMERICA
BT FIELD MUSEUM PRESS

n

SPERMATOPHYTES, MOSTLY PERUVIAN— IV

J. FRANCIS MACBRIDE

In presenting another paper written at the Berlin-Dahlem her-
barium, I acknowledge with gratitude my continued indebtedness
to my friendly hosts. Because of the complete freedom of study
permitted at Dahlem, it has been possible to elucidate Peruvian
types deposited there, to, I trust, their better understanding. Misde-
termination of collections is so frequent because of misinterpretation
of descriptions, even when perfectly drawn, that any aid, however
imperfect, should prove helpful. I hope, therefore, that my attempt
to interpret the Dahlem types, as they concern Peru, will be a prac-
tical evidence of my appreciation of the privileges enjoyed in research
in that great herbarium.

The present paper includes descriptions of two new species of
Scleria by Mr. R. Gross of Berlin, based upon material obtained by
Field Museum expeditions.

1. NEW CYPERAGEAE

Scleria Williamsii R. Gross, spec. nov. Culmus 3-1 mm. in
diam., triqueter, simplex (?), firmus, sparsim hirtus, in angulis
laevis, partes circa 56 cm. altae. Vaginae angustae, circa 3 cm.
longae, multistriatae, minute hirtae, ore truncatae, dense hirtae.
Folia circa 24-29 cm. longa, 6 mm. lata, plana, subrigida, supra
subtilissime celluloso-reticulata, nervis 2 prominentibus, glabra vel
glabrescentia, in parte superiore perscabra, subtus carinata, minutis-
sime hirta, in marginibus laevia. Panicula terminalis altera partiali
axillari multo minore addita 6-8 cm. longa, 4-7 cm. lata; paniculae
axillares (2) 7-11 cm. lg., 4 cm. lat., pedunculi 4.5, 2.5, (7) cm. vel
4, 3.5, (11) cm. lg.; rami 3, 4 vel 7, divaricati, basi ochrea vaginata
et bractea (circa 1-2 cm. lg.) praediti, inter se circa 10, 8, 7, 8, 5,
10, 5 mm. vel 10, 9, 9, 7, 7, 10, 10, 7 mm. distantes, circa 4, 3, 2.5,
2, 1.5, 1, 1, 0.5 cm. lg., puberuli, 5-, 4-, 3-, 2- vel 1-stach. Spiculae
unisexuales, solitariae, interdum longe pedicellatae, 6 mm. longae.
Squamae 4-6, distichae, chartaceae, ovato-acuminatae, carinatae,
fere glabrae, viridescentes, margine rubescentes. Nux quam squamae
brevior globoso-trigona, circa 4 mm. longa, 3 mm. lata, alba (rarius
pallide virescens), maculis leviter obscuris, minute verrucosa, parce
minute pilosa, rostrum conicum plus minusve recuryum. Discus
trilpbus, lobi rotundati, concavi, marginibus involutis, chartaceo-
coriacei, pallide brunnei. — Peru: Alto Rio Huallaga, Dept. San
Martin, alt. 360-900 m., August 12, 1929, Llewelyn Williams 5823
(type, Field Museum).

39

40 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

Die Species ist entfernt verwandt mit S. tenacissima Nees oder
S. flagellum Sw., aber sie weicht ganz erheblich ab. Recht auffallig ist
die umgebogene Spitze der Nuss, ein ganz characteristisches Merkmal.

Scleria Macbrideana R. Gross, spec. nov. Culmus 3-1 mm. in
diam., triqueter, in angulo retrorsum remote tenuiter spinulosus,
scandens (?), vero-similiter subramosus, partes (solum visae) 37,48 cm.
altae. Vaginae 3.5-3 cm. lg., submicroscopice hirtae, multistriatae,
ore subtruncatae vel abbreviato-ovatae, breviter ciliato-hirtae. Folia
(vagina inclusa) circa 23-31 cm. longa, 5 mm. lata, anguste yaginata,
linearia, longe acuminata, plana, supra nervis 2 prominentibus, fere
glabra vel basi glabrescentia, subtus leviter carinata carina margine-
que retrorsum tenuiter spinulosa. Paniculae 5 adsunt, 4 axillares,
laxe divaricatae, 14.5, 14.5, 12.5, 6, 5.5 cm. longae, 3.5 cm. latae;
pedunculi 9, 9, 8, 2.5 cm. lg.; rhachis scabriuscula; rami 6-7, plus
minusve horizontals, scabri, inter se 10, 10, 4, 4, 6, 4, 5 mm. vel
12, 10, 9, 5, 6, 4, 4 mm. distantes, basi ochrea crasso-vaginata bractea
setacea (circa 3 mm. lg.) praediti; inferiores plerumque 4-stach.
vel -flor. (rarius 2), 2, 2.5 vel 3.5 cm. lg., partes circa 11, 5, 4,
3 mm. lg. ; superiores 1- vel 2-stach. Spiculae unisexuales, solitariae,
pedicellatae, ellipticae, 5 mm. lg., 3 mm. lat.; masculae paucae.
Squamae circa 9, sursum longitudine accrescentes, ovatae, carina
excurrente, glabrae, rubridae. Nux conspicue exserta late conica,
3 mm. longa, 2 mm. lata, minutissime transversim albo-pilosa,
intense violacea, basi saepe alba. Discus obpyramidalis, lobi crasso-
coriacei, irregulariter undulati, flavido-rubri. — Peru: La Victoria on
the Amazon River, Dept. Loreto, August 19, 1929, Llewelyn Williams
2571 (type, Field Museum).

Diese Art ist leicht zu erkennen an den kleinen, breit kegelfb'rm-
igen, intensiv violetten, am Grunde oft weissen Niissen, die in
Querreihen mit ausserst f einen Harchen bedeckt sind ; f erner an den
lockeren Rispen mit fast wagerecht abstehenden Aesten, an denen
entfernt die einzelnen Aehrchen stehen. Die vegetativen Teile
erinnern an S. flagellum Sw. und S. tenacissima Nees.

Rhynchospora Killipii R. Gross, spec. nov. Species imperfecte
nota. Culmus circa 1.20 m. altus (pars superior 60 cm.), 3-2 mm.
in diam., strictus, triqueter faciebus concaviusculis, dense striatus,
glaber. Internodia 22.5, 16 cm. longa. Vaginae superiores 3.5, 2.5
cm. lg., glabrae, profunde sinuosae (circa 1 cm.). Folia superiora
(1 et 0.5) circa 70 cm. longa, circa 12-14 mm. lata, linearia, longe
attenuate acuminata, multistriata, supra subtilissime celluloso-
reticulata; subtus tenuiter carinata, minutissime puncticulata,
margine brevissime spinulosissima. Inflorescentia paniculata, stricta,
21.5 cm. longa, 8 cm. lata, interrupta (10, 1.5, 1, 1, 1, 1 cm., 5 mm.,
5, 5, 5, 5 mm.); terminalis (altera partiali axillari 10 cm. distante
addita). Bracteae 3 inferiores inflorescentiam longe superantes
(bractea tertia 18 cm. longa, 7 mm. lata) plus minusve puberulae

PERUVIAN PLANTS 41

margine ciliatae, interdum spinulosissimae; superiores perangustae,
valde reductae. Rhachis, rami, ramuli dense breviter piloso-hir-
sutuli. Ramus axillaris stricte erectus, 9 cm. longus; rami 4 plus
minusve patentes (6, 4, 4, 3.5 cm. longi) ; superiores (circa 8) approxi-
mati (2.5-1 cm.); ramuli circa 1.5-0.5 cm. longi, dense ramosi.
Bracteolae 2 diversae, setaceae, ovatae. Spiculae brevissime pedun-
culatae vel subsessiles, solitariae, superiores saepe approximatae,
oblique patentes, conico-lanceolatae, acutae, ad 17 squamae, 1-
nucigerae, 5-7 mm. lg., 1 mm. lat. Squamae chartaceae, ellipticae,
carinatae, mucronatae, stramineae, inferiores 10 vacuae, 11 nucu-
ligerae; superiores involutae, lanceolatae, breviores, steriles. Stamina
3; filamenta brevia; antherae 3 mm. lg. Setae 5, aequilongae, antice
scabridae. Nux nimis juvenilis; stylus divisus. — Peru: Dept. Junin,
Pichis Trail, Yapas, 1,350-1,600 m., dense forest, Kittip & Smith
25566 (type, Field Museum).

Vorhanden ist ein festgeklebtes Bruchstiick von 60 cm. Lange;
neben der iiblichen Standortsangabe steht auch die Notiz "to 4 ft.
high." Auffallig ist die steife dickstielige terminale Infloreszenz mit
einer 10 cm. tiefer stehenden axillaren, die dicht am Halm auf warts
ragt. Die einzelnen Aeste sind mit ihren ziemlich gleichmassig
verteilten Aestchen einander sehr ahnlich. Ein recht auffalliges
Merkmal sind die 3 sehr langen (2 sind allerdings nur unvollstandig
vorhanden) gewimperten, unteren Brakteen, an der Terminal-In-
floreszenz und auch die sehr zahlreichen (cr. 17) Schuppen der
Aehrchen. Die Niisse sind leider so wenig entwickelt, dass dariiber
nichts gesagt werden kann.

Calyptrocarya Poeppigiana Kunth, forma grandifolia R.

Gross, f. nov. Folia ad 2.5 cm. lata, 6 dm. longa. — Peru: Mishuyacu
near Iquitos, Klug 377 (type, Field Museum).

Although a number of collections of this species are before me,
the leaves of none approach in size those of this material. It seems,
therefore, worthy formal designation.

2. MONOCOTYLEDONS, MOSTLY SCITAMINEAE
Podocarpus Harmsianus Pilger, Pflanzenr. IV. 5: 68. 1903.
The Arnold Arboretum apparently regards this species and P.
montanus (Willd.) Lodd. as the same, or at any rate that institution
is authority for the identification of Killip & Smith 22480 as the
latter, although Professor Pilger refers it to the former. Little
known as both species are, they appear to be well distinct, as the
leaf nerve in P. montanus is impressed, and in P. Harmsianus not
at all.

42 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

Elodea Potamogeton (Bert.) Espinosa, Rev. Chil. Hist. Nat.
31: 150. 1928. Anacharis Potamogeton (Bert.) Viet. Contr. Bot.
Montr. 18: 41. 1931.

In the last paper of this series I inadvertently published the above
combination as my own. Frere Marie- Victorin has given a helpful
resume of the status of this group, but his arguments for discarding
the name Elodea, which might have been well taken before the
botanical congress of 1930, are of questionable significance now,
since the case is one requiring interpretation of the rules and to be
decided legally only by committee.

Dichromena Kuntzei (Clarke), comb. nov. Rhynchospora
Kuntzei Clarke, Kew Bull. Add. Ser. 8: 39. 1908. R. Uleana Kiikenth.
fide Gross in Herb. Dahlem.

Mr. R. Gross has referred the plant of Ule in Herb. Dahlem to
this species which, as remarked by Clarke, simulates Pleurostachys
macrantha. It supports, from a practical standpoint, my suggestion
(Field Mus. Bot. 11: 6. 1931) that at least in regional treatments
the group Pleurostachys is best regarded as a section of Dichromena.

Dichromena cariciformis (Nees), comb. nov. Rynchospora
cariciformis Nees in Mart. Fl. Bras. 21: 145. 1842. R. trichophora
Nees, loc. cit. fide R. Gross.

This very unusual species with carex-like spikes has been redis-
covered by Williams (3457) on the Rio Itaya, and in the same
region by Killip and Smith, according to Mr. Gross. These are,
apparently, the first Peruvian records.

Cyperus cayennensis (Lam.) Britton, var. redolens (Maury)
R. Gross, comb. nov. C. redolens Maury, Me"m. Soc. Phys. Geneve,
31: 126. pi 36. 1889.

This plant is distinguishable from the typical form only by the
more numerous flowers — 6-8 (10) — and it may, therefore, be treated
as a variety. The necessary combination does not seem to have
been made under the earliest specific name, i.e. C. cayennensis (C.

flavus Nees).

Carex fecunda Steud., var. atropurpurea (Boeckl.), comb,
nov. C. atropurpurea Boeckl. Linnaea 39: 150. 1875.

In Field Mus. Bot. 8: 113. 1930 I suggested that this is a
recognizable plant. Recently, in looking at the type again with
Mr. Gross, the fact has been called to my attention with renewed
force, therefore the Boeckler name may be removed from synonymy,
as indicated above.

PERUVIAN PLANTS 43

Paepalanthus aequalis (Veil.), comb. nov. Dupatya aequalis
Veil. Fl. Flum. 1: pi. 85; text 36. 1827. P. blepharocnemis Mart,
ex Koern. Fl. Bras. 31: 376. 1863.

As recent an authority on this group as Ruhland has questioned
not at all the equality of the names of Vellozo and Martius; the
former, therefore, having priority, is the name to be retained.

Juncus andicola Hook. Icon. 8: pi. 714- 1848 and J. Lesueurii
Boland. Proc. Calif. Acad. Sci. 2: 179. 1863 are undoubtedly very
close and perhaps the latter were better treated as a variety,
but so far as Peruvian material is concerned the problem is easy,
as the Californian plant certainly is not represented there, in spite
of herbarium determinations to the contrary. It very doubtfully
occurs in Chile and Argentina, the flowers and anthers being smaller.
Buchenau's separation is not clear (cf. Pflanzenr. IV. 36: 126, 147-
148. 1906) because he distinguishes the species on the proportion-
ate length of anthers and filaments. The diagnostic character of
J. Lesueurii seems to me to be its large anthers, nearly 2 mm. long,
on a filament 0.5 mm. long. It is typically a slender plant with few-
flowered or capitate inflorescence, the segments bicolored and the
stems shortly sheathed at base (3-6 cm.). J. andicola Hook, has
small anthers at most 1.25 mm. long or shorter, on filaments 0.5-1
mm. long. The proportion between anther and filament is therefore
not a satisfactory means of distinguishing the species, but there
seems to be a constant difference in the size of the anthers.

J. andicola was described from Ecuadorian material, and typically
its anther is 1.25 mm. long, its filament nearly 1 mm. long. It has
rather slender, wiry stems sheathed at base for a length of 1.5 dm.
and many nearly concolored and dark flowers. It seems to be ex-
actly matched by Colombian and Bolivian specimens, and Argen-
tina and Chile material (referred to J. Lesueurii) probably belongs
to it, at least as a variety.

In Peru also it has been collected, but not in exactly typical form,
occurring there only as a much more robust plant, the stems above
the elongate basal sheaths 5-12 mm. thick, and greatly prolonged,
not infrequently 2 dm., above the inflorescence. This Peruvian
state is puzzling in that it seems to exist in two forms. One, Weber-
bauer 3283 from the Yananguco See in Ancash, referred by Buchenau
to J. Lesueurii, is nearly J. andicola except for the greatly elongate
stems. It is very young, but the anthers, though about twice as
long as the filaments, are small, that is, 1 mm. long. The other form,
referred by Buchenau to J. andicola, is so at variance that it seems

44 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

worth while to regard it as a variety, even if the plant first described
is a connecting state. This is Weberbauer 1367 from Pucara, Puno.
In its almost gigantic size, its pale flowers, and particularly its smaller
stamens it seems nearly specifically separable, but these characters
are admittedly variable. This plant, so interesting even if only a
variant, may bear the name of Mr. Schulz-Korth, and give me also
the opportunity of expressing thus my indebtedness for many
courtesies in herbarium work.

Juncus andicola Hook., var. Schulz-Korthii, var. nov.,
robustus, 10-12 dm. altus; caulibus crassis diam. circa 11 mm.;
cataphyllis ad 1.5 dm. longis; bracteis valde elongatis; inflorescentiis
dense congestis; tepalis 5-6 mm. longis pallide brunneis vel lateribus
castaneis; antheris 1.2 mm. longis; filamentis circa 0.5 mm. longis;
capsula vix 4 mm. longa; seminibus fuscis circa 0.6 mm. longis.—
Peru: Pucard, Dept. Puno, 3,600 m., forming large stands in a lake,
Weberbauer 1367 (type, Berlin-Dahlem).

I acknowledge with appreciation the kindness of the Director of
the Dahlem Museum and of Dr. Markgraf , the curator of the family,
in permitting the above publication, based entirely on a study of
material in their care.

Smilax gilva, spec, nov., gracilis; ramis ramulisque teretibus
3 mm. crassis laevibus viridibus non lineatis, punctatis, tuberculatis
vel aculeatis; petiolis tortuosis circa 1 cm. longis, basi anguste vagi-
nantibus, vaginis valde apiculatis ut videtur non cirrhiferis; foliis
ovatis vel ovato-lanceolatis circa 10-12 cm. longis, 3.5-5.5 cm. latis
vel interdum bracteiformibus, oblongo-lanceolatis, circa 2 cm. latis,
basi late cuneatis, acutis, apicem versus paullo angustatis vulgo
subabrupte obtuseque caudato-acuminatis, acumine 5-7 mm. longo,
utrinque nitidulis, chartaceo-coriaceis fere opacis sed pellucido-
punctatis, haud lineatis, undique bene laxeque reticulatis, solum
subtus nervis 3 prominulis praecipue cpsta media, omnibus basa-
libus; pedunculis masc. circa 3 mm. longis; receptaculis cylindraceo-
globosis 4 mm. longis; pedicellis plerumque 6 mm. longis, sepalis
spathulato-obovatis 4 mm. longis, petalis paullo longioribus; antheris
non apiculatis filamentis duplo longioribus. — Peru: Mishuyacu near
Iquitos, Dept. Loreto, King 874 (type, Field Museum).

Dr. Gleason has referred this to S. cumanensis Willd., which he
perhaps interprets broadly enough to include it. On the specific
basis established by DeCandolle, Monogr. 1. 1878, it is at variance
to the type (Herb. Willd. 18396) in its much heavier, obtusely cau-
date-acuminate leaves, ellipsoid receptacles, nearly twice as large
flowers, and obtuse anthers. In DeCandolle's treatment it would
be sought, it seems to me, among species Nos. 57-64 or 136-141.
Among the first it is most comparable to S. floribunda Kunth and

PERUVIAN PLANTS 45

S. staminea Griseb. Its caudately acuminate, laxly reticulate leaves
and smaller flowers distinguish it from the former and its subopaque
leaves, longer petioles, more shortly vaginate, and broader perianth
segments from the latter. Among the second group (Nos. 136-141)
it is more at variance in its smaller flowers and different foliage.

In proposing a number of new species of Smilax it may not be
amiss to remark that they are probably new forms only in the sense
that their characters do not conform exactly to those of any species
described, at least so far as I have been able to determine. And,
more to the point, it has not seemed to me practical to interpret
described species liberally enough to include considerable variations,
for to do so would have the effect of breaking down the entire classi-
fication carefully worked out by DeCandolle. When Smilax is really
known, the number of species recognized today will be considerably
reduced, but for Peru, at least, there is yet a paucity of material,
so that the true limits of species can only be guessed. Especially
is the problem difficult because few species are known positively in
fruit and flower, collectors always securing one or the other in accord
with the particular season. Furthermore, few collections show en-
tirely mature leaves and old stems or branches. Some species, includ-
ing mine, may thus be juvenile states, or their apparent differences
may be sexual rather than specific.

Smilax colubrina, spec, nov., ubique tortuosa; ramis ramulisque
gracilibus teretibus haud lineatis et obscure vel vix nigrp-punctatis
exaculeatis ad 2 mm. crassis; petiolis flexuosis tarde cirrhiferis solum
ad basin inconspicue vaginatis, 1-1.5 cm. longis; foliis undique
nitidulis laxe reticulatis, nervis lateralibus 2-3 basalibus supra haud
subtus paullo prominulis solum costa media praecipue subtus valde
notata, subcoriaceo-opacis (non punctatis), late ovatis vel ovato-
lanceolatis, breviter acuminatis vel acutis plerumque circa 9 cm.
longis et 4 cm. latis, margine undulatis; pedunculis masc. 3-5 mm.
longis; pedicellis circa 5 mm. longis; floribus albis vel ut videtur
interdum flavis 2-2.5 mm. longis numerosis (circa 25), segmentis
anguste oblongis; antheris filamentis multo brevioribus exapiculatis. —
Peru: Mishuyacu near Iquitos, Dept. Loreto, Klug 1327 (type,
Field Museum); also, apparently, 886 and 867.

Referred with Smilax gilva to S. cumanensis Willd. by Dr. Gleason,
from both of which its completely opaque leaves appear to distin-
guish it. From the former it is further separable by its smaller and
different flowers. It seems to me to belong to species Nos. 97-118,
and to resemble most S. irrorata Mart., from which, according to
description, it differs in its elineate stems, shortly vaginate petioles,

46 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

and many flowers. In aspect it simulates S. staminea Griseb., which,
however, has larger flowers.

Smilax vaga, spec, noy., liana; ramis teretiusculis vel suban-
gulatis 2 mm. crassis, scabriusculis et ad nodos tuberculatis ubique
plus minusve aculeatis, aculeis breyibus fere minutis; ramulis an-
gulatis tortuosis fere laevigatis; petiolis 5-7 mm. longis, bene vagi-
natis et cirrhiferis; foliis oyatis basi rotundatis mucronulato-acutis
plerumque circa 7 cm. longis, 3.5 cm. latis, margine distincte crasso-
cartaligineis, membranaceis, densissime lucido-punctatis, ubique
mediocriter reticulatis, nervis 5 paullo notatis, pallido-viridibus vix
nitidulis; pedunculis fern. 3-4 mm. longis; receptaculis globosis yix

3 mm. crassis; pedicellis (ut videtur 12-15) circa 5 mm. longis;
fructibus circa 1 cm. crassis. — Peru: Rumizapa near Tarapoto,
Dept. San Martin, Williams 6757 (type, Field Museum).

Among Peruvian species to be distinguished particularly only
from S. cumanensis Willd., which is unarmed and smooth; but per-
haps it has wandered, in which case its character seems to ally it
to species 97-119, possibly to S. cissoides Griseb. or even to some
other species, since its flowers are unknown. Precisely, however,
it "fits" no description. If it is a new endemic, its prickles and well-
developed tendrils suggest that its name, considered locally, may
not be inappropriate.

Smilax Williams!, spec, nov., valde tprtuosa; ramis ramulisque
subangulatis vel teretiusculis laevibus gracilibus epunctatis et elinea-
tis paullo angulato-striatis vix 1.5 mm. crassis; petiolis 7-10 mm.
longis ad medium vaginatis plerumque valde cirrhiferis; foliis ellip-
ticis vel ovato-ellipticis basi apiceque rotundatis, vel apice fere
truncatis brevissime mucronulatis, plerumque circa 8 cm. longis,

4 cm. latis, subcoriaceis, fere qpacis, haud punctatis (vel obscure
pellucidis), undique dense prominulo-reticulatis et pernitidis, nervis
lateralibus vix notatis; pedunculis junioribus circa 3 mm. longis;
receptaculis globosis, 3 mm. crassis; floribus ignotis. — Peru: In
sandy soil, Tarapoto, Dept. San Martin, Williams 5432 (type, Field
Museum).

Notwithstanding the lack of flowers I venture to give this speci-
men a name because the foliage seems to be entirely distinctive.
No species approaches it more, apparently, than the ornate-lineate
S. irrorata Mart., also of Peru.

Smilax magnifolia, spec, noy., liana venusta; ramis vel ramulis
subteretibus leviter striatis laevigatis haud punctatis mediocriter
robustis circa 5 mm. crassis; petiolis valde curvatis circa 2.5 cm.
longis ad basin vaginatis et plus minusve breviter cirrhiferis; foliis
elongato-ovatis basi rotundatis, breviter acutis, apice ut videtur
breviter obtuseque acuminatis, circa 3 dm. longis ad 11.5-12.5 cm.

PERUVIAN PLANTS 47

latis, e basi ad apicem gradatim sed paullo angustatis, fere subcpri-
aceis subnitidulis opacis, nervis venisque solum subtus mediocriter
prominulis, nervo central! praecipue distinctp, supra valde impressis;
pedunculis valde compressis circa 1 cm. longis; receptaculis oyalibus
fere 7 mm. longis, circa 5 mm. latis; floribus ignotis; fructibus ut
videtur negris. — Peru: Mishuyacu near Iquitos, Dept. Loreto, Klug
1000 (type, Field Museum).

Apparently comparable only to S. phyllobola Griseb., and seem-
ingly distinct by the much larger leaves, longer and flat peduncles,
and oval receptacle.

Stnilax bella, spec, nov., liana verruculpsa; ramis (vel ramulis?)
leviter angulatis breviter parceque aculeatis et dense tuberculatis
circa 3.5 mm. crassis; petiolis circa 2 cm. longis breviter vaginatis;
fpliis fere rigido-coriaceis supra nitidis griseo-viridibus subtus pal-
lidioribus supra obscure laxeque reticulato-venosis, nervis lateralibus
haud prominulis vel impressis, oblpngp-elliptico-lanceolatis, basi rotun-
dato-acutis, apice late acuminatis, circa 2.5 dm. longis et 7.5-9 cm.
latis; pedunculis fern, vix 5 mm. longis; pedicellis circa 3.5 mm.
longis; floribus pernumerosis (circa 50-60), viridibus, vix 2 mm.
longis, segmentis lineari-oblongis; ovario subglobosp. — Peru: Mishu-
yacu near Iquitos, Dept. Loreto, Klug 37 ^ (type, Field Museum).

In foliage there is a striking similarity to the smooth-stemmed
S. syphilitica Willd., and no doubt the relationship lies in the vicinity
of that species; perhaps with S. insignis Kunth or S. japicanga
Griseb., both with much thinner leaves; or, more likely, with S.
aequatorialis (Griseb.) DC. or S. cinnamomea Desf., both with longer
peduncles or shorter petioles, but, more especially, with deeply
impressed lateral nerves above.

Eucharis Castelnaeana (Baill.), comb. nov. Calliphruria Cas-
telnaeana Baill. Bull. Mens. Soc. Linn. Paris 143: 1135. 1894.

This little-known plant must, if one may judge from description,
be a true Eucharis, notwithstanding its simulation of Calliphruria
in its small and narrow corolla, for the latter genus has free filaments
winged below and toothed at each side.

Eustephia armifera, spec, nov., conspicue foliosissima; bulbi
oyoidei 4 cm. diam. in collum fere 1.5 dm. longum attenuati; foliis
circa 8 pblongo-linearibus 2-3 dm. longis fere ubique 12-14 mm.
latis, laxis vel ut videtur interdum prostratis; pedunculis circa 1.5
dm. longis plus minusve compressis; bracteis 2-3, majoribus fere
4 cm. longis, circa 1 cm. latis, subtruncatis; floribus circa 5, subses-
silibus; perigonip 2.5-3 cm. longo, tubo vix nptato in faucem sensim
ampliato, laciniis definite inaequalibus ovatis subacutis 6-8 mm.
longis; filamentis liberis 5 mm. longis anguste alatis ad apicem undique
valde dentatis; antheris circa 7 mm. longis; stylo stamina paullo

48 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

superante sed incluso, stigmate irregulariter discoideo-folioso. — Peru:
Grassy places between shrub-wood, 3,100 m., Marcapata, Dept.
Cuzco, Weberbauer 7806 (type, Field Museum).

This interesting plant seems referable only to Eustephia, from
which it differs in its subsessile flowers and the position of the two
lateral filament teeth immediately beneath the anther. The stamen
therefore suggests in form the ancient spear or trident. The illusion
is perfect upon the fall of the anther. Thus an innocuous-appearing
plant is armed with concealed weapons! The collector noted the
flowers as "blood-red without, yellowish within."

Dioscorea fodinarum Kunth, Enum. PI. 5: 405. 1850. D.
venosa Uline ex Knuth, Notizbl. 7: 190. 1917. D. venosa Uline, var.
effusa (Griseb.) Uline ex Knuth, op. cit. 191. D. effusa Griseb. Kjoeb.
Vidensk. Meddel. 161. 1875.

Knuth regards these collections as representing one variable
species. He has, however, failed to accord with modern nomen-
clatorial practice of using the first published name as the species
name. If his interpretation of the plant is correct, as seems probable,
his two varieties must be referred to the name of Kunth rather than
to that of Uline.

Heliconia Standleyi, spec, nov., robusta circa 2 m. alta glabra;
petiolis circa 1.75 m. longis; foliis ad fere 3 m. longis, ad 6 dm.
latis, chartaceo-coriaceis, nervis lateralibus mediocriter prominen-
tibus, costa media 1 cm. crassa; inflorescentia pendula ad fere 2 m.
longa; rhachidibus valde flexuosis manifesto 4-angulatis ad 1 cm.
crassis; bracteis 2- fere 4 cm. remotis 9-10 cm. longis 3-3.5 cm.
latis, obtusis vel vix acutis, late cymbiformibus; pedicellis numerosis
circa 1 cm. longis; bracteolis pallidis anguste lanceolatis acuminatis
circa 6 cm. longis; floribus ignotis. — Peru: Iquitos, Dept. Loreto,
Killip & Smith 27444- (type, Field Museum). ,

A magnificent plant, nearest, I think, to H. rostrata R. & P.,
but widely at variance in the character of rachis and bracts. The
short and blunt bracts are markedly different from the tapering ones
of H. rostrata and H. Bihai L., to which species it is also allied.

Heliconia tenebrosa, spec, nov., ut videtur glabra 5 dm. alta;
petiolis manifesto nitidis 1.5-2 dm. longis, vix 2 mm. crassis; foliis
nitidulis ut videtur intense viridibus, subellipticis, basi breviter
acutis, apice longe acuminatis, circa 2.5 dm. longis, 7-8 cm. latis;
inflorescentia erecta vix 2 dm. longa; bracteis inferioribus 7-8 cm.
longis ad basin 1.5 cm. latis ad apicem valde angustatis, acutis, circa
1 cm. remotis; rhachidibus distincte curyato-flexuosis; floribus 3 cm.
longis leviter curvatis ad basin 4 mm. latis; staminibus liberis exsertis,
antheris 5 mm. longis. — Peru: Mishuyacu near Iquitos, Dept. Loreto,
King 1089 (type, Field Museum).

PERUVIAN PLANTS 49

Allied to H. cannoidea, but the petioles long and slender. The
relatively close inflorescence distinguishes it easily from H. Schumann-
iana and H. aureorosea.

Costus zingiberoides, spec, nov., ut videtur herba humilis;
caulibus gracilibus 5-6 mm. crassis; foliis breviter petiolatis lineari-
lanceplatis 1.&-2 dm. longis, 12-15 mm. latis, longe acuminatis,
glabris, vaginis sparse hirsutulis; spicis obovoideo-cylindraceis vel
demum oblongo-cylindraceis ad 10 cm. longis et 2.5 cm. crassis;
bracteis late oyatis yix acutis coriaceis striatis puberulentis vel
glabris infra apicem linea callosa ad 4 mm. longa munitis; corollae
tubo 7 mm. longo; bracteolis hyalino-membranaceis obscure parceque
puberulis truncatis 7 mm. longis; calyce obtuse trilobate circa 22
mm. longo; corolla intense flava 3 cm. longa; staminis nlamento
petaloideo elliptico apice integro; ovario triloculari. — Peru: Yuri-
maguas, Dept. Loreto, Williams 3985 (type, Field Museum); also
3924; Killip & Smith 28000.

A very strange plant with something of the aspect of Zingiber
in the narrow light-colored leaves and orange-reddish bracts; and
with partially the character of Renealmia, but the connective is
certainly petaloid and the inflorescence is typically that of Costus.
The ovary is definitely 3-celled. I have not succeeded in dissecting
well a perfect corolla.

Costus scaber R. & P. Fl. Peruv. 1: 2. pi. 3. 1798.

Klug 922, referred recently to C. villosissimus Jacq., is rather
this species.

Costus puchucupango, spec, nov., vix robustus ut videtur
mediocriter altus, caulibus circa 1-1.5 cm. crassis; vaginis sordide
fulvo-puberulis marginibus dense longe lanuginoso-ciliatis; foliis
sessilibus oblongo-obovato-subellipticis, circa 2-2.5 dm. longis, &-10
cm. latis, basi angustatis, apice caudato-acuminatis, supra glabris
praeter costam mediam utrinque marginesque prominulos fulvo-
pubescentes, subtus ubique minutissime parceque pallide puberulis;
inflorescentia turbinato-oyoidea; bracteis late ovatis apice rotundato-
acutis circa 1.5 cm. longis ut videtur rubricundis infra apicem pro-
minente calloso-lineatis. — Peru: Yurimaguas, Dept. Loreto, Williams
4570 (type, Field Museum).

The habit and foliage, even to color, of this plant are strikingly
like those of C. laevis R. & P. The latter, however, is glabrous, and so
marked are the two forms of pubescence on C. puchucupango that it
seems reasonable to think it is more than a pubescent variety. Too,
its inflorescence is very young, and more than likely there are dis-
tinctive floral characters. It bears some resemblance to C. Ulei
Loes., but that species has much narrower bracts, as also C. tarmicus

50 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

Loes., the latter with glabrous leaves. The native name may serve
also as the specific one.

Costus tarapotensis, spec, npv., C. amazonico ut videtur similis,
foliis membranaceis supra glabris subtus conspicue fulvo-villosis;
oblongo-ovato-lanceolatis, basi apiceque sensim angustatis, acu-
minatis, circa 3 dm. longis, 5-7 cm. latis; inflorescentia turbinato-
ovoidea glabra; bracteis circa 5 cm. longis et 7 mm. latis acutis ad
apicem submembranaceis; floribus ut videtur 2 cm. longis. — Peru:
Tarapoto, Dept. San Martin, Williams 6529 (type, Field Museum).

In spite of the imperfect material, I think the alliance of this
plant must be with C. amazonicus (Loes.) Macbr., which has much
broader leaves.

Dimerocostus Williams!, spec, nov., ut videtur affinis D.
Tessmannii sed foliis solum 10-12 cm. longis, circa 3.5 cm. latis,
bracteis definite dentatis et cum calloso obscure ornatis, sepalis 2 cm.
longis et seminibus brunneis 4 mm. longis; yaginis ad marginem
conspicue lanuginoso-ciliatis; petiolis breyissimis circa 2 mm. latis.—
Peru: Yurimaguas, Dept. Loreto, Williams 1+291 (type, Field
Museum).

The foliage of this plant is nearly that of the little-known D.
Guttierezii Ktze., but the pubescent sheaths, narrow petioles, and
brown instead of black seeds would seem to distinguish it. Except
in their much smaller size, the leaves resemble those of D. Tessmannii,
but as the inflorescence is in fruit (detached, however), it does not
seem likely that the leaves are not fully grown. Besides, there appear
to be other differences, as indicated. D. bolivianos (Rusby) Loes.
is apparently related, having also the bicarinate bracteoles of all
these species, but they are described as ciliate, which is not at all
the case for D. Williamsi. To be considered is D. elongatus Huber,
but it has deeply fissured bracteoles and puberulent leaves. It seems,
therefore, necessary to admit another species into this rapidly grow-
ing genus and it, the fifth for Peru, may fitly bear the name of the
energetic collector.

Renealmia lativagina, spec, nov., robusta circa 3 m. alta;
foliis longe petiolatis oblongo-ellipticis abrupte et breviter caudato-
acuminatis, basi sensim gradatim angustato-cuneatis, circa 8 dm.
longis fere 2 dm. latis, glabris vel fere glabris; pedunculis basalibus
dense cum vaginis glabris chartaceo-coriaceis imbricatis 1-1.5 dm.
longis circa 2.5 cm. latis, utrinque vestitis; racemis simplicibus 3
dm. longis; floribus solitariis; rhachidibus dense sed haud velutine
pubescentibus; bracteolis calycibusque leviter puberulis 3-dentatis
circa 2.5 cm. longis; pedicellis 7 mm. longis; corollis fere 3 cm.
longis, lobis et labello ut videtur subaequalibus. — Peru: Iquitos,
Dept. Loreto, Killip & Smith 27099 (type, Field Museum).

PERUVIAN PLANTS 51

Undoubtedly referable to the Racemosae, but not to any species
as yet described therein. So large are the sheaths, considered pro-
portionately to the rest of the plant and to those of the other species,
that they could not inaccurately be described as enormous. The
pubescence of the inflorescence is a dense but not velvety puberulence.
The bracts, except young ones about the buds, are too broken for
description; the buds are oblong-obovate, acute, 2.5 cm. long, and
1 cm. wide.

At the New York Botanical Garden this was considered the same
as R. thyrsoidea (R. & P.) P. & E., with spike dense even in fruit
and rarely 1.5 dm. long, the pedicels 2 mm. long, and the sheaths
soft, elliptic, and only 3-5 cm. long.

Calathea Loeseneri, spec, noy., ut videtur circa 3 dm. alta;
petiolis circa 13 cm. longis ad medium vaginatis cum yaginis mem-
branaceis minute parceque pubescentibus; foliis yalde inaequilateris
ovato-ellipticis basi acutis apice anguste acuminatis fere 1.5 dm.
longis, 5.5 cm. latis, supra glabris, subtus minutissime strigillosis
praecipue ad nervos; pedunculis circa 3 dm. longis minute adpresseque
pilosis vel superne plus minusve longe villosis; spicis subturbinatis
yel anguste ovoideis circa 7 cm. longis et 2-2.5 cm. latis; bracteis
inferioribus subcoriaceis solum puberulis vel interdum etiam praeci-
pue ad basin villosis, 2-2.5 cm. longis, apice rotundatis, superioribus
(sterilibus) erectis anguste ovato-lanceolatis acutis vel acuminatis
4-5 cm. longis; bracteolis indurato-clavaliculatis; floribus ignotis;
capsula glabra circa 7 mm. longa. — Peru: Marsh, Mishuyacu, near
Iquitos, Dept. Loreto, King 940 (type, Field Museum).

No other species referred to the Comosae with indurate bracteoles
has the small, sharply acuminate leaves of this plant. Nevertheless
it has recently been determined as C. peruviana Koern., a species
with much larger leaves, little oblique at the rounded-apiculate
tip, and with soft bracts, rusty-pilose throughout. Perhaps the
species is highly variable in these respects, but I can not believe it.
The collector noted the flowers as lilac-color.

Calathea peruviana Koern. Bull. Soc. Nat. Mosc. 351: 128.

1862.

Besides the Pavon specimen from Chicoplaya, I have seen Weber-
bauer 1817 from La Merced (det. Loesener) and Killip & Smith
23664 from Rio Pinedo, north of La Merced. The last was deter-
mined as C. velutina (P. & E.) Koern., a species belonging in another
group because of its thin bracteoles.

That the woods of Peru are filled with unclassified plants may
be accepted as a truism after a glance at nearly any revision or

52 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

current taxonomic paper. Nevertheless, the numerous species I am
proposing here, particularly in Calathea and Ischnosiphon, are, I am
certain, mostly not new in the sense of new species acceptable by
me iii the flora of Idaho, for instance, although some of them are
undescribed entities positively distinct from any known form. All
of them, however, key out from all other Peruvian species, at least
as interpreted by me, and accordingly I propose them as new because
in working on a partially known flora it is the most practical device
in building up its classification. To treat these different plants as
varieties or to include them in previously described species by liberal-
ization of original descriptions only, in most instances would make
more difficult the monographer's work in unraveling the true rela-
tionships and determining the real specific lines.

Calathea silvosa, spec, nov., planta gracilis ad 6 dm. alta; foliis
subtus, petiolis pedunculis inflprescentiisque conspicue sed vix dense
cum pilis brunneis Ipngissimis patentibus pubescentibus; petiolis
solum inferne yaginatis circa 6 dm. longis; foliis supra glabris nitidis
ut videtur utrinque viridibus, fere aequilateris et ellipticis, ad basin
paullo angustatis, subacutis, apice subrotundatis et abrupte apicu-
lato-acuminatis, circa 3 dm. longis, 1.5 dm. latis; pedunculis ut
videtur elongatis; spicis subglobosis circa 6 cm. crassis; bracteis
membranaceis valde depresso-patentibus ovatp-lanceolatis, exteri-
oribus ad basin coriaceis, glabris, acuminatis, circa 5 cm. longis, 12
mm. latis, interioribus vacuis late ovatis acutis circa 2-2.5 cm. longis,
1.5-1.7 cm. latis; sepalis glabris 17 mm. longis; corollae tubo 2.8 cm.
longo, lobis oblongo-lanceolatis subacutis 10 mm. longis; staminodio
obovato, circa 11 mm. longo, calloso paullo breviore ceteris simili;
ovario glabro. — Peru: Woods, Mishuyacu, near Iquitos, Dept.
Loreto, Klug 11 (type, Field Museum).

A member of the series Comosae, but very distinct in pubescence.
The long brown hairs, though uniformly distributed and conspicuous,
do not conceal the surfaces. The flowers were noted by the collector
as green and brown. The widely spreading, not at all imbricated,
and subequal bracts give the inflorescence a striking appearance.

Calathea ulotricha, spec, nov., caulibus petiolis pedunculisque
ignptis; foliis mediocriter inaequilateris late ellipticis ad basin rotun-
datis basi ipse acutis ad apicem abrupte longiacuminatis vel fere
caudatp-acuminatis 4 dm. longis et fere 2 dm. latis supra glabris
nitidulis subtus minute denseque adpresse pilosis ubique viridibus;
spicis cylindraceis 8 cm. longis 3 cm. latis vel cum floribus 5 cm.
latis pallidis et dense cum pilis ochroleucis longissimis (ad 2 mm.) et
brevissimis intermixtis molliter pubescentibus (sterilibus minus dense
indutis); bracteis spiraliter dispositis, inferioribus subrhpmboidep-
ovatis rotundo-truncatis 2 cm. longis 1 cm. latis, superioribus steri-

PERUVIAN PLANTS 53

libus ovato-lanceolatis pbtusis sed ad apicem aliquid angustatis circa
5 mm. latis; bracteolis indurato-clavaliculatis fere 2 cm. longis;
sepalis 10 mm. longis; floribus 2.5 cm. longis conspicue exsertis;
corollae tubo glabro 18 mm. longo, lobis lanceolatis acuminatis 7
mm. longis; staminodium exterius 9 mm. longum obovatum obscure
emarginatum. — Peru: Santa Rosa, Dept. Loreto, Lower Rio Hua-
llaga, Williams 4921 (type, Field Museum).

In all probability a handsome plant, tall, with large leaves, and
entirely at variance to all other species of the series Comosae. In
its rather narrow, densely pale-pubescent inflorescence, indurate
bracteoles, and acuminate leaves it possesses an unusual combination
of characters.

Calathea jocosa, spec, nov., caulibus petiolis pedunculisque
ignotis; foliis ubique viridibus supra glabris nitidulis subtus tenuis-
sime denseque puberulis inaequilateris late yel rotundo-ellipticis
basi rotundato-acutis apice rotundatis et brevissime acutis, 23 cm.
longis, 15 cm. latis; spicis globosis circa 7 mm. crassis; bracteis glabris
laxe imbricatis et valde divaricatis, exterioribus fere 5 cm. longis
solum ad basin subcoriaceis oblongo-obovatis breyiter acutis, steri-
libus numerosis patentibus similibus paullo brevioribus; bracteolis
membranaceis; floribus ignotis; oyario glabro. — Peru: Recreo, near
Yurimaguas, Dept. Loreto, Williams 3968 (type, Field Museum).

Meager as the material is, it seems that the plant must belong
to the Comosae, although its glabrous inflorescence and subrotund
leaves, pubescent beneath, comprise a character not found in any
other species in that group. Possibly it is a member of the Nudi-
scapae, but there, even on its few known characters, it fits the descrip-
tion of no other species. Among Peruvian Calatheas it resembles
most C. silvosa.

Calathea enclitica, spec, nov., gracilis vix alta glabra; foliis
fere aequilateris elliptico-ovatis basi subacutis vel subrotundatis
apice obtusis vel subacutis circa 1.5 dm. longis, 6.5-7 cm. latis, ut
videtur pallide viridibus et concoloribus; pedunculis ut videtur 1.5
dm. longis; spicis turbinato-globosis cum bracteis 5-6 cm. longis;
bracteis exterioribus coriaceo-chartaceis ovato-lanceolatis acutis vel
acuminatis ad 4 cm. longis, interioribus valde patentibus membran-
aceis acutis circa 3 cm. longis, 10-12 cm. latis; bracteolis membran-
aceis; sepalis 12 mm. longis; corollae tubo ignoto, lobis circa 10 mm.
longis; staminodium exterius obovatum 17 mm. longum, callosum
ut videtur circa 12 mm. longum; ovario glabro. — Peru: Pebas, on
the Amazon River, Dept. Loreto, Williams 1955 (type, Field
Museum) .

More than likely this is a close ally of C. colorata (Hook.) Benth.,
notwithstanding the abundant technical differences. Its divaricate,
not at all imbricate bracts seem to be a very similar development.

54 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

Its small leaves are noteworthy. It is unfortunate that only a portion
of the petiole and peduncle is known.

Calathea Williams!, spec, nov., ut videtur gracilis et humilis;
petiolis longissime vaginatis 5-6 cm. longis; yaginis paullo angustatis
fere 4 mm. latis dense strigosis; foliis plus minusve inaequilateris fere
ellipticis basi subrotundatis apice definite acuminatis circa 1.5 dm.
longis, 6-7 cm. latis, supra ut videtur viridibus sed molliter cum pilis
rufis subpatentibus pubescentibus, subtus atropurpureis (vel rubris?)
etiam dense subpatente albo-pilosis; pedunculis strigosis, recurvatis
e centre foliorum, 3 cm. longis; spicis anguste cylindraceis 5 cm.
longis, circa 1 cm. latis, rufo-villosis; bracteis distichis ad apicem
fertilibus obovatis circa 1.5 cm. longis, 8 mm. latis vel infimis longior-
ibus; bracteolis indurato-claviculatis; sepalis glabris, 8 mm. longis;
corollae tubo glabro, 12 mm. longo, staminodium exterius circa 6
mm. longum; ovario glabro. — Peru: La Victoria on the Amazon
River, Dept. Loreto, Williams 2541 (type, Field Museum).

At first glance this species suggests C. Legrelleana (Lind.) Reg.,
but its spikes are evidently floriferous to the apex. I think, never-
theless, that its relationship is with that species, which probably
belongs in the Comosae. The character of spiraled or distichous
bracts is not always well marked.

Calathea Standleyi, spec, nov., ut videtur altissima et robusta;
petiolis elongatis ad 10 dm. longis dense rufo-villoso-hirsutis exceptis
2 cm. ad apicem; foliis late ellipticis circa 4 dm. longis, 2 dm. latis,
acuminatis ad basin rotundatis et breviter decurrentibus glabris
utrinque metalico-caeruleis-coloratis; pedunculis circa 1.5 dm. longis
ad medium vaginantibus; vaginis hirsuto-villpsis fere 1 dm. longis
et 1 dm. anguste productis; spicis ovoideis, circa 1 dm. longis, 5-6
cm. crassis, glabris ut videtur metalico-caeruleis; bracteis exterioribus
late ovatis 4.S-5 cm. longis, 3-3.5 cm. latis vix acutis, interioribus
fere acuminatis; bracteolis membranaceis; floribus paullo exsertis
pallide flavis circa 4 cm. longis, glabris; calyce glabro 17-20 mm.
longo; corollae tubo circa 3 cm. longo; staminodiis ut videtur subae-
qualibus circa 5 mm. latis, 8 mm. longis. — Peru: Puerto Yessup,
Dept. Junin, Killip & Smith 26268 (type, Field Museum).

This beautiful Calathea is surely comparable to C. pachystachya
(P. & E.) Koern., of the Scapifoliae, a glabrous plant with elongate
spikes but otherwise not, apparently, very diverse. In the single
inflorescence I have not succeeded in removing a perfect flower.
C. contamanensis Huber, though placed by the author in the Nudi-
scapae, seems from description to resemble our plant. Its leaves,
however, are described as dark violet beneath and its bracts as only
1.5 cm. long.

Calathea mishuyacu, spec, nov., C. exscapae peraffinis; ubique
fere glabra; petiolis longissime vaginatis saepe fere ad articulum,

PERUVIAN PLANTS 55

vaginis membranaceis inferne 5 mm. latis; foliis baud manifesto
inaequilateris obovato-ellipticis ad basin gradatim attenuatis, apice
subabrupte et breviter acuminatis, ad 10.5 cm. latis, 1.5 dm. longis,
ut videtur obscure zonatis; floribus fere ignotis, pallide flavis, ut
videtur vix exsertis; bracteis coriaceis; sepalis circa 12 mm. longis;
ovario apice paullo sericeo. — Peru: Mishuyacu, near Iquitos, Dept.
Loreto, Klug 1^.16 (type, Field Museum).

Probably this species is much more distinct from C. exscapa (P.
& E.) Koern. than I have realized. The material is not in good
condition, the one flower remaining being broken. Williams 7868
from Yurimaguas I have referred to C. exscapa, and it seems quite
different in foliage and in texture of bracts, the latter being much
softer than in our plant. The petioles of C. exscapa are figured by
Poeppig and Endlicher a& vaginate only at base. Unfortunately
the material seen by me shows only the upper part of the petioles.
A plant may bear a native name without modification. So, the
Indian word "mishuyacu" may serve as the scientific name of this
species.

Calathea Klugii, spec, nov., glabra humilis vix ultra 2 dm. alta,
vulgo humilior radicibus rhizomatosis; petiolis longissime vaginatis,
vaginis sensim attenuatis, 4-10 cm. longis; foliis anguste ovato- vel
oblongo-lanceolatis aequilateris basi rotundatis apicem versus sensim
acute acuminatis ad 12 cm. longis et 2 cm. latis interdum brevioribus
et 1.5 cm. latis, utrinque yiridibus vel ut videtur subtus plus minusye
purpureis; spicis subturbinatis cum floribus circa 2 cm. longis et vix
1 cm. latis; bracteis paucis oblongp-ovatis subacutis glabris vel minu-
tissime parceque puberulis, exterioribus circa 7 mm. longis, 4 mm.
latis; sepalis glabris fere 9 mm. longis; floribus glabris albis et viola-
ceis; corollae tubo 10 mm. longo, lobis lanceolatis vix 4 mm. longis,
staminodio paullo longiore. — Peru: Mishuyacu, near Iquitos, Dept.
Loreto, Klug 532 (type, Field Museum); also 559.

Certainly a member of the series Rhizanthae, but differing from
all species with soft bracteoles in its narrow leaves, glabrous (or
nearly) bracts, and small flowers. The peduncles may arise directly
from the leaf base, apparently, and such individuals suggest the
subgenus Microcephalum, which contains no species with so narrow
leaves at the same time acuminate and equilateral.

Ischnosiphon Killipii, spec, nov., suffrutescens, scandens;
caulibus ut videtur ramosissimis e nodis incrassatis; vaginis glabris
circa 6 cm. longis; petiolis ad 7-7.5 cm. longis, articulo 3-6 mm.
longo supra paullo hirsute; foliis fere ellipticis basi rotundatis apice
valde inaequilateri-acuminatis viridibus haud pallidioribus subtus
glabris vel costa media et apice strigosa, plerumque ad 12 cm. longis
et 5 cm. latis; spicis 12 cm. longis, 6 mm. crassis; bracteis glabris

56 FIELD MUSEUM OF NATURAL HISTORY— BOTANY, VOL. XI

circa 3.5 cm. longis; bracteolis induratis fere 3.5 cm. longis; floribus
intense flavis geminatis; sepalis paullo puberulis 2 cm. longis; corollae
tubo fere 4 cm. longo glabro, lobis anguste lanceolatis 12 mm. longis
breviter pilosis, staminodio circa 1 cm. longo; ovario parce piloso.—
Peru: Iquitos, Dept. Loreto, Killip & Smith 26929 (type, Field
Museum).

Apparently in habit allied to /. gracilis (Rudge) Koern., with
different foliage (in color, shape, and size) and narrower spikes. In
other respects it seems much nearer I. surinamensis (Miq.) Koern. and
its apparent relatives, but its crowded leaves with exceptionally
1-sided tips afford a character, if important, as accepted by Schu-
mann, that forbids its reference to any of these species.

Notwithstanding these affinities the same collection has recently
been determined as /. bambiisaceus (P. & E.) Koern., a plant with
narrowly ovate-lanceolate (2-2.5 cm. wide), long-acuminate leaves,
the tips not at all oblique. Of course, there is the question of specific
values, but with only the two sheets before me and no intermediates,
I can not consider them even varietally related.

Ischnosiphon verruculosus, spec, nov., scandens ad nodos
tumescens; caulibus vaginisque verruculoso-tuberculatis; petiplis
ad articumm vaginatis (articulus 1-2 cm. longus) scabris, vaginis
valde prpductis; foliis utrinque viridibus glabris (vel apice puberulis)
ovato-ellipticis basi rotundatis ad apicem definite inaequilateralibus
acuminatis circa 1 dm. latis et fere 2 dm. longis; racemis solitariis
vel geminatis ad 3 dm. longis circa 7 mm. crassis; bracteis minutis-
sime puberulentis plerumque 3-3.5 cm. longis; bracteolis apice
induratis 4 cm. longis conspicue exsertis; floribus flavis, rubris et
viplaceis circa 2.5 cm. exsertis; sepalis leviter pilosis lineari-acu-
minatis 3 cm. longis; ovario piloso; corollae tubp 2 cm. longo parce
puberulo, lobis anguste lanceolatis distincte pilosis; staminodio circa
7 mm. longo. — Peru: Mishuyacu near Iquitos, Dept. Loreto, Klug
430 (type, Field Museum).

The exceedingly rough upper internodes and sheaths recall 7.
gracilis var. scabra Peters, but this is a much coarser plant with
larger and green leaves. It differs also from /. gracilis (Rudge)
Koern. in its pilose ovary and small staminodium. Apparently the
rough stems are distinctive in comparison with all species except
/. gracilis; nevertheless its true relationship is with I. aruma (Aubl.)
Koern., a large-leaved plant, with several racemes of larger flowers,
to which species it has recently been referred but in which it could
most doubtfully be included as a variety.

Ischnosiphon obliquiformis Loes. Notizbl. 6: 272. 1915.
This species, to which I had referred Klug 1006, may be considered
by some as identical with I. obliqua (Rudge) Koern., since I find

PERUVIAN PLANTS 57

the collection has recently been determined as the latter. However,
Professor Loesener's judgment, in the absence of intermediate
material, may be followed, and the material mentioned properly
represents his species.

Ischnosiphon ornatus, spec, nov., I. sphenophyllo similis;
vaginis pedunculis bracteisque fulvo-pilosis; petiolis ad 1.5 dm.
longis (articulus 1.2 cm. longus) infra medium vaginatis supra gla-
bratis vel puberulis; foliis aequilateris (acumine haud excentrico)
oblongo-ellipticis basi cuneatis circa 6 cm. latis, 1.5-2 dm. longis,
praeter costam mediam apicemque glabris, subtus pruinosis; spicis
ad 17 cm. longis circa 6 mm. crassis; bracteis 2-2.3 cm. longis;
bracteolis induratis solum 17 mm. longis; sepalis linearibus pilosis
15 mm. longis; corollae tubo puberulo 18 mm. longo, lobis lineari-
lanceolatis 6 mm. longis; stfoninodio circa 5 mm. longo; ovario apice
sericeo. — Peru: Yurimaguas, Dept. Loreto, Williams 3870 (type,
Field Museum).

Very possibly only a variety of /. sphenophyllus Schum., but
spikes almost shaggy-pubescent and sheaths far from "glabrous."
Besides, the bracteoles are much shorter, the corolla tube minutely
pubescent, and the corolla lobes shorter and linear. I. hirsutus
Peters, which I have not seen, seems to differ in its villous petioles
and much larger leaves. Its flowers are unknown.

Ischnosiphon neotericus, spec, nov., glaberrimus ut videtur
altus, erectus; petiolis solum inferne vaginatis ad 2.5 dm. longis,
articulo 2 cm. longo; foliis membranaceis utrinque viridibus ellipticis
aequilateris basi rotundatis apice fere abrupte caudato-acuminatis,
acumine 2 cm. longo tenuissime producto; spicis solitariis circa 12
cm. longis 4 mm. crassis; bracteis valde convolutis 3 cm. longis glabris
apice minute dentatis; bracteolis clavato-induratis 3.3 cm. longis;
sepalis linearibus minutissime puberulis 18 mm. longis; floribus
geminatis ut videtur roseis; corollae tubo 3.5 cm. longo glabro, lobis
glabris lineari-lanceolatis 9 mm. longis; staminodio crispe crenulato
11 mm. longo; ovario glabro. — Peru: Iquitos, Dept. Loreto, Williams
3761 (type, Field Museum).

Perhaps as near /. surinamensis (Miq.) Koern. as any species,
but at once distinct by its paired flowers and long petioles. The
plant seems to me to be entirely at variance to any described form.

Ischnosiphon wyomingensis, spec, nov., scandens, praeter
inflorescentiam haud pubescens; vaginis 4-5 cm. longis; petiolis
1-1.5 cm. longis; foliis oblongo-ellipticis vel fere ellipticis basi
rotundatis apice sensim mediocriter oblique acuminatis, acumine
ipso subcaudato, 5-6 cm. lath et circa 1.5 dm. longis, supra fusco-
olivaceis, subtus paullo pallidioribus; racemis ut videtur solitariis
circa 2 dm. longis, 7 mm. crassis haud flexuosis; bracteis circa 3.5
cm. longis minute adpresseque sed definite puberulis; bracteolis

58 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

filiformibus apice induratis paullo vel vix exsertis; sepalis 3 cm.
longis manifeste adpresse pilosis; corollae tubo conspicue yilloso
circa 3 cm. longo, lobis anguste lanceolatis 1 cm. longis; staminodio
oblongo-lanceolato. — Peru: Mishuyacu, Dept. Loreto, Klug 854
(type, Field Museum).

This could very easily be included in /. surumuensis Loes., but
the bracts are closely puberulent, the corolla tube conspicuously
pubescent, and the staminodium much narrower. I. Uleanus Loes.
is similar, but it has larger leaves and slender flexuous spikes. The
flowers were described by the careful collector as yellow and brown.
The colors of Wyoming appear to have been planted in Peru. It
gives me pleasure to give them botanical recognition there!

Monotagma anathronum, spec, nov., ut videtur peraffinis
M. doloso sed petiolis longissime vaginatis (ad vel fere ad articulum),
foliis 8-12 cm. latis et 2 dm. longis et paniculis haud spiciformibus;
bracteis 2.5- fere 3 cm. longis superne pilosis; sepalis 1 cm. longis;
corollae tubo glabro fere 2.5 cm. longo, lobis ellipticis circa 3 mm.
longis; ovario glabro. — Peru: San Antonio, Rio Itaya, Dept. Loreto,
Killip & Smith 29363 (type, Field Museum). Alto Rio Itaya,
Williams 3251.

Allied also to M. spicatum (Aubl.) Macbr. in the complete lack
of annulus at the base of the callus, but the inflorescence accompanied
by a leaf. It is therefore comparable also to M. Ulei Schum., to
which Gleason refers it, but that species has densely villous spikes
only a few centimeters long and a pubescent corolla tube only 1 cm.
long.

Monotagma contrariosum, spec, nov., herba ut videtur humi-
lis; petiolis late et fere ad articulum vaginatis circa 2 dm. longis,
articulo fere 2 cm. longo supra breyissime piloso, annulo (a petiolo)
prominente denseque hispido; foliis conspicue inaequilateris basi
rotundatis, basi ipsa acuta breviter acuminatis, acumine valde excen-
trico, fere 2 dm. longis, 1 dm. latis, subtus cum vaginis pedunculis
bracteisque pilosis, supra praeter costam glabris, ut videtur zonatis;
paniculis angustis circa 12 cm. longis, rhachidibus conspicuis; bracteis
circa 2 cm. longis; sepalis 1 cm. longis, hyalinis, glabris; corollae
lobis valde cucullatis 5 mm. longis; ovario apice villoso. — Peru:
Mishuyacu near Iquitos, Dept. Loreto, Klug 116 (type, Field
Museum).

This species would key only to M. plurispicatum (Koern.) Schum.,
because of the conspicuously annulate-barbed calloused petioles,
but in its few-flowered pilose inflorescence it is markedly contrarious.

Mpnotagma dolosum, spec, nov., gracilis; caulibus superne
petiolis foliisque subtus minute pilosis; petiolis circa medio vaginatis
ad 18 cm. longis infra laminam crasse (5 mm.) callosis; foliis haud

PERUVIAN PLANTS 59

vel vix inaequilateris ovato-ellipticis ad basin acutis ad apicem acu-
minatis, 6-7 cm. latis, plerumque 15-18 cm. longis, supra praeter
costam dense et longe pilosam glabris; paniculis spiciformibus ut
videtur axillaribus breviter pedunculatis paucis circa 1 dm. longis
fere glabris; rhachidibus internodiis cpnspicuis; bracteis yalde inpyatis
circa 17 mm. longis; floribus ut videtur albis solitariis sessilibus;
sepalis 10 mm. longis; corollae tubo circa 12 mm. longo, glabro,
lobis oblongo-ellipticis circa 2.5 mm. longis; staminodio (?) fere 5
mm. lato profunde emarginato. — Peru: Near Yurimaguas, Dept.
Loreto, Williams 5025 (type, Field Museum); also 5115.

Simulating M. parvulum Loes., but in reality nearest, I think,
to M. anathronum, from which it is distinguished chiefly by its nearly
evaginate petioles and spikelike inflorescence.

Monotagma laxum (P. & E.) Schum., var. oblongifolium,
var. nov., foliis oblongis 5.5 cm. latis, ad 3 dm. longis, ad apicem
paullp obliquis; vaginis (semper?) ad basin dense longo-pilosis;
sepalis circa 17 mm. longis. — Peru: Between Rio Nanay and Rio
Napo, Dept. Loreto, Williams 721 (type, Field Museum).

A plant apparently worthy a name, but there are connecting states.

Myrosma stromanthoides, spec, nov., caulibus petiolisque
ignotis; foliis glabris late ellipticis basi cuneato-rotundatis apice
breviter acuminatis subtus pallide viridibus supra ut videtur plus
minusye purpureo-zonatis circa 7 dm. longis, 1.5-1.8 dm. latis;
racemis plerumque 2-3; pedunculis laxe villpsis, 1.5-5 cm. longis,
longe bracteatis; bracteis inflorescentiarum circa 8-12 vix laxe im-
bricatis scaripso-chartaceis circa 1.5 cm. longis subrotundatis persis-
tentibus; floribus 7-8 geminatis, pedicellis brevibus vel longioribus
vix 2 mm. longis; sepalis 7 mm. longis; ovario glabro; capsula laevi
5 mm. Ipnga. — Peru: Tarapoto, Dept. San Martin, Williams 6563
(type, Field Museum); also 6017. "Bijahuillo."

Pubescent Colombian forms of Myrosma lutea (Jacq.), comb,
nov. (Maranta lutea Jacq.) are, from herbarium material, scarcely
distinguishable in aspect from our plant. Typically, however, M.
lutea is glabrous and the bracts finally caducous, and always, appar-
ently, the flowers are two or few, the terminal with a pedicel 3 mm.
long. It is unfortunate that the attachment of the leaves of our plant
is unknown.

This plant calls to attention the similarity of Stromanthe Loud.
Hamb. Gartenz. 5: 225. 1899, and Myrosma L.f. Schumann relied
upon the diverse placement of the leaves (in the former antitropous,
in the latter homotropous) to separate them. If this is properly a
generic character, Maranta Ruiziana must be removed from Ma-
ranta, for its leaves are antitropous. If the character were supported
by others it might seem to have significance, but in this case it is

60 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

not. On other characters there is no case for Stromanthe. The sta-
minodium is equally developed and the bracts may persist in fruit
in S. lutea, as evidenced above. Whether Myrosma should also
include Saranthe Eichl. and Ctenanthe Eichl. is open to question,
but probably at least the latter should be reduced.

For that matter, all generic lines in the Marantaceae, as drawn
by Schumann, are certainly open to question. There is no evidence
that many characters used by him to separate the groups result in
natural divisions or in divisions that, practically in some instances
at least, were not better regarded as sections of a well-defined genus.
In the Maranteae, for example, there is a general segregation that
is in logical accord with the generic acceptances in the Phrynieae,
but why should one regard characters as of generic value in one
instance and not in another, especially when they result in the
disassociation of plants of apparent group affinity? That is, if the
arrangement of the bracts is of importance as a generic character,
why is it given prime recognition in one place and none at all in
another? If the number of flowers is significant here, why not
generically elsewhere? Or again, if the duration of the bracts is
fundamentally important sometimes, why not always? Or the
degree development (to absence) of staminodia? Obviously one
should be able to answer these queries by pointing out that the
characters mentioned in each instance are supported by others that,
taken together, indicate natural groupings. Unfortunately such is
not the true state of affairs. Even a casual consideration of the plants
themselves suggests that in this family, as in many others, the lines
of development have been so close that the only practical and reason-
able treatment will be based not primarily on logic and arithmetic
(as Schumann's so obviously is), but on all characters as found plus
common sense. The result will be a classification less erudite in
appearance, no doubt, but basically sound in so far as variable and
perplexing nature permits. Discarding the commutation method,
then, the Maranteae could conveniently include probably six instead
of ten genera, namely, Maranta, Myrosma, Ischnosiphon, Monotagma,
Thalia (of course excluding the section Ammothalia illogically main-
tained here by Schumann in defiance of his own generic key), and
perhaps Monophyllanthe. A similarly elucidating realignment could
readily be made for the Phrynieae.

3. OTHER PLANTS, CHIEFLY MORACEAE

Killip & Smith 27932 has been referred by Professor Mildbraed
to Pourouma sapida Karst. but by the collectors to P. cecropiaefolia

PERUVIAN PLANTS 61

Mart. I think this is one instance of misdetermination more apparent
than real, for the validity of the Karsten species may well be
questioned.

Cecropia Standleyi, spec, npv., arbor 3-6 m. alta; ramulis
petiolisque plus minusve hispidulis haud lanuginosis; petiolis 1-2
dm. longis sulcato-striatis, basi intense rufo-tomentoso-callosis
interdum etiam hispidulis; foliis subrotundatis ut videtur plerumque
circa 3.5 dm. latis, supra viridibus ubique cum pilis nitidis submollibus
haud asperis mediocriter pubescentibus, chartaceo-membranaceis
nunquam bulboso-rugosis, subtus in areolis griseo- sed haud lanugino-
tomentosis, costis venulisque plus minusve hirsutulis; foliis ad paullo
infra medium 6-7-lobis, lobis sinu rotundato sejunctis late obovatis
breviter acuminatis, superioribus multo majoribus, lobo medio 2-2.5
dm. longo, 10-12 cm. lato, neryis (in lobo medio) utrinque 12-15,
1-1.5 cm. distantibus; pedunculis breviter denseque hirsutulis circa
1-1.5 dm. longis; amentis masc. circa 8-10, 1 cm. longe pedicellatis,
9 cm. longis, circa 3 mm. latis, filamentis aequalibus, antheris non
caudatis; amentis fern. 4 subsessilibus circa 18 cm. longis, 1 cm.
crassis, perigpnio leviter tomentoso fere 2 mm. longo, stigmatibus
minute penicillatis. — Peru: Mishuyacu, near Iquitos, Dept. Loreto,
King 399 (type, Field Museum); also Killip & Smith 27033 (fern.).

Among the twenty-five species of Cecropia known to grow in
Peru are several that resemble this one in a general way, but none
precisely, and in distinguishing Cecropias precision is necessary.
Nearest among these several appears C. ficifolia Warb., with obtuse
or barely acute leaf lobes, asperous above and white-tomentose
beneath. Perhaps as nearly allied is C. Klotzschyana Miq., with
distinctly rugose-bullate leaves, white beneath in the areoles with
a felt-like pubescence and with fewer aments. Then there is C.
bicolor Kl., but its leaf lobes are oblong; and the obscure C. dentata
Tre"c., its leaves described as 9-lobed, the lobes "dentate-subserrate,"
and white-tomentose beneath. Among extra-Peruvian species, C.
palmata Willd. and C. adenopus Mart, only need be mentioned. The
former differs in its tomentose pubescence and the latter in its thick
aments. In the case of all these species I have indicated only the
most obvious differences.

Pseudolmedia Mildbraedii, spec, npv., arbor 15 m. alta; ramis
ramulisque brevibus griseo-brunneis glabris vel minutissime sparseque
puberulentis, juyenilibus plus minusve angulatis; petiolis crassis 3
mm. longis; foliis numerosis elliptico-ovatis plerumque 8 vel 9 cm.
longis et 4 vel 4.5 cm. latis, basi haud oblique sed breviter cuneato-
acutis, apice plus minusve obscure obtuseque caudato-acuminatis vel
interdum apicem versus solum angustatis et obtusis, glabris, coriaceis,
supra ut videtur intense viridibus, tenuissime denseque reticulate-

62 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

venosis, nervis lateralibus circa 15 mediocriter undique prominulis
brunneo-flavidis, venis subtus obscure reticulatis; receptaculis masc.
discoideis 6 mm. latis; bracteis valde imbricatis circa 3-seriatis, sparse
puberulis, marginibus tenuioribus sed vix yel paullo ciliatis, exterior-
ibus late rotundatis interdum fere latioribus quam longis, circa 2
mm. latis, interioribus similibus sed 4 mm. latis et longis vel margin-
alibus (2-4) lanceolatis et 5 mm. longis; bracteolis lineari-spathulatis
apicem versus tenuiter ciliatis circa 4 mm. longis; filamentis vix
0.5 mm. longis; antheris 1.5 vel 2 mm. longjs, connectivo producto
apiculatis et minutissime ciliatis. — Peru: Mishuyacu, near Iquitos,
Dept. Loreto, King 633 (type, Field Museum).

P. laevigata Tre"c. has, according to description, thinner and
smaller, oblong-lanceolate leaves, somewhat oblique at base, the
15-18 lateral nerves rather obscure, and scarious-ciliate-fimbriate-
margined receptacle bracts. Otherwise P. Mildbraedii appears to
be similar. The collector noted the flowers as "light yellow." They,
i.e. the receptacles, are star-like, the narrow and elongate innermost
bracts extending as points.

Professor Mildbraed should have many plants named for him in
recognition of his friendly helpfulness in the furtherance of their
study.

Pseudolmedia scabra, spec, nov., arbor 4 m. alta; ramulis
foliisque undique dense scabro-puberulis; petiolis 5-7 mm. longis;
foliis oblongo-ellipticis basi acutis paullo vel haud inaequilateris
apice fere obtusis vel obtusis et apiculatis, plerumque 8 cm. longis
et 3 cm. latis vel demum 11 cm. longis et 4.5 cm. latis, rigidiuscule
coriaceis, supra viridibus, subtus pallidioribus; nervis lateralibus
circa 9 cum veniis reticulatis, supra vix prominulis, subtus conspicuis;
receptaculis masc. 2-3 in axillis foliorum fasciculatis, pedunculis 5^-7
mm. longis tomentuloso-puberulis, globosis circa 5 mm. crassis;
bracteolis spathulatis paullo apice incrassatis circa 1 mm. longis;
filamentis tenuibus 1.5 mm. longis; antheris fere 1 mm. longis.—
Peru: Mishuyacu, near Iquitos, Dept. Loreto, Klug 384 (type,
Field Museum).

To my surprise, upon dissection I found this Perebea-like tree
must be referred to Pseudolmedia, for certainly there is no perianth
as such, although two or three stamens may be borne together sepa-
rated by as many bractlets from the nearest stamen or stamens.
No other Pseudolmedia has the fine dense scabrosity of this species,
nor has "Brosimum Uleanum Mildbr." to which it has recently been
referred. But compare also Perebea elegans!

Admittedly, something is wrong in the classification of the Mora-
ceae. Collections made since the days of Tre"cul, Bentham and Hooker,
and Engler have presented many new problems that are not solved

PERUVIAN PLANTS 63

satisfactorily by interpolation into the work of these earlier botanists,
the method followed by Pittier and Ducke. As a result, we have a
flock of small groups designated as genera that are purely academic,
based on differences in one sex only, and, ignoring similarities, that
result in the most similar of trees being separated generically, to no
practical or scientific purpose. Or, at least, as regards the last, to
no scientific purpose that could not be served more naturally and
more conveniently by sectional groupings under fewer common
(i.e. generic) names.

Perebea elegans, spec, nov., arbor usque 20 m. alta; ramulis
junioribus leviter angulatis fulvo-tomentulosis demum glabratis;
petiolis 5-7 mm. longis; foliis chartaceo-coriaceis oblongo-ellipticis
basi paullo oblique acutis apice breviter caudato-acuminatis, ad 12
cm. longis, 5 cm. latis, plerumque minoribus, supra glabris, subtus
pallidipribus minute praecipue ad nervos venasque puberulis; nervis
lateralibus circa 12 solum subtus cum yenis reticulatis mediocriter
prominulis; receptaculis masc. pseudaxillaribus fasciculatis circa 3
(pedunculis yix 5 mm. longis) globosis 7 mm. crassis; bracteis minutis
squamiformibus imbricatis; tepalis 4.1 mm. longis, apice valde
incrassato-concavis vel -cucullatis minutissime puberulis; filamentis
4 crasso-cylindraceis, 2 tepalis distincte longioribus; receptaculis
fern, ignotis. — Peru: Mishuyacu, near Iquitos, Dept. Loreto, Klug
603 (type, Field Museum).

The floral structure is very similar to that of Helicostylis, and in
the absence of pistillate material the generic identity can not be
positive. The validity of Helicostylis, however, on the basis of a
completely superior ovary and slender styles, may be questioned.

Among Perebeas our tree resembles most, apparently, P. lauri-
folia, from which its much smaller receptacles separate it. Neverthe-
less it has recently been named "Brosimum Uleanum Mildbr."
Compare also Pseudolmedia scabra!

Perebea chimiqua, spec, nov., ad P. guianensem valde vergens,
sed foliis basi rotundato-acutis haud cordatis, subtus distincte ubique
scabriusculis, nervis lateralibus haud vel rarissime parceque hirsu-
tulis; receptaculis masc. 5 mm. diam. (pedunculis circa 6 mm. longis) ;
bracteis 3-4-seriatis late ovatis obtusis vel vix acutis fere 1.5 mm.
latis, plus minusve strigosis; tepalis 4, apice paullo incrassatis;
filamentis brevissimis; antheris subrotundatis fere 0.75 mm. latis.—
Peru: San Antonio, Alto Rio Itaya, Dept. Loreto, Williams 3^12
(type, Field Museum).

Ducke has remarked that the foliage aspect of P. guianensis
Aubl. is that of Castillo, Ulei Warb. One would never so describe
our plant, for the leaves are not even slightly cordate. It is more
scabrous than Aublet's tree and lacks (practically) the hirsutulous

64 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

hairs that, at least in some degree, are always present on the leaf
nerves of the latter. The native name was recorded by the collector
as "chimiqua."

Ogcodeia Tessmannii Mildbr. Notizbl. 10: 189. 1927.

Williams under number 3511 has found the hitherto unknown
pistillate plant of this species — at least I refer it here with no
hesitancy, so exactly does the foliage match the staminate material
of Tessmann. Williams collected it along the Rio Itaya at San
Antonio, and noted the Quichua name as "naccho-huasu." The
heads are subglobose, nearly 2.5 cm. thick, the inner series of large
bracts triangular-ovate, acuminate, most minutely puberulent as
also the rigid-aculeate bractlets 4-5 mm. long.

Ogcodeia tamamuri, spec, nov., ut videtur proxima 0. calo-
neurae sed ramulis solum minutissime sparseque puberulentis et
nervis venisque inconspicuis; foliis breviter (circa 1 cm.) caudato-
acuminatis, plerumque circa 2 dm. longis, 6 cm. latis. — Peru: Tim-
buchi on the Rio Nanay, Dept. Loreto, Williams 992 (type, Field
Museum).

This staminate specimen has the receptacles, bracts, and bract-
lets of 0. ternstroemiiflora Mildbr., but more nearly the foliage of
0. caloneura (Hub.) Macbr. It differs at once from 0. glabra (Spruce)
Mildbr. in its short petioles and relatively fine venation. So, as it
is aberrant if referred to any of these three species, it at present at
least, so far as the trees compared are concerned, requires a name
of its own, which may well be the native one.

Ogcodeia Ulei (Warb.) Macbr. has been collected in Yurimaguas
at Puerto Arturo by Williams (5194) and also by Killip and Smith
who, however, consider their specimen as representing Naucleopsis
macrophylla Miq., a disposition in which I can not agree.

Sorocea opima, spec, nov., fruticosa; ramulis sparse puberulis;
petiolis 7-9 mm. longis; foliis pblongo-ellipticis integris vel undulato-
marginatis chartaceis yix nitidis glabris supra obscure subtus con-
spicue reticulato-venosis, nervis lateralibus primariis circa 12, basi
paullo inaequilateris subacutis vel rotundatis, apice caudato-acu-
minatis, acumine ad 2 cm. longo, ad medium 5 mm. lato, interdum
fere 2 dm. longis et 6 cm. latis, vulgo minoribus; racemis fern, extra-
axillaribus, dense puberulis, rhachidibus crassis, usque ad 5 cm.
longis; fructibus demum glabris 1.5 cm. longis, 1 cm. crassis. — Peru:
Caballo-cocha, Dept. Loreto, Williams 2357 (type, Field Museum).

In its foliage most like S. Ulei Warb., with broader elliptic-
obovate leaves. The large fruits appear to be distinctive.

PERUVIAN PLANTS 65

Roupala Dielsii, spec, nov., arbuscula ubique glabra; ramulis
gracilibus pauciramosis dense foliatis; foliis oblongo-ellipticis vel
interdum paullo oboyatis in petiolum brevem 3-5 cm. cuneato-
attenuatis apice breviter late acuminatis, demum subcoriaceis, 10
cm. longis et 4 cm. latis supra nitidulis vix manifeste venosis, subtus
opacis, venis lateralibus leviter elevatis obscure reticulatis, ad apicem
baud prominente denticulatis infra medium undique integris; racemis
folio subaequalibus glabris vel glabrescentibus, splum ovario rufo-
tomentoso, cum floribus vix 2 cm. latis; pedicellis circa 3 mm. longis,
floribus circa 3.5 mm. longis; stylo circa 6 mm. longo, stigmate dis-
tincte incrassato-clavato. — Peru: Along Rio Itaya, Dept. Loreto,
Williams 114 (type, Field Museum).

The aspect of this shrub is not altogether that of Roupala, but
the anthers are pedicellate and the stigma is terminal, in accord
with the generic character. Its alliance is probably with Roupala
rhombifolia but the red pubescence that, at least in some degree, is
present in that species is here lacking, and the somewhat differently
shaped leaves are borne on much longer petioles. There is also a
floral difference.

Roupala Raimondii, spec, nov., ut videtur arbuscula; ramulis
glabris, apice evanescenter rufo-pubescentibus, gracilibus, dense
folipsis; foliis 4-7-jugis demum ad 2 vel 2.5 dm. longis, glabris;
petiolis ad 4 cm. longis; foliolis oblique ovatp-lanceolatis basi valde
inaequalibus subintegris cuneatis, basi apiceque exceptis grosse
spinescenti-serratis, ad 10-14 cm. longis et 3.S-4 cm. latis, terminal!
paullo majore, sed omnibus plerumque minoribus, venis venulisque
utrinque obscure laxeque reticulatis, supra subnitidulis subtus opacis
rufidulis; acumine integro, 1.5-2 cm. longo; dentibus plerumque 4
mm. longis, basi 2.5 mm. latis vel ad 6 mm. longis, basi 3.5 mm.
latis, haud vel vix incurvis sensim apiculatis, sinubus definite acutis;
racemis ignotis. — Peru: Crown of Sierras, Tarapoto, Dept. San
Martin, 750 m., December, 1929, Williams 5997 (type, Field Museum).
Between Moyobamba and Tarapoto, Raimondi (Herb. Berol.).

With a number of names already available for sterile Roupalas
with pinnate leaves, the wisdom of adding another to the list may be
questioned, but expediency seems to sanction it in this instance.
The original vegetation at Tarapoto has been largely destroyed, and
since no collector except Williams of the many who have sought
species in that locality have apparently found this shrub since
Raimondi, it must be, as Williams indicates, rare. A definite record
of it, therefore, may be worth while, especially as it does not seem
to be the young foliage of a species with simple leaves on the older
branches or it would by now have been found in the same region.
Roupala complicata is known from Moyobamba, but I think it is
always simple-leaved. Among pinnate-leaved species our plant

66 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

suggests R. heterophylla, R. adiantifolia, and R. affinis, but in the
shape of the leaflets and in the nature of the teeth it is, precisely at
least, separable. My description is drawn in part from Raimondi's
leaves, but undoubtedly they are of the same species as Williams'.

Roupala complicata HBK. To this species I recently referred
Kittip & Smith 24891 from San Ramon, Dept. of Junin, several
other collections from the same region, and one from Tarapoto
(Williams 5703}. Weberbauer had earlier collected the same tree
a number of times and in the Dahlem herbarium his specimens are
determined without question as R. complicata HBK. A sheet of
the latter, "ex herb. Humboldt," is at Dahlem; it is not too poor,
and I should never have questioned the identity of the recent material
with it if someone had not determined the latter as R. Gardneri
Meisn.!

I think the diverse opinion rests mostly upon the question of
the validity of the latter species. The Peruvian plant matches it
slightly better in leaf form, the Colombian tree having mostly less
acuminate leaves. In a series of specimens the differences in pubes-
cence, including color, brought out by Meisner in DC. Prodr. 14:
428, appear to exist in degree. I must say that at present I see no
reason for distinguishing the Peruvian tree, even if the Brazilian
R. Gardneri is really another species.

Panopsis acuminata (Meisn.), comb. nov. Andripetalum
rubescens Schott, var. acuminatum Meisn. in DC. Prodr. 14: 346.
1856.

Klug under his number 1540 has collected at Mishuyacu near
Iquitos an exact match of the type of this species. The example
found by Mr. Klug was 8 m. high with cream-colored flowers. With
this confirmation of character, the plant seems to represent a species
amply distinct from both A. Sprucei and P. rubescens in its thread-
like but stiff pedicels 10-12 mm. long and glabrous inflorescence.
Its leaves vary from acute to long-acuminate. It seems evident
that A. Sprucei Meisn. op. cit. 347 is only a variety or more likely
only a race of P. rubescens. In herbaria there is some confusion in
the Spruce labels between A. Sprucei and P. acuminata, but the
latter is at once marked by its pedicels.

Williams, under numbers 1121 and 1044, has collected on the
Rio Nanay a variant of P. rubescens which has the characteristic
fulvous-pubescent racemes and lax short pedicels of that species
(including A. Sprucei) but conspicuously acuminate leaves. This
small tree may be known as

PERUVIAN PLANTS 67

Panopsis rubescens (Pohl) Schott, yar. simulans, var. nov.,
foliis longe acuteque acuminatis. — Peru: Timbuchi on the Rio Nanay,
Dept. Loreto, Williams 1044 (type, Field Museum). Also at Man-
finfa, Williams 1121.

Vochysia diversa, spec, nov., glaberrima praeter inflorescentiam
minutissime parceque puberulam; ramulis subteretibus fere nigres-
centibus ut yidetur decorticantibus; petiplis 0.5-1 cm. longis; foliis
verticillatis rigido-coriaceis in sicco plus minusve nigrescentibus supra
nitidulis subtus fere opacis, costis lateralibus ubique tenuibus paullo
prominulis mediocriter reticulatis, oblongo-lanceolatis vel -ellipticis,
basi sensim in petiolum attenuatis apice rotundatis interdum brevis-
sime emarginatis baud mucronulatis, 8-12 cm. longis, 2.5-3.5 cm.
latis; inflorescentiis anguste cylindraceo-pyramidatis circa 1-1.5 dm.
longis, ad 3 cm. latis, cicinnis 1-2-floris; bracteis fere lineari-subulatis,
ad 7 mm. longis; alabastris suberectis paullo clavatis circa 8 mm.
longis; calycis laciniis quatuor majoribus, calcare clavato calyci haud
aequilongo adscendenti-incurvo; petalis paullo obovatis; stylo glabro
apice clavato. — Peru: Mishuyacu, near Iquitos, Dept. Loreto, Klug
685 (type, Field Museum).

A tree 20 m. high with yellow flowers; apparently very similar
to V. venulosa Warm., but the petioles much shorter and the leaves
narrower. It is distinguishable from V. Weberbaueri Beckmann by
its oblongish leaves, solitary (-2) flowers, and prominent bracts.

Qualea impexa, spec, nov., glaberrima pedicellis calycibusque
plus minusve (yix sericeis) puberulis exceptis; ramulis ut videtur
non decorticantibus; petiolis gracilibus 1 cm. longis; foliis ellipticis
circa 1 dm. longis, 4.5 cm. latis, basi rotundatis, apice abrupte
caudato-acuminatis, acumine obtuso ad 8 mm. longo, imprimis
praecipue subtus ad marginem junctim reticulatis, marginibus valde
undulatis, flexili-coriaceis; cicinnis in racemos 1- vel paucifloros dense
paniculatim dispositis; alabastris ovato-conicis, pedicellis subae-
qualibus circa 1.5 cm. longis; calcare quam pedicellis bene breviore,
dependente, elongato-obovoideo, 7 mm. longo, 3 mm. crasso; petalo
late obcordato ad 3.5 cm. longo ad basin breviter unguiculato gla-
bro; anthera marginibus haud barbato; ovario sericeo-pubescente.—
Peru: Mishuyacu near Iquitos, Dept. Loreto, Klug 748 (type, Field
Museum).

There are probably more names already for Amazonian Qualeas
than there are species, as many descriptions read very much alike
and many specimens look enough alike (I have not seen all the
recently described ones). No description or specimen that has come
to my notice seems to conform in all respects to Q. impexa — perhaps
most nearly Q. macropetala Spruce. Mr. Klug noted the tree as
24 m. high, with yellow flowers. They are so congested that they
appear snarled.

68 FIELD MUSEUM OF NATURAL HISTORY — BOTANY, VOL. XI

Dichapetalum stipulatum, spec, nov., scandens; ramulis
praecipue ad apicem dense fulvo-villosis, demum glabratis, stipulis
tarde deciduis ovato-lanceolatis circa 5 mm. longis fere pinnatifido-
fimbriatis; petiolis 3 mm. longis; foliis coriaceo-chartaceis supra
costa media excepta glabris subtus ad neryos solum plus minusve
fulvo-villosis, nervis lateralibus primariis circa 6 supra cum venis
reticulatis rugoso-impressis subtus valde prominentibus, late ellip-
ticis, circa 14 cm. longis, 8 cm. latis, basi rotundatis vel subcordatis,
apice abrupte breyiter acuminatis; inflorescentia corymboso-panicu-
lata dense rufo-villosa multiflora, ad 1 dm. longa, 1.5 dm. lata;
pedicellis 5 mm. longis; petalis "pallide brunneis" circa 2 mm. longis,
late spathulatis, profunde bifidis; filamentis anguste marginatis gla-
bris; styli ramis 2; ovario dense piloso. — Peru: Mishuyacu near
Iquitos, Dept. Loreto, Klug 1081 (type, Field Museum).

An extraordinary species by virtue of the stipules and the sunken
leaf veins, but the flowers very much like those of D. pedunculatum
Baill.

Tapura coriacea, spec, nov., arbor glabra; ramulis griseo-albis
leviter angulatis subcontortis; petiolis 5-8 mm. longis; foliis oblongo-
ellipticis, basi cuneato-subrotundatis, acutis, apice subabrupte cau-
dato-acuminatis (acumine ad 1 cm. longo, obtuso), plerumque 8-10
cm. longis, 3.5-4 cm. latis, coriaceis, supra nitidis, nervis vix notatis,
subtus opacis nervis venisque inconspicuis; flpribus circa 10 ad
medium fere petioli insertis, flavis, 5 mm. longis; pedicellis 2 mm.
longis cum calyce 4 mm. longo sordide puberulis; sepalis inaequalibus
obtusis; corollae lobis ut videtur solum 3 valde dissimilibus, majori-
bus 2 pbovatis apice breviter bifidis, valde 2-cucullatis, intus cum
tubo villoso-lanatis, minoribus subrotundatis; stamina fertilia 3 inter
se aequalia. — Peru: Mishuyacu near Iquitos, Dept. Loreto, Klug
602 (type, Field Museum).

Perhaps nearest T. guianensis Aubl., but the flowers much smaller
and the stamens equal. There are certainly only 3 fertile stamens,
but otherwise the structure of the corolla and the foliage seem to
resemble greatly T. capitulifera Baill. (ex char.).

Trigonia virens, spec, nov., liana; ramulis et imprimis costis
petiolisque plus minusve subadpresse hirsutulis; foliis adultis demum
circa 12 cm. longis, 5-6 cm. latis praeter petiolum et costas fortiores
utrinque glabris vix nitidulis, basi acutis vel obtusis, apice subab-
rupte in acumen longiusculum acutum productis, membranaceis;
costis circa 5-8 subtus mediocriter prominulis; inflorescentiis (calyci-
bus solum extus) subviridi-cano-tomentosis 4-8 cm. longis; cicinnis
nunc 1- mine 2-3-floris racemose dispositis; pedunculis 2-3 mm. longis
quam pedicellis vix longioribus; bracteis anguste ovato-lanceolatis
ad 4 mm. longis; alabastris circa 5 mm. longis; calycis laciniis fere
5 mm. longis; petalis circa 6 cm. longis, 2 carinatis late oblongis.—
Peru: Wooded banks of Rio Itaya above Iquitos, Dept. Loreto,
Killip & Smith 29539 (type, Field Museum).

PERUVIAN PLANTS 69

Probably as nearly allied to T. hypoleuca Griseb. as to any other
species, but glabrate; otherwise comparable to T. Spruceana Benth.
and T. macrantha Warm. Its thin and broad leaves distinguish it
from the former; its pedunculate flower clusters and hirsutulous leaf
nerves from the latter.

Killip and Smith have collected at the same locality (as well as
Klug at Mishuyacu) T. sericea HBK. of Colombia, or at least I see
no pronounced differences. More to have been expected was T.
najadum Warm., from the upper Amazon, but our material has
lenticellate branches and matches well enough the Dahlem Humboldt
specimen.

AUG 2 6 1931

f ILLINOIS

THE
CHENOPOD]

To. 3

ASSOCIATE CT

•UINOIS