XI B R.AR.Y
OF THL
UNIVERSITY
OF ILLINOIS
58,0.5
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 2 11! 70
NOV G<|1984
L161 O-1096
FIELD MUSEUM OF NATURAL HISTORY
FOUNDED BY MARSHALL FIELD, 1893
PUBLICATION 288
BOTANICAL SERIES VOL. XI, No. 1
SPERMATOPHYTES, MOSTLY PERUVIAN-III
BY
J. FRANCIS MACBRIDE
~
B. E. DAHLGREN
ACTING CURATOR, DEPARTMENT OF BOTANY
EDITOR
CHICAGO, U. S. A.
MAY 29, 1931
FIELD MUSEUM OF NATURAL HISTORY
FOUNDED BY MARSHALL FIELD, 1893
PUBLICATION 288
BOTANICAL SERIES f$L.J !
5- ,931
SPERMATOPHYTES, MOSTLY PERUVIAN-III
BY
J. FRANCIS MACBRIDE
ASSISTANT CURATOR OP TAXONOMY
B. E. DAHLGREN
ACTING CURATOR, DEPARTMENT OP BOTANY
EDITOR
CHICAGO, U. S. A.
MAY 29, 1931
PRINTED IN THE UNITED STATES OF AMERICA
BY FIELD MUSEUM PRESS
f]
l\
SPERMATOPHYTES, MOSTLY PERUVIAN III
J. FRANCIS MACBRIDE
Few if any American botanists have had my privilege of so-
journing over a long period at the great herbaria of the Old World.
I could easily write a book upon the happy experiences that have
been mine and the thanks that I owe so abundantly to the members
of the staffs at the several institutions visited. Here I must content
myself with a limited mention of the many who have generously
cooperated with Field Museum, to the benefit, let us hope, of science
generally, in permitting the most complete freedom in carrying on
the work delegated to me. The cordial friendship I have met has
contributed immeasurably to the consummation of this special work.
Apart from it, I have been able to continue to some extent my
interest in the flora of Peru. In presenting the following paper I
acknowledge with pleasure my special indebtedness to Dr. Briquet
and staff, Conservatoire Botanique, Geneva; to Dr. R. Chodat and
staff at the University, Geneva; and to Professors Goebel, Ross, and
Suessenguth at the Botanical Institute, Munich, at which institu-
tions, particularly the first, the paper has evolved. As in the case
of the last number of this series, I am again happy to record my
thanks to Professor Diels and to Professor Pilger and their friendly
associates at Berlin-Dahlem.
1. SOME PERUVIAN CONVOLVULACEAE
WITH A NEW VARIETY
Merremia glabra Hall, f ., var. pubescens van Ooststr., n. var.
Differt ramis petiolis et foliis quoad nervos paginae inferioris in
eodem specimine nunc pubescentibus nunc plus minusve glabris.
Peru: Pampayacu, Hacienda at mouth of Rio Chinchao, about 3,500
feet, July 19-25, 1923, Macbride 5045. "On sunny thickets, fls.
white."
Ipomoea (sect. Eriospermum) clavata (Don) van Ooststr.,
comb. nov. Convolvulus clavatus Pa von, mss. Colony Action clavatum
Don, Gen. Syst. 4: 264. 1838. Ipomoea lactescens Benth. PI.
Hartw. 120. 1839; Hall. f. in Jahrb. Hamb. Wiss. Anst. 16: Beih.
3: 50. 1898. Peru: Pozuzo, about 2,000 feet, June 20-22, 1923,
Macbride 4671. La Merced, about 2,000 feet, Aug. 10-24, 1923,
Macbride 5394; "on brush."
4 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
Ipomoea Plummerae Gray, Syn. Fl. N. Am. ed. 2. 2 1 : Suppl.
434. 1886.
In the Supplement of his Synoptical Flora of North America (434.
1886), Asa Gray gives a diagnosis of this species, followed by a
short description of I. cuneifolia Gray (Proc. Amer. Acad. 19: 90.
1884). In the latter he writes that this species has the "tuber,
peduncles, flowers and habit of the preceding," i.e. /. Plummerae,
but it differs in the leaves, which are "simple, cuneate, laciniate-
dentate at the broad apex, tapering into a short petiole," etc.
In the collections made by Weberbauer in Peru I found two
numbers (7275a and 7275), the first of which I suppose to be identical
with /. Plummerae. The second has the cuneate leaves of /. cunei-
folia and seems to be identical with that species or nearly allied to it.
As the two forms seem to grow together, I assume that the
cuneate-leaved plants represent a variety of Plummerae, which I
call var. cuneifolia, a variety probably identical with /. cuneifolia
Gray. Peru: Carumas, Prov. Moquegua, 2,700 m., Feb. 21-Mar. 6,
1925, Weberbauer 7275 and 7275a. "Open mixed formation. Decum-
bent. Tuberous. Flowers purple."
2. NEW AND OLD PERUVIAN PLANTS
Elodea Potamogeton (Bert.), comb. nov. Diplandra Potamo-
geton Bert. Merc. Chil. 612. 1829; Bull. FeYus. 20: 110. 1830. Anacha-
ris chilensis Planch. Ann. Sci. Nat. III. 11: 75. 1849. Elodea chilensis
Gasp. Monatsb. Berl. Acad. 47. 1857. A. Matthewsii Planch, op.
cit. 74?
The original publication of this transferred name is valid; it is,
not a nomen nudum as has been assumed. The description, in
Spanish, is informal, but there is no mistaking the identity, and
furthermore, this is the only Elodea in the region. The generic name
Philotria Raf. is not acceptable legally in place of Elodea, or at
least not until another congress.
Echinodorus palaefolius (Nees & Mart.), comb. nov. Sagit-
taria palaefolia Nees & Mart. Nov. Act. Acad. Nat. Cur. 11: 21.
1823. Alisma ellipticum Mart, in R. & S. Syst. 7 2 : 1607. 1830. Echi-
nodorus ellipticus Micheli in DC. Monogr. 3: 51. 1881.
Since this species is known to occur in Mexico and in Uruguay
as well as at various intermediate stations, the fact that it does not
seem to have been recorded from Peru is annoying rather than
significant. The var. pubescens (Mart.), comb. nov. E. ellipticus
(Mart.) Micheli, var. pubescens (Mart.) Micheli has a scabrous-
pubescent rather than glabrous inflorescence.
PERUVIAN PLANTS 5
Bulbostylis tenuifolia (Rudge), comb. nov. Scirpus tenuifolius
Rudge, PI. Guian. 18. pi. 22. 1805. Stenophyllus tenuifolius Britton,
Bull. Torr. Club 43: 448. 1916.
Bulbostylis capillaris (L.) Clarke, var. coarctata (Ell.), comb,
nov. Scirpus coarctatus Ell. Bot. S. C. & Ga. 1: 83. 1816. Stenophyllus
capillaris Britton, Bull. Torr. Club. 21: 30. 1894.
It is possible that the variety with a composite umbel of many-
flowered spikelets is merely a robust form.
Pfeiffer, just before the last botanical congress (Repert. Spec.
Nov. 27: 85-91. 1929 and 28: 24-26. 1930) raised a great hue and
cry for the conservation of the generic name Bulbostylis, and with
a notable display of righteousness wrote: "The [next] most essential
point in nomenclature is ... the avoidance of any useless introduc-
tion of unnecessary names." He then proceeded to coin a super-
fluous name as follows: "Bulbostylis conspicua (Boeck.) H. Pfeiff. n.
comb, vel Stenophyllus conspicuus (Boeck.) H. Pfeiff. n. comb."
Bulbostylis should be conserved, which can still be done with reason
since there has not yet been a general transfer of the names to Steno-
phyllus.
Dichromena setacea (Berg.), comb. nov. Schoenus setaceus
Berg. Act. Helv. 7: 130. 1772.
As pointed out by me in previous papers, the genera Rynchospora
and Dichromena constitute in fact but a single genus, for which the
latter is the prior name. After the species are all transferred, some
one will probably demand the employment of the former name.
Accordingly, in continuing my legal adoption of the name Dichro-
mena (cf. Field Mus. Bot. 4: 165-166. 1929), I am making the above
and the following new transfers of species that it has been desirable
or necessary to associate with the flora of Peru.
Dichromena Linkii, nom. nov. Rynchospora tennis Link,
Jahrb. 3: 76. 1820, not D. tennis Steud. Syn. Cyp. 135. 1855.
Dichromena barbata (Vahl), comb. nov. Schoenus barbatus
Vahl, Eclog. 2: 4. 1798. Rynchospora barbata Kunth, Enum. 2:
280. 1837.
Dichromena Marisculus (Lindl. & Nees), comb. nov. Ryncho-
spora Marisculus Lindl. & Nees in Mart. Fl. Bras. 2 1 : 142. 1842.
Dichromena distans (Michx.), comb. nov. Schoenus distans
Michx. Fl. Bor. Amer. 1: 36. 1803. Rynchospora distans Vahl,
Enum. 2: 235. 1805.
Dichromena cymosa (Ell.), comb. nov. Rynchospora cymosa
Ell. Bot. S. C. & Ga. 1: 58. 1816.
6 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
Dichromena cyperoides (Sw.), comb. nov. Schoenus cyperoides
Sw. Prodr. 19. 1788. Rynchospora cyperoides Mart. Denkschr. Akad.
Wiss. Muench. 6: 149. 1816-17.
Dichromena gigantea (Link), comb. nov. Rynchospora gigan-
tea Link, Jahrb. 3: 76. 1820.
Dichromena triflora (Vahl), comb. nov. Rynchospora triflora
Vahl, Enum. 2: 232. 1806.
Dichromena Schoenus, nom. nov. Scirpus cephalotes L. Sp.
PL ed. 2. 76. 1762, not D. cephalotes Britton.
Dichromena amazonica (Poepp. & Kunth), comb. nov. Ryn-
chospora amazonica Poepp. & Kunth in Kunth, Enum. 2: 292. 1837.
Dichromena corymbosa (L.), comb. nov. Scirpus corymbosus
L. Cent. 2: 7. 1756. Rynchospora corymbosa Britton, Trans. N. Y.
Acad. Sci. 11: 84. 1892.
Dichromena Pearcei (Clarke), comb. nov. Pleurostachys
Pearcei Clarke, Kew. Bull. Add. Ser. 8: 41. 1908.
In a regional treatment, at least, there is no useful purpose served
in maintaining the segregate genus Pleurostachys, the generic value
of which is not on a par with that of the other genera within the tribe.
Dichromena peruviana (Clarke), comb. nov. Pleurostachys
peruviana Clarke, Kew Bull. Add. Ser. 8: 42. 1908.
Dichromena aberrans (Clarke), comb. nov. Rynchospora
aberrans Clarke, Kew Bull. Add. Ser. 8: 35. 1908.
This Brazilian species, allied to D. exaltata, has been found
recently in Peru, according to Pf eiff er.
Rhodospatha Poepp. in Poepp. & Endl. Nov. Gen. 3: 91. 1845.
This genus may well be amended to include the plants that have
been referred to Stenospermatium Schott (Gen. Ar. 70. 1858), since
the latter in general has no distinction except the basal instead of
lateral attachment of the ovules. If this character properly forms
a basis of generic definition, other groups in the family now included
under one name (as Philodendron, for example) should be segregated.
Rhodospatha in this broader sense is too near Raphidoflora Hassk.
(which no doubt should include Afroraphidophora Engler and Epi-
premnum Schott, also based on ovule placement), but that is entirely
Old World in distribution and is usually distinguishable by its 1- or
partially 2-celled ovary. The alliance of all these plants is very close
to Monstera, but here there is the very practical distinction in the
difference of leaf nervation (with one or two connecting species as
PERUVIAN PLANTS 7
exceptions). The disposition of the several groups in the tribe may
be perplexing if emphasis is placed on the one or two aberrant species
uniting genera that, from a standpoint of convenience at least,
should be retained, but this confusion is avoided when these aberrant
species are left as such; that is, as "exceptions," which are, inci-
dentally, rarely met with. They will also fit better into a rational
scheme of classification where their position will be clearly that of
connecting links of diverging groups. Accordingly, it seems most natu-
ral to merge Stenospermatium with Rhodospatha and to treat other
aberrant plants (cf. Engler & Krause, Pfl. IV. 23B: 16. 1908), such
as Anepsias Schott, Alloschemone Schott and Amydrenium Schott,
also as connecting species and not as separately developed genera.
The Peruvian species of Stenospermatium to be transferred to Rhodo-
spatha are:
Rhodospatha amomifolia (Poeppig), comb. nov. Monstera
amomi/olia Poeppig, Nov. Gen. 3: 88. 1845. Stenospermatium
amomifolium Schott, Prodr. 348. 1860.
Rhodospatha crassifolia (Engler), comb. nov. Stenospermatium
crassifolium Engler, Bot. Jahrb. 37: 114. 1905.
Rhodospatha flavescens (Engler), comb. nov. Stenospermatium
flavescens Engler, op. cit. 111. 1905.
Very doubtfully more than a variety of the next.
Rhodospatha Mathewsii (Schott), comb. nov. Stenosperma-
tium Mathewsii Schott, Gen. Ar. 70. 1858.
Rhodospatha Spruceana (Schott), comb. nov. Stenosperma-
tium Spruceanum Schott, Gen. Ar. 70. 1858.
Rhodospatha popayanensis (Schott), comb. nov. Stenosper-
matium popayanense Schott, Oesterr. Bot. Zeitschr. 9: 39. 1859.
This species, similar to the preceding but with a much shorter
spadix stipe, is to be expected in Peru. The former is found also
in Colombia.
Rhodospatha Weberbaueri (Engler), comb. nov. Stenosper-
matium Weberbaueri Engler, Bot. Jahrb. 37: 110. 1905.
This species, originally from Huanuco, has been found recently
at La Merced in Junin by Schunke, according to Professor Krause.
Anthurium Burchellianum (Engler), comb. nov. A. pandur-
atum Mart. var. Burchellianum Engler, Pfl. IV. 23B: 279. 1905.
Rather similar to A. davigerum Poepp. & Endl., but the petioles
shorter, the leaf divisions often fewer, and the peduncle twice and
8 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
the spadix half as long. Huber, who found it in eastern Peru, noted
the leaves as attaining a width of 1 m.
Xanthosoma Poeppigii Schott, var. mafaffa (Schott), comb,
nov. X. mafaffa Schott, Araceen Betreff. 2: 5. 1855.
X. Poeppigii was published a year earlier than X. mafaffa.
Engler in Flora Brasiliensis treated the former as a variety of the
latter. In the var. mafaffa the basal leaf lobes are a little introrse
and the tube of the spathe is yellowish green.
Paepalanthus peruvianus (Ruhl.), comb. nov. Syngonanthus
peruvianus Ruhl. Pflanzenr. IV. 30: 253. 1903.
The genus Syngonanthus was established by Ruhland (op. cit.
30) to include those species of Paepalanthus with more or less connate
(at the middle) petals, in the case of the female flowers. This seems
to me to be a character that serves most usefully as a means of
grouping merely sectionally the supposedly related forms. I think
the natural genera in the family are defined only by the variation in
the number of the stamens and by the number of the anther cells.
On this basis Blastocaulon Ruhl. (op. cit. 223) and Phylodoce Mart.,
widely separated by Ruhland because the petals of the latter are
partially adnate, are to be merged. The only character in this case
remaining to Phylodoce that defines it in contrast to Tonina Aubl.
is the presence of well-developed petals; and in the otherwise similar
Lachnocaulon Kunth the petals are reduced to hairs. These four
groups defined as genera constitute, therefore, from a purely dis-
interested standpoint, one natural genus. As they exhibit some
habital differences, their maintenance in regional treatments may
sometimes be convenient.
Dichorisandra hexandra (Aubl.) Standley, var. persicariae-
folia (Clarke), comb. nov. D. Aubletiana R. & S., var. persicariae-
folia Clarke in DC. Monogr. 3: 274. 1881.
Apparently the species is represented in Peru only by this form
with transversely striate leaves, which perhaps is a distinct species.
I collected it in Junin at La Merced.
Dichorisandra Ulei, nom. nov. D. longifolia Ule, Verh. Bot.
Verein Brandenb. 50: 71. 1908, not Martens & Gal., 1842.
Distinguished by Ule from D. ovata Mart, by the longer and
glabrous leaves. It may be only a variant of D. thyrsiflora, which
typically, however, has leaves gradually narrowed to a petiole, a
subpaniculate inflorescence, and obtuse petals. The species of the
genus are either poorly defined or poorly understood.
PERUVIAN PLANTS 9
Juncus Liebmanni, nom. nov. Juncus brevifolius Liebm. Vid
Meddel. For. Kjoben. 40. 1850, not Hoffm. & Link ex Rostk., 1801.
Although Buchenau has proposed two forms of this Mexican and
Ecuadorian species as varieties, his name for the typical state (var.
mexicana) has already been used in the genus. His other variety
represents the Ecuador plant, perhaps distinct.
Fortunatia, gen. nov. Scilla L. sect.? Hesperoscilla Benth. in
Benth. & Hook. f. Gen. 3: 815. 1883.
The type is:
Fortunatia biflora (R. & P.), comb. nov. Scilla biflora R. & P.
Fl. Peruv. 3: 69. pi. 102. 1802. Ornithogalum biflorum Don in Sweet,
Br. Fl. Gard. Ser. 2: 4. pi. 246. 1838.
Bentham, evidently with hesitancy, referred this Peruvian lily to
Scilla but called attention to its discrepant characters the 3-ovulate
cells, the complanate filaments, and the open inflorescence of gemi-
nate, long-pediceled flowers. It is more diverse from Ornithogalum.
Indeed, from Krause's key (Pflzf. ed. 2. 15a: 243-254), there appears
to be no reason for excluding Fortunatia from the Asphodeloideae
where it would find a place in the Chlorogalinae and be no more
aberrant than in the Scilloideae. As genera in the Liliaceae are
commonly accepted, this plant constitutes most properly and con-
veniently a generic entity differing technically and in aspect from
the exclusively Old World groups Scilla and Ornithogalum. Since
neither custom nor rule requires the advancement of the sectional
name, I have proposed Fortunatia in well-merited commemoration
of the work of the Peruvian botanist, Fortunate Herrera. Professor
Herrera has kindly sanctioned the choice.
Zephyranthes tubiflora (L'H4r.) Schinz, var. flammea (R.
& P.), comb. nov. Amaryllis flammea R. & P. Fl. Peruv. 3: 56. pi.
286. 1802. Z. flammea Baker, Amaryl. 37. 1888.
Possibly this is a color form, but I have seen no specimen. Z.
flava (Herb.) Baker appears to be scarcely distinct specifically, since
its only noteworthy difference is the absence of scales in the perianth
tube. Z. Beustii Schinz (Viertelj. Zur. Nat. Ges. 424. 1915) seems,
from character, to be indistinguishable from Z. flava.
Zephyranthes Briquetii, spec. nov. Tota planta pygmaea
prostrata 5-8 cm. longa; bulbi anguste ovato-oblongi 8 mm. diam.
in collum circa 1-2 cm. longum attenuati; foliis 2-3 linearibus
3-6 cm. longis circa 1 mm. latis ad apicem subulato-acuminatis;
scapo 1-3 cm. longo unifloro; spatha membranacea fere ad basin
bifida circa 1.5 cm. longa; pedicello 3-4 mm. longo; ovario 6 mm.
longo; perigonio albo (interdum violaceo-maculato) circa 2 cm.
10 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
longo, tubo gracili circa 5 mm. longo in faucem subabrupte ampliato,
laciniis subaequalibus ut videtur subrotundatis et inter se connatis;
filamentis liberis basin versus dilatatis circa 3 mm. longis ad apicem
tubi affixis; stylo stamina aequante vel superante, stigmate trilobato-
folioso. Peru: With cushion and rosette plants, Carumas, Prov.
Moquegua, Feb. 27, 1925, Weberbauer 7322 (type, Field Museum).
Notwithstanding the large number of species of Zephyranthes
described in recent years, this little Peruvian one seems to be refer-
able to none of them. Perhaps it is as nearZ. gracilis Herb. (Amaryl.
172. pL 29. 1837) as any species, but its differently shaped flowers
and much shorter style with foliose stigmas readily separate it. The
last character apparently has not been described for any species with
the habit of Z. Briquetii, although it is entirely characteristic for the
genus. Dr. Weberbauer observed that the leaves and flowers were
both prostrate.
In studying this collection at Geneva, Dr. Briquet kindly verified
my analysis and I welcome the opportunity to associate his name
with an apparently unclassified plant.
Stenomesson Elwesii (Baker), comb. nov. Callithauma viridi-
florum (R. & P.) Herb., var. Elwesii Baker, Gard. Chron. n. s. 9:
756. 1878. S. viridiflorum (R. & P.) Benth., var. Elwesii Baker,
Amaryl. 116. 1888.
S. viridiflorum with a merely crenate stamineal cup and S.
Elwesii with the cup deeply 6-cleft and its lobes quadrate and emar-
ginate must be treated as distinct if there is any taxonomic signifi-
cance in extreme variation in the nature of this organ. Otherwise
a number of Peruvian species must be regarded as variants.
Stenomesson viridiflorum (R. & P.) Benth., var. angusti-
folium (Herb.), comb. nov. Callithauma angustifolium Herb. Bot.
Mag. 67: pi. 3866. 1841.
Baker suggested (Amaryl. 116. 1888) that this form could be
regarded as a variety, but he did not make the transfer. Compared
with the typical state of the species, its leaves are narrower, its
stamens shorter than the perianth, and its style exserted, all char-
acters which often are variable in the genus and family.
Stenomesson Macleanicum (Herb.), comb. nov. Clitanthes
Macleanica Herb. Bot. Reg. 25: Misc. 87. 1839. Coburgia Mac-
leanica Herb. op. cit. 28: Misc. 55. 1842.
Similar to S. luteum (Herb.) Baker, but the scape 1-flowered and
the stamineal cup truncate between the short but slender filaments.
Both these species have been confused with S. recurvatum, which has
elongate filaments and several reddish-yellow flowers.
PERUVIAN PLANTS 11
Stenomesson pauciflorum (Lindl.) Herb.,var. curvidentatum
Herb.), comb. nov. S. curvidentatum Herb. Bot. Mag. 53: pi 2640.
1826.
Apparently like the type, but the perianth slender and the bifid
stamineal teeth elongate and curved.
Stenomesson variegatum (R. & P.), comb. nov. Pancratium
variegatum R. & P. Fl. Peruv. 3: 55. 1802. P. incarnatum HBK.
Nov. Gen. 1: 280. 1815. S. incarnatum Baker, Saund. Ref. Bot. sub
pi. 308. 1873.
Variable, particularly in color, and a number of forms have been
proposed as species.
Hymenocallis narcissiflora (Jacq.), comb. nov. Pancratium
narcissiflorum Jacq. Fragm. 86. pi. 138. 1809. P. calathinum Ker,
Bot. Reg. 3: pi. 215. 1817. H. calathina Nichols. Diet. Gard. 2:
165. 1886.
This beautiful plant was introduced in 1794 but has scarcely been
known except in cultivation. According to Herbert, it came from
the Andes of Bolivia or Peru.
Hymenocallis pedunculata (Herb.), comb. nov. Ismene pedun-
culata Herb. Amaryl. 222. pi 35. 1837. /. Macleana Herb. Bot.
Mag. 65: pi 3675. 1839. H. Macleana Nichols. Diet. Gard. 2:
165. 1886.
Similar to H. narcissiflora but for the straight, slender, and much
shorter tube, the linear segments, and the shorter (about 3.5 cm.) cup.
Hymenocallis longipetala (Lindl.), comb. nov. Elisena longi-
petala Lindl. Bot. Reg. 24: Misc. 79. 1838.
Hymenocallis sublimis (Herb.), comb. nov. Elisena sublimis
Herb. Bot. Mag. 67: sub pi 3873. 1841.
It is almost amazing that these plants have not before been in-
cluded in Hymenocallis, where they so obviously belong. Herbert
himself in Bot. Reg. 25: Misc. 142. 1839, and again in Saund. Bot.
Ref. 4: pi 264- 1871 called attention to their generic similarity.
H. deflexa (Herb.) Baker, a typical Hymenocallis, has been produced
by hybridizing H. narcissiflora and H. longipetala.
Urceolina peruviana (Presl), comb. nov. Sphaerotele peruviana
Presl, Rel. Haenk. 1: 119. pi 16. 1827. S. coccinea Link, Kl. & Otto,
Ic. pi 38. 1840. Pentlandia miniata Herb. Bot. Reg. 25: pi 68.
1839. U. miniata Benth. & Hook. Gen. PI. 3: 732. 1883.
When Link, Klotsch and Otto described S. coccinea they were
concerned only with distinguishing their plant from S. peruviana,
and so far as description and plate indicate, there is no reason what-
soever to question their judgment that the plants are generically the
12 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
same. Furthermore, there seems also to be no real specific difference.
Nevertheless every one since Bentham has regarded the Presl plant
as a species of Stenomesson and referred S. coccinea to U. miniata.
Presl makes no mention of the stamineal cup that typifies the former
genus (although he characterizes the floral structure of his plant in
detail), nor does the plate show one. There appears, therefore, to be
no reasonable doubt that these three plants are congeneric and
conspecific. The flowers, as Herbert long ago remarked, simulate
those of Stenomesson croceum, so weakly developed is the urceolate
character that for other species so readily marks the genus, but, lack-
ing the stamineal cup, the plant is best retained in Urceolina and
regarded as a connecting species.
Urceolina urceolata (R. & P.), comb. nov. Crinum urceolatum
R. & P. Fl. Peruv. 3: 58. pi 287. 1802. U. pendula Herb. Amaryl.
193. 1837.
This species is readily known by its almost filiform, green tube.
Phaedranassa dubia (HBK.), comb. nov. Haemanthus dubius
HBK. Nov. Gen. 1: 281. 1816. Crinum quitense Spreng. Syst. 2:
55. 1825. P. chloracea Herb. Bot. Reg. 31: pi 17. 1845.
This is an Ecuadorian species, well marked by its red flowers
deeply stained with green at the tip.
Nothoscordum fictile, spec, nov., bulbo ovoideo 2-3 cm. crasso
collo longo (1-1.5 dm.) instructo; caulibus brevissimis ut videtur
nullis yaginis foliorum omnino involutis; foliis prostratis lineari-
oblongis anguste longiacuminatis 1-2 dm. longis, 4-10 mm. latis;
pedicellis flexuosis gracilibus valde inaequalibus 1-2 cm. longis;
floribus albis 4-5 mm. longis; segmentis oblongo-ellipticis ad basin
breviter connatis, filamentis subulatis circa 2.5 mm. longis; antheris
brunneis vix 1 mm. longis; stylo yix 1 mm. longo; capsula immatura
4 mm. longa ut videtur subovoidea. Peru: Carumas, Moquegua,
Feb. 21-Mar. 6, 1925, Weberbauer 7262 (type, Field Museum).
Among the species included in Beauverd's painstaking synopsis
of the genus (Bull. Herb. Boiss. II. 8: 993-1007. 1908) which, by the
way, is not mentioned in the recent edition of the Pflanzenfamilien
along with other omissions, not to mention commissions, including
the too frequent violation of the International Rules of Botanical
Nomenclature N. fictile is not especially dissimilar to N. sessile
(R. E. Fries) Beauv., which heretofore was unique in habit. The latter,
however, has narrower leaves, a much longer style (about twice as
long as the ovary), and smaller fruit. So far as known, it is Argentine.
Dioscorea monadelphoides, nom. nov. Helmia monadelpha
Kunth, Enum. 5: 421. 1851, not D. monadelpha Griseb. 1875. D.
PERUVIAN PLANTS 13
monadelpha Pax, Pflanzenf. 2 5 : 133. 1888. D. subhastata Veil. Fl.
Flum. 10. pi. 121. 1827?
Kunth as recently as 1924 (Pflanzenr. IV. 43: 126) has used the
name of Pax for this plant, when modern custom in nomenclature
clearly requires that the right to the name be restricted to the
species of Grisebach. The Vellozo name, since its application is
questionable and not determinable, should be dropped.
Nemastylis Huyanae, spec, nov., adscendens-erecta glaberrima;
bulbo ovoideo-conico, circa 2 cm. longo et 1.5 cm. crasso, tunicis
fuscescentibus exterioribus abrupte subulato-caudato-acuminatis;
caulibus 1.5-3 dm. longis plerumque simplicibus subtortuosis; foliis
2 (1 radical!) 1.5-4 dm. longis, ad basin et versus apicem subulatim
attenuatis ad 6 mm. latis; inflorescentiis terminalibus; spathis sub-
aequalibusque similibus acutis circa 4.5 cm. longis, albido-papyraceis;
floribus circa 4, mediocriter exsertis, subviridibus atque definite pur-
pureo-maculatis circa 12 mm. longis; lobis anguste obovatis; fila-
mentis in tubum gracilem connatis; antheris vix 4 mm. longis, plus
minusve tortis; styli ramis 2-partitis antheris paullo longioribus;
pedicellis gracilibus circa 1.5-2 cm. longis; capsulis ut videtur ad
basin acutis. Peru: Matucana, Macbride & Featherstone 469 (type,
Field Museum).
Apparently this species is related to N. nana Wats., which is
entirely different in habit. The few species known with greenish
and spotted flowers are not Peruvian and appear to be essentially
different from this plant. The name commemorates King Huyana,
father of the last Inca kings, Huascar and Atahuallpa.
N. Pearcei Baker, or rather a variant of it, also was collected at
Matucana. The specimens have the large purple flowers and the
long anther column of that species, but the anthers are longer than
the style branches.
Tigridia lobata (Herb.), comb. nov. Hydrotaenia lobata Herb.
Bot. Reg. 30: Misc. 66. 1844.
Similar to T. grandi flora (Cav.) Diels but apparently distinct, the
flowers campanulate, with oblong-cuneate erect segments densely
brown-spotted at base.
Gostus amazonicus (Loes.), comb. nov. Costus Malortieanus
Wendl., var. amazonicus Loes. Notizbl. 10: 710. 1929.
As species in Costus are at present defined, this Peruvian plant
seems to exhibit characters that give it that rank.
Heliconia Schumanniana Loes. Bot. Jahrb. 54: Beibl. 117:
12. 1916. H. Schumanniana Loes., var. basirubra Loes. loc. cit.
This is a species of Heliconia that admittedly shows considerable
variation in the color of the bracts and flowers, and the fact raises
14 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
the question whether many of the twenty collections recorded as
distinct species in Peru are not rather similar color variants in at
least some instances. However this may be, the following color
forms of this plant may take that taxonomic status.
Heliconia Schumanniana Loes., forma apicirubra (Loes.),
comb. nov. H. Schumanniana Loes., var. apicirubra Loes. Bot.
Jahrb. 54: Beibl. 117: 12. 1916.
Heliconia Schumanniana Loes., forma acreana (Loes.), comb,
nov. H. Schumanniana Loes., var. acreana Loes. loc. cit.
Renealmia cernua (Sw.), comb. nov. Costus cernuus Sw. ex
R. &S. Syst. 1:25. 1817.
This unusual species, with the aspect of Costus, has, apparently,
never been "properly" christened, that is, in accord with the usage
of modern nomenclature.
Monotagma spicatum (Aubl.), comb. nov. Maranta spicata
Aubl. Hist. PI. Gui. 1 : 4. 1775. Phrynium Parkeri Rose. Monandr.
PI. pi. 4.2. 1828. Monotagma Parkeri Schum. Pflanzenr. IV. 48:
168. 1902.
The Peruvian plant seems to be identical with the type of the
species from French Guiana, a not particularly remarkable fact and
yet at the same time not particularly usual. No other Monotagma
known in Peru has pilose-annulate petioles.
Hedyosmum Kanehirae, spec, nov., arbuscula ad 5 m. alta;
ramulis subquadrangulatis dense verruculosis; foliis numerosis rigido-
coriaceis opacis breve (3-6 mm.) petiolatis glabris minute denseque
crenato-denticulatis oblongo- vel lanceolato-ellipticis, basi subacutis,
apice abrupte obtuseque acuminatis, plerumque circa 6.5 cm. longis
et 2.5-3 cm. latis; nervis lateralibus subtus mediocriter prominentibus
et plus minusve reticulatis; floribus masc. ignotis, cymulis fern,
racemoso-spicatis subapproximatis 4-5 mm. longis, 3-4 mm. latis;
bracteis drupas subaequantibus vel paullo brevioribus; drupis nigris
trigonis acutis 1.5 vel vix 2 mm. longis. Peru: Pan de Azucar,
Huanuco, Sawada 64 (type, Field Museum).
Having had the pleasure recently of naming a species for Mr.
Sawada, well merited by his active interest in the flora of Peru, I
dedicate this Hedyosmum to his friend, Dr. Kanehira, who has kindly
shared his collections and those of Sawada with Field Museum. The
species, among Peruvian ones, resembles most H. Lechleri Solms,
from which its scurfy branchlets, small elliptic leaves, and tiny black
drupes easily separate it. These characters in conjunction with its
glabrous leaves appear to distinguish it from all species.
PERUVIAN PLANTS 15
Hedyosmum Huascari, spec, noy., ut videtur dioicum; ramulis
subteretibus striatis glabris; petiolis circa 5 mm. longis; foliis ovato-
ellipticis, vix acutis, rigido-coriaceis glabris obscure denseque denti-
culatis plerumque circa 6 cm. longis et 2-2.5 cm. latis; venis haud
prpminentibus supra vix notatis; spicis masculis ignotis; cymulis
spicato-aggregatis, 5-6 mm. longis, 3-4 mm. latis; bracteis drupis
brevioribus; drupis pallide brunneis circa 3.5 mm. longis. Peru:
above Tabaconas, Cajamarca, Weberbauer6113 (type, Field Museum).
Similar to H. Lechleri, but apparently distinguishable by its
coriaceous and broader, obtusish leaves. Its name commemorates
Huascar, son of the Inca king Huyana. He died in battle over the
kingdom inherited jointly with his brother Atahuallpa.
The species of Hedyosmum are obscure in character. While
proposing the above plants as new, I am (as usual) unable to distin-
guish some of the species described by others. For example, the
characters relied upon by Solms to distinguish H. integrum Cord,
and H. Sprucei Solms from H. racemosum (R. & P.) G. Don, seem,
at least in part, explained as degrees of maturity. They may be
varieties of one species, but more material is needed than the types
or cotypes seen by me to decide this. Some forms of this group appear
to approach H. arborescens Sw. too closely. Melchior (Notizbl. 9:
1036. 1926) has called attention to the weak distinction between
H. racemosum and H. brasiliense Mart. Besides my species and
H. racemosum, there are perhaps four others in Peru that are fairly
well marked: H. scabrum (R. & P.) Solms, H. Dombeyanum Solms,
H. Lechleri Solms and H. glaucum (R. & P.) Cord.
Glarisia nitida (Allem.), comb. nov. Soaresia nitida Allem.
Palestr. Sc. Rio Jan. 142. pi 1, 2. 1857 (i.e., Revista Braz. 1:
209. 1857).
A specimen (Ducke 16606) of this desirable timber tree of Brazil,
the "guariuba," in herb. Delessert (examined by the courtesy of the
Director, Dr. Briquet) was distributed as Clarisia racemosa R. & P.
I judge from the original description and plate that it is the tree
described by Allemao, loc. cit. Certainly it is not the tree of Ruiz
and Pavon, the "tulpay" of Peru, a tree also of value.
Fruiting material of the latter (as well as of C. biflora R. & P.)
is preserved in herb. Boissier of the University, Geneva, and through
the kindness of Professor Chodat I have been able to study it also.
It has rather thin, elliptic, acuminate leaves about two decimeters
long by half as broad, glabrous and smooth above but the lateral
nerves obvious, these with the midrib prominent beneath and shortly
rusty-villous. The fruiting spikes are 2-3 cm. long, the sessile,
16 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
closely crowded, velvety-puberulent, globose fruits, apparently nearly
mature, only 5-6 mm. in diameter. These characters are strikingly
at variance with those of the Brazilian tree with so much smaller,
glabrous leaves and large, glabrous fruits. The latter may resemble
more C. biflora R. & P. with large (2 cm. in diameter), somewhat
verruculose fruits borne on stout pedicels about 5 mm. long. The
very reticulate-veined, lustrous, subcoriaceous leaves of this species
are about 1.5 decimeters long by half as broad, abruptly caudate-
acuminate, and glabrous.
Helicostylis tomentosa (Poepp. & Endl.), comb. nov. Olmedia
tomentosa Poepp. & Endl. Nov. Gen. 2: 32. pi. 145. 1838. H. Poep-
pigiana [Mart.] Trc. Ann. Sci. Nat. III. 8: 134. 1847.
Staminate specimens of this tree are doubtfully distinguishable
from Perebea, and it is very probable that the genus, along with
many others in the family, will some day be regarded as having no
more than an academic interest, treated as sections of a few more
naturally defined groups.
Sorocea Sprucei (Baill.), comb. nov. Pseudosorocea Sprucei
Baill. Adans. 11:296. 1875.
This species was described as glabrous, but two specimens of
the type collection Spruce 4483 from Tarapoto which I have seen
have minutely pubescent branchlet tips and petioles, with more or
less pubescence also on the under leaf surfaces. Except for the
pubescence, it resembles generally S. muriculata Miq.
Pseudolmedia Huberi, nom. nov. Pseudolmedia obliqua (Hub.)
Ducke, Arch. Jard. Bot. Rio 3: 31. 1922, not P. obliqua (Karsten)
Benth. & Hook.
This Amazonian tree, since its original name is already pre-
occupied in the genus, may be rechristened as above, in memory of
the capable Swiss botanist who, in a remarkably short time and
under discouraging conditions, did so much toward the classification
of the plants of the region.
Pseudolmedia laevis (R. & P.), comb. nov. Olmedia laevis
R. & P. Syst. 258. 1798.
The material seen by me is young or imperfect, but the species
seems without question referable to this genus.
Ogcodeia Ulei (Warb.), comb. nov. Acanthosphaera Ulei Warb.
Verh. Bot. Ver. Brandenb. 48: 150. pi. 2. 1907. Naucleopsis Ulei
Ducke, Arch. Jard. Bot. Rio 3: 38. 1922.
PERUVIAN PLANTS 17
Ogcodeia caloneura (Hub.), comb. nov. Olmedia(t) caloneura
Hub. Bol. Mus. Para. 5: 336. 1909. Naucleopsis caloneura Ducke
Arch. Jard. Bot. Rio 3: 38. 1922.
Professor Mildbraed has recently revived and redefined the genus
Ogcodeia Bur. The first of the above species is Peruvian. It may be
distinguished readily from the other species definitely known from
that country by its broad leaves with more numerous nerves (25-30).
The second, Brazilian, suggests 0. Tessmannii Mildbr. of Peru,
but its bracts and young parts are puberulent-tomentose.
Perebea australis (Hemsl.), comb. nov. Castilla australis Hemsl.
in Hook. Icon. 7: pi. 2676. 1901.
Dubiously enough I am transferring this plant, which I have
not seen, from Castilla. Several factors induce me to do so: it is,
presumably, from southern Peru; there is no reason to question
the accuracy of the plate or the description; Castilla is otherwise
unknown from Peru, but Perebea is represented there by species
remarkably similar. The latter genus is distinguishable from the
former particularly by its short style with short stigma, characters
well delineated in the plate of P. australis. Unfortunately only
pistillate inflorescences are known, so the other distinctive character
of Castilla, i.e. the presence of scales among the staminate flowers,
can not be proved; but its style is certainly the typical style of
Perebea, and this genus becomes very weak indeed if this tree can
not be referred there. Accordingly I am transferring it as indicated,
notwithstanding its inclusion by Pittier in Castilla in his revision
of the genus (Contr. U. S. Nat. Herb. 13: 7. 1910).
Phrygilanthus longebracteatus (Desr.), comb. nov. Loran-
thus longebracteatus Desr. in Lam. Encyc. 3: 599. 1792. L. glaucus
R. & P. Fl. Peruv. 3: 45. pi. 275. 1802. L. corymbosus F. G. Dietr.
Vollst. Lexic. Gaertn. Nachtr. 4: 468. 1815-21. P. corymbosus
Eichl. in Mart. Fl. Bras. 5 2 : 46. 1868.
This well-defined Peruvian species has been re-collected (det.
Krause) by Raimondi and Weberbauer, the latter's number 2489
from the department of Ancash under the native name "pupa."
Oryctanthus ovalifolius (R. & P.), comb. nov. Loranthus
ovalifolius R. & P. Fl. Peruv. 3: 50. pi. 177. 1802.
As suggested by Eichler (in Mart. Fl. Bras. 5 2 : 91. 1868), this
can scarcely be referred to 0. florulentus (Rich.) Urban, that is,
0. ruficaulis (P. & E.) Eichler. It is glabrous, simple-stemmed, and
alternate-leaved .
Phthirusa paniculata (HBK.), comb. nov. Loranthus pani-
culatus HBK. Nov. Gen. 3: 442. 1820. L. conduplicatus HBK.
18 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
op. cit. 441. L. Theobromae Willd. ex R. & S. Syst. 7: 132. 1829.
P. Theobromae Eichl. in Mart. Fl. Bras. 5 2 : 56. 1868.
Accepting Eichler's conclusion that the above names refer to the
same species, I take up the earliest and choose paniculatus, as it
describes the diagnostic feature of the plant.
Acrodiclidium limbatum (Nees), comb. nov. Nectandra
limbata Nees, Linnaea 21: 509. 1848. A. limbosa [R. & P.] Mez,
Berl. Jahrb. 5: 89. 1889.
The Nees name is based on a specimen by Tafalla from Tacna,
Peru. Mez refers doubtfully to Nectandra another collection from
Tarma that Nees, hesitatingly, included in his species. In any case,
the Nees name is the first validly published, whether or not restricted
to the Tacna tree.
The species is one of a number illustrated by Ruiz and Pavon
in their Laurograph or in the extremely limited edition of volume
four of their well-known Flora of Peru and Chile. Most of the names
signed to the plates were taken up by later authors as written and
usually before another name had been validly published for the same
plant. There are some instances, however, in which the name was
either changed (as here) when finally published or separate valid
publication made without reference to the Ruiz and Pavon work.
In these latter cases precision if nothing else seems to require the
use of the name first validly published. Accordingly as above 1
accept the name as actually published and not as written by Mez.
Persea Ruizii, nom. nov. P. ferruginea [R. & P.] Mez, Berl.
Jahrb. 5: 154. 1889, not HBK. Nov. Gen. 2: 159. 1817.
Mez erred in publishing for Ruiz and Pavon in Persea a name
that was already in valid use in that genus. To avoid the conflict
he changed the established name(!) to P. Humboldtii Mez.
Ocotea cuneata (Nees), comb. nov. Oreodaphne cuneata Nees,
Syst. Laur. 385. 1836. Ocotea cuneifolia [R. & P.] Mez, Berl. Jahrb.
5:259.1889.
Here is another instance in which the name in use is not the one
first published validly.
Ocotea Poeppigiana (Nees), comb. nov. Oreodaphne Poep-
pigiana Nees, Syst. Laur. 404. 1836. Ocotea multiglandulosa [R. & P.]
Mez, Berl. Jahrb. 5: 280. 1889.
And this is a further example of a name technically never
published having supplanted one proposed validly. The species in
exact character approaches Phoebe, and is one of many that illustrate
PERUVIAN PLANTS 19
the fundamental unity of the four or five groups traditionally accepted
as "genera."
Ocotea sublanuginosa (Nees), comb. nov. Oreodaphne sub-
lanuginosa Nees, Linnaea 21: 515. 1848. Ocotea ovalifolia [R & P 1
Mez, Berl. Jahrb. 5: 261. 1889.
Nees having published this species before Mez took up the name
of Ruiz and Pa von, the former's name, of course, must be used.
Nectandra acuminata (Nees & Mart.), comb. nov. Persea
acuminata Nees & Mart, in Nees, Syst. Laur. 170. 1836. N. acutifolia
[R. & P.] Mez, Berl. Jahrb. 5: 409. 1889.
This shrub or small tree has been collected in Peru near Monzon
by Weberbauer (det. Mez) and in the department of Cajamarca by
Raimondi (det. Schmidt).
The laurels of Peru comprise an interesting group of shrubs and
trees, some of the latter magnificent. When Mez monographed the
family, he did, it seems to me from a rather casual knowledge of the
Peruvian species, a beautiful work, painstaking and, for its day,
conservative. He brought order out of near-chaos and gave what
is in all probability a classification essentially natural. But, as he
himself indicated, the generic lines, especially as concerns Persea,
Phoebe, Nectandra, and Ocotea, are broken by species that in one or
more essential characters do not entirely conform. As more species
become known, more modifications will in all probability be disclosed
and thereby prove more definitely that certain generic characters
regarded as fundamental, such as the degree of development of stami-
nodia and the position and relative position of anther cells, are them-
selves merely variable characters. If such is the case, and in niy
opinion even if it were not, these characters would serve a more
useful purpose and be interpreted more accurately if they were
permitted to define merely sections of a generic unit where they would
still express the apparent group relationships. The fact that there
are no good concomitant characters of fruit or foliage, all species
considered in the genera mentioned, suggests very strongly that the
really natural limitations of the group or groups have even yet not
been correctly defined. Therefore it may be hoped that the next
monographer of the laurels will, even as Mez, break from tradition,
and classify the group on a broader, more practical, and probably
more natural basis. Were the genera not so particularly drawn, the
rather clear vegetative characters of many species, if indeed not of
all proper species, would permit of their determination in any con-
dition. The situation now, to anyone but perhaps a professional
20 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
taxonomist, is open to ridicule: they are among the most conspicuous
trees of the American tropics, but when not in flower, scarcely a
man can venture to state with certainty even a generic name.
Fortunately more and more students believe that classification
of flowering plants can serve a dual purpose, philosophical and
economical, without any sacrifice of the former. When the results
of this belief become manifest, then, and only then, will the raison
d'etre of the science be satisfied. A taxonomy limited in usefulness
is unsatisfactory. The present accepted classification of the American
laurels is a pointed illustration.
Calandrinia ciliata (R. & P.) DC., var. Menziesii (Hook.),
comb. nov. Talinum Menziesii Hook. Fl. Bor. Amer. 1 : 223. pi. 70.
1833. C. caulescens HBK. var. Menziesii (Hook.) Gray, Proc. Amer.
Acad. 22: 277. 1886.
The identification of C. ciliata (R. & P.) DC. Prodr. 3: 359.
1828, has been impossible from the meager description. There are
in herb. Delessert three specimens collected by Pavon that, though
unnamed or not so labeled, seem to me to represent the plant de-
scribed as Talinum ciliatum. Unfortunately, one specimen so labeled
is certainly not that species, as it has a totally different inflorescence
and is quite glabrous. The rest of the material, however, fits the
description, and I can detect no differences between it and C. caules-
cens HBK., as that later described species has come to be interpreted
over its wide range from California to Peru. The more common
form of western North America with more numerous stamens (often)
and larger flowers must, therefore, be transferred to the first specific
name, as indicated above. I think it is more than probable that
Phacosperma peruviana Haw., C. Phacosperma DC., that is, is the
same, in spite of its 13 stamens, as the number of stamens is known
to be variable. Notwithstanding its name, it is doubtfully Peruvian.
Calandrinia crenata (R. & P.), comb. nov. Talinum crenatum
R. & P. Syst. 115. 1798.
This species is apparently unknown except for the original collec-
tion, and except for the crenulate petals it is scarcely distinguishable
from the next.
Calandrinia Ruizii, nom. nov. Talinum polyandrum R. & P.
Syst. 115. 1798, not C. polyandra Benth.
This seems to be related to C. grandiflora, but is smaller in all
its parts. It therefore approaches C. adenosperma, recently segre-
gated by Johnston, but the seeds, though minutely white-hispidulous,
PERUVIAN PLANTS 21
are eglandular. It has been found near Lima, and re-collected near
Arequipa, the type region.
Calandrinia quivensis, spec, nov., annua (vel biennis) glabra;
caulibus gracilibus plus minusye ramosis foliosissimis strictis circa
3 dm. altis; foliis linearibus sessilibus acutis ad 4 cm. longis et 2 mm.
latis; pedunculis aphyllis usque ad 6 cm. longis; racemis simplicibus
vel 1-2-furcatis circa 1 dm. longis; bracteis inconspicuis vix 1 mm.
longis; pedicellis fructiferis 5-6 mm. longis haud valde patentibus
vel interdum suberectis; floribus (circa 6-12) purpureis ut videtur
5 mm. longis; capsula 4-4.5 mm. longa sepalis subrotundatis fere
duplo longiore; seminibus globosis opacis minutissime et obtuse
tuberculatis. Peru: Open rocky slope, Quive, Lima, Pennell 14299
(type, Field Museum).
Apparently this plant is very similar to C. lingulata from the
same region, but it appears to be distinguishable by the conspicu-
ously elongate capsules. There are no perfect flowers in my material.
Portulaca Haughtii, spec, nov., annua ut videtur erecta circa
8 cm. alta, superne plus minusve patenter ramosa, axillis longe
albo-pilosis; foliis sparsis alternis in sicco mediocriter persistentibus
sublinearibus sessilibus vix acutis ut videtur subteretibus glabris
plerumque 10-15 mm. longis, 1-1.5 mm. latis; internodiis 5-15 mm.
longis; capitulis haud dense albo-pilosis (pilis 3-5 mm. longis)
bracteis (foliis superioribus) circa 10 involucratis ad 1 cm. longis
3-5-floris; calyce supra rupturam 3 mm. longo, lobis basi connatis
triangularibus acuminatis rubescentibus; corolla flava vix ad 4 mm.
longa, petalis oblongo-ellipticis, abrupte acutis; staminibus circa 10;
stylo supra medium quadrifido; capsula globosa, vix 2 mm. diametro,
breviter stipitata paullo infra medium circumscissa; seminibus brun-
neis 0.35 mm. diametro, acute et minute tuberculatis, tuberculis
elevatis circumcirca vix manifeste stellatim radiantibus. Peru:
North of Parinas Valley, Piura, March 25, 1929, Oscar HaughtF-180
(type, Field Museum).
Among Peruvian species this Portulaca keys to P. lanuginosa,
which has procumbent stems, shorter leaves, and many stamens.
Species other than Peruvian with brown seeds and yellow flowers
that have come to my notice all seem, from description, to be obvi-
ously different from this plant, with the possible exception of P.
Milleri of Margarita, which has more numerous leaves and stamens
and a six-parted style.
Portulaca tingoensis, spec, nov., annua erecta vel suberecta
basi ramosa 5-10 cm. alta; ramis simplicibus plus minusve rubes-
centibus in sicco fere 2 mm. crassis, apice (sub involucre) valde
incrassatis; internodiis 1-2 cm. remotis; foliis alternis inferioribus
deciduis laxis fere ignotis sed ut videtur pblongo-spathulatis subtere-
tibus circa 1 cm. longis, superioribus similibus paucis ad 2 vel aliquid
22 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
3.5 mm. latis; floribus ignotis; capsulis plerumque capitato-confertis
terminalibus sed etiam solitarip-axillaribus cum pilis sordide albis
5-7 mm. longis dense involutis, 2-3 mm. longis medium versus
circumscisse dehiscentibus; seminibus circa 0.6 mm. diam. atro-
opalescentibus obtuse stellato-tuberculatis. Peru: Arequipa, open,
sandy and rocky slopes, Tingo, Pennell 13111 (type, Field Museum).
A plant of dubious specific standing, evidently allied to P. pilosa
and its numerous variants. However, as far as other Peruvian speci-
mens are concerned, its metallic-lustrous seeds and its habit distin-
guish it readily enough.
Gynandropsis gracilis (Tr. & PI.), comb. nov. Cleome gracilis
Tr. & PI. Prodr. 1:74. 1862.
Gynandropsis macrothyrsis (Tr. & PI.), comb. nov. Cleome
macrothyrsis Tr. & PL Prodr. 1: 72. 1862.
Gynandropsis puberula (Tr. & PL), comb. nov. Cleome
puberula Tr. & PL Prodr. 1: 71. 1862.
It is entirely apparent from a reference to the above-cited work
that the authors did not regard Gynandropsis as a genus distinct
from Cleome, and the fact is emphasized by their own transfer of
species described in the former to the genus Cleome. To accredit the
above names to Gynandropsis with Triana & Planchon as authority
is therefore an error that the authors could not sanction. I think,
however, that the genus is acceptable and express this opinion in
the case of the above species, which have come to notice in the course
of study of some related Peruvian forms, and the type collections of
which I have seen.
Echeveria Harmsii, nom. nov. Oliverella elegans Rose, Bull.
N. Y. Bot. Gard. 3: 2. 1903, not E. elegans Rose, 1905. Cotyledon
elegans N. E. Br. Bot. Mag. 131: pi. 7993. 1905. E. elegans Berger,
Pflanzenf. ed. 2. 18a: 472. 1930.
As I have had occasion to point out at least once in nearly every
paper published by me, the transfer of a specific name into a genus
that already contains the same name is not in accord with accepted
nomenclatorial custom. The plant Oliverella elegans, therefore, when
treated as an Echeveria, requires another specific designation because
there is already an Echeveria elegans. This seems to be a nomen-
clatorial maneuver which is so simple, sensible, and natural that
it is liable to be followed always, supported by botanical opinion
regardless of what future law-tinkerers may decide to do about it.
With some confidence, therefore, I make this particular name change,
and with much pleasure select a new name for this beautiful plant.
PERUVIAN PLANTS 23
Escallonia Atahuallpae, spec, nov., arbor 10-15 m. alta; ramis
brunneis adultis nudis valde in laminas papyraceas exfoliatis, ramulis
novellis granule- tomentulosis; foliis lanceolatis basi apiceque cuneato-
attenuatis acutis minutissime denseque denticulatis glabris vel
glabratis sessilibus (vel interdum in petiolum breviter decurrentibus)
demum circa 10 cm. longis, 2.5 cm. latis, saepius minoribus; racemis
pendulis terminalibus simplicibus circa 1 dm. longis parce glandulosis;
bracteis setaceis contortis 3-5 mm. longis; pedicellis circa 4 mm.
longis; calycis laciniis triangularibus acutis vix 2 mm. longis; petalis
erectis oblongis 3 mm. longis; receptaculo cupulato; stylo 3 mm.
longo, stigmate peltato; antheris fere 2.5 mm. longis filamenta per-
brevia multo superantibus; capsula ignota. Peru: Gravelly river
valley, Tambo de Pariocpta, Ancash, Oct. 8, 1922, Macbride &
Featherstone 2541 (type, Field Museum).
The generic character of Escallonia calls for elongate slender
filaments; in the species described above they are very much shorter
than the conspicuous large anthers. If it were not for the fact that
two botanists of reputation have kindly verified my reference of the
plant to Escallonia, I should be inclined to think that again I have
described a species in the wrong genus. The tree simulates E. pendula
in aspect, but the narrower leaves are glabrous and the flowers are
considerably smaller.
Whether King Atahuallpa murdered his brother or merely
defended himself, his famous name may appropriately be perpetuated
in the botany of the land of the Incas. Anyway, it may be mentioned,
for the pleasure of the cynics, that it would not be the first time that
a plant has been named for a reprobate.
Desmodium immerens, spec, nov., erectum vel fere erectum
ad 6 dm. altum; caulibus superne petiolis foliisque plus minusye
pilosis vel interdum glabratis; stipulis liberis longe acuminatis;
petiolis 3-4 cm. longis; foliolis 3 late ovatis vel rotundato-ellipticis
vel ovato-lanceolatis, apice rotundatis vel rarq acutis semper apicu-
latis; racemis laxis 1-2 dm. longis; pedicellis fere 2 cm. longis;
floribus pallide rubro-purpureis 8-9 mm. longis; calycis labio supe-
riore integro ovato vix 3 mm. longo; vexillo exauriculato late obovato
emarginato ad basin late cuneato sed haud unguiculato medio 6 mm.
lato, 8 mm. longo; alis carinam subaequantibus circa 8 mm. longis,
3 mm. latis; leguminibus breviter stipitatis, leviter arcuatis, sutura
superiore definite, inferiore profunde sinuata; articulis plerumque
5 prehensili-pubescentibus 8 mm. longis, 5 mm. latis. Peru:
Muna, Dept. of Huanuco, Macbride 4000 (type, Field Museum).
Yanano, Dept. of Huanuco, Macbride 3725.
Not a Meibomia, as restricted by Schindler, because of the sessile,
not at all clawed standard, nor a Nephromeria, because of the numer-
ous pod segments, this plant must be considered a true Desmodium,
24 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
and in that group it seems to approach closely only D. affine Schl., the
pods of which are exactly straight on the upper edge. The existence
of these collections, which, incidentally, have caused me no end of
trouble for which only my own limitations may be blamed, appears
to support the opinion expressed in a former paper that Desmodium
may rightly include the groups Meibomia and Nephromeria as
sections.
Parosela Sawadae, spec, nov., fruticosa ut videtur erecta sed
diffusa et laxe ramosa; ramulis puberulis demum glabris parce glandu-
losis; stipulis setaceis glabris; foliolis plerumque 11, petiolatis ellip-
ticis 4-5 mm. longis, 2-3 mm. latis, supra glabris subtus puberulis et
plus minusve pallide glanduloso-punctatis; spicis longe pedunculatis
vel interdum subsessilibus demum 4-8 cm. longis; bracteis glabris
paullo et obscure punctatis persistentibus caudato-acuminatis 4-6
mm. longis; calycis dentibus inaequalibus basi ipsa ovatis superne
fere spinuloso-subulato-apiculatis tubo dense villoso et valde nervato
subaequilongis ad circa 2 mm. longis; floribus circa 8 mm. longis;
carina alisque ut videtur albis, vexillo purpureo. Peru: Huanuco,
Sawada Pll (type, Field Museum). San Rafael, Dept. of Huanuco,
Sawada Pi 12.
Mr. Sawada's continued interest in the flora of Huanuco is
hereby, with pleasure, given merited recognition.
As specific lines are drawn in this genus at present, it is necessary
to add another name to take care of this diverse form. It is P. nova
as previously interpreted by me, at least in part, but it is not Dalea
nova Ulbrich, a species similar in foliage and pubescence but with a
very different calyx, the teeth minute.
I notice that some one has proposed that Dalea supplant Parosela
by adding the former name to the list of nomina conservanda. This
proposal is admission that the question as to which name is really
"correct" under the Rules can not be determined clearly enough to
make the decision satisfactory to every one. Therefore one name
or the other must be "conserved" and, in the spirit of the Rules of
course the one to choose, other considerations, as here, being about
equal, is the one under which fewer species have been described.
One of the factors that could be employed in making such a decision
is commonly known as "common-sense."
Astragalus Dielsii, spec, nov., subacaulis dense caespitosus 2-3
cm. altus sericeo-pilosus demum glabratus; stipulis yaginantibus
valde imbricatis persistentibus; fpliis 8-10 mm. longis, 5-7-jugis,
foliolis confertissimis vel demum distinctis sessilibus obovatis retusis
vix 1 mm. longis; floribus solitariis brevissime pedicellatis 8 mm.
longis pallide caeruleis; bracteis membranaceis circa 1 mm. longis;
PERUVIAN PLANTS 25
calycis dentibus anguste triangularibus circa 1.5 mm. longis tubo
oblpngo-campanulato fere duplo brevioribus, dense cum pilis sericeis
albis et nigris intermixtis pubescentibus; vexillo subrotundato 6-7
mm. lato; alls 3-4 mm. longis; legumine subovoideo, 3 mm. longo,
2 mm. lato, apiculato chartaceo glabro (vel fere) dorso sulcato
biloculari, loculis ut videtur monospermis. Peru: Yanashallos, west
of Huallanca, Dept. of Ancash, over 5,000 m., Oct. 2, 1922, Mac-
bride & Feather stone 2479 (type, Field Museum).
Peruvian Astragali are either extremely local or variable; few are
completely known and many are misinterpreted because compared
with misdetermined material. This tiny plant appears to be truly
comparable to only two Andean species (which, it happens, are among
the few not seen by me) : namely, A. alpamarcae Gray and A. casapal-
tensis Ball. From the former it differs, as to description, by the fewer
and smaller leaflets that are densely and equally pubescent on both
sides until in age equally glabrate on both surfaces, and by the
narrower, longer, and densely pubescent calyx tube; from the latter
by the distinctly fewer leaflets and by the shorter calyx teeth in
proportion to the tube.
This silvery-pubescent, closely tufted little plant grew without
competition on loose stony slopes, and when collected its pale blue
flowers were all but hidden by a light fall of snow.
Astragalus Pilgeri, spec, nov., suffrutex humilis laxe caespitosus
subglaber; caulibus diffusis 1^4 cm. longis; foliis viridibus minutissime
parceque hispidulis 4-8-jugis (plerumque 6) circa 1.5 cm. longis,
foliolis 2-3 mm. longis obovatis retusis conduplicatis; stipulis parvis
solum ad basin connatis; floribus 7 mm. longis perbreviter racemosis
ad apicem pedunculi axillaris folio interdum multo vel interdum
paullo brevioris; calyce parce pilis brevibus nigris pubescente 2-2.5
mm. longo, laciniis triangulari-subulatis quam tubo campanulato
distincte brevioribus; vexillo subrotundato fere 5 mm. lato; ovario
circa 5-ovulato. Peru: Hacienda Chuchapaya, Valle del Paucar-
tambo, 3,800 m., Feb., 1929, Herrera 2315 (type, Field Museum).
Perhaps nearest A. micranthellus Wedd. among species recorded
from Peru, but very distinct in its green, retuse leaflets, free stipules,
and larger flowers. It has a superficial resemblance to A. Hieronymi
Ulbr., but the smaller flowers and short, triangular calyx teeth readily
separate it.
Astragalus salubris, spec, nov., pusillus dense caespitosus
subpatenter canescenti-strigosus vel subyillosus; caulibus ad 2 cm.
longis; stipulis vaginantibus valde imbricatis; foliis numerosissimis
longe (circa 1 cm.) petiolatis, plerumque 3-4 cm. longis, erecto-
patentibus 10-12-jugis; foliolis ellipticis obtusis vel paullo retusis
circa 3 mm. longis, fere 1.5 mm. latis conduplicatis supra glabris;
26 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
spicis subsessilibus 6-8 mm. longis; floribus 4-6 ut videtur purpureis
vix 6 mm. longis; calycis nigro-pubescentis fere 4 mm. longi dentibus
subulatis tubo suboblongo distincte brevioribus; vexillo late obovato
3.5 mm. lato; ovario biloculari, loculis monospermis. Peru: Cerro
de Pasco, Oct. 28, 1927, Sawada P86 (type, Field Museum).
Very possibly this thrifty, hardy form should be included in one
of the several Andean species that are all so similar in general and yet
not quite alike. Its subspreading pubescence, however, separates it
from most species. In many respects it resembles A. Brackenridgei
Gray, but that, typically, at least, has larger, oblong leaflets and
longer, peduncled spikes with more flowers. The calyx bract of A.
salubris is white, rather than black-pubescent.
Pithecolobium Merrilli, nom. nov. P. multiflorum Merr.
Phil. Journ. Sci. Bot. 10: 11. 1915, not Benth. in Hook. Lond. Journ.
3: 320. 1844.
The earlier use of the name multiflorum by Bentham for a Brazilian
species in this genus is not recorded in the Index Kewensis. The
Philippine plant described by Merrill may most appropriately bear
the name of its well-known author.
Acalypha Hookeri, nom. nov. A. cordifolia Hook. f. Trans.
Linn. Soc. 20: 186. 1847, not Griseb. Abh. Ges. Wiss. Gott. 24: 60.
1879.
Pax, contrary to usual botanical nomenclatorial practice, has
retained the synonymous name A. flabellifera Rusby for the Grisebach
species of Argentina and Peru. It is, of course, the more recently
described A. cordifolia of the Galapagos Islands that requires another
name.
Briquetina, gen. nov. Calyx breviter 4-5-lobus. Petala 5,
yalvata, libera, oblonga, lamina medio incurva et inflexa, laevigata,
intus linea longitudinal! elevata praedita et etiam ad medium trans-
verse incrassata. Stamina 5 petalis paullo breviora et eis alterna,
filamentis superne valde dilatatis glabris; antherae erectae, loculis
qvato-oblongis, connective haud crassq. Discus nullus. Ovarium
liberum, 1-lpculare; stylus brevis, conicus, stigmate subdiscoidep;
ovula 2 ut videtur vulgo 1 evolutum, pendulum, altero ante anthesin
abortiente. Fructus drupaceus subglobosus, endocarpio lignoso.
Arbores. Folia integerrima, coriacea, penninervia. Flores parvi
spicati, spicis brevibus in racemos elongates dispositis.
Briquetina incarum, spec, nov., arbor magna; ramis teretibus
glabris, cortice pallide brunneo nitidulo longitrorsum rimoso; ramulis
pedunculisque glabratis vel sparse puberulis; petiolis crassis, 7-10
mm. longis; foliis fere ellipticis vel late ovato-ellipticis, basi plus
minusve obliquis, apice breviter obtuseque acuminatis, plerumque
circa 1.5 dm. longis et 6-8 cm. latis, supra paullo nitidulis, subtus
PERUVIAN PLANTS 27
opacis glabris vel interdum cum pilis crispis brevibus pilosis, tenuis-
sime cartilagineo-undulato-marginatis; nervis medio lateralibusque
primariis supra vix notatis, subtus cum venis transversis yalde con-
spicuis; inflorescentiis 5-13 cm. longis, laxis, angustis, fructiferis raro
2 cm. latis, ramulis 0.5-1 cm. longis dense floriferis valde recurvato-
incurvatis; floribus sessilibus flavo-viridibus satis fulvo-pilosis; petalis
glabris minutissime papillosis suboblongis circa 1.25 mm. longis ad
medium definite angustatis et margine inflexis; nervis (vel lineis)
mediocriter prominentibus vel elevatis ad medium laminarum incras-
sato-contractis sed appendicibus destitutis; filamentis glabris, late
cuneatis; ovario ovoideo; drupa ellipsoidea obtusa circa 1 cm. diam.
Peru: Mufia, Dept. of Huanuco, about 2,200 m., May 23-June 4,
1923, Macbride 4050 (type, Field Museum). Yanano, Dept. of
Huanuco, about 2,000 m., May 13-16, 1923, Macbride 3748.
It is rather astonishing if this handsome tree, which I found grow-
ing on open rocky hills in central Peru, is really undescribed. It is
quite possible that it has already received a name in some family
other than the Icacinaceae in which I place it, especially in view of
the fact that this family for a long time was a part of another now
generally regarded as far removed. But its characters, kindly verified
for me by Dr. Briquet, certainly seem to forbid its reference to any
other group, and in this family it is not referable to any genus as
described. The feature most obviously at variance is probably the
inflorescence, which is unusually open and elongate for the family,
the flowers more generally being borne in rather close cymes, often
panicled. Perhaps it is nearest Poraqueiba Aubl. but the petals in
that genus are ciliate-appendaged across the middle; Mappia Jacq.
has pubescent petals and filiform filaments; Kummeria Mart,
pubescent filaments; and so the list could be extended to include also
the new genera described recently for the Amazon region. In
addition to these floral differences which traditionally separate
genera in the family, there is always for our plant the distinctive
inflorescence.
So it has seemed necessary to propose a new genus for it; and in
choosing the name Briquetina I give myself the pleasure of honoring
one to whom botanical honor is so abundantly due.
Paullinia Williams!, spec, nov., scandens fruticosa subglabra;
ramis trigonis leviter vel obsolete striatis; foliis 5-foliolato-pinnatis;
petiolis 2-5 cm. longis, 4-sulcatis vel interdum anguste marginatis,
parce pulverulentis et ciliato-hirsutulis; rhachidibus conspicue alatis
5 mm. latis; stipulis prominentibus lineari-lanceplatis acuminatis,
7-17 mm. longis; foliolis ovato-lanceolatis subsessilibus vel breviter
petiolatis, acute acuminatis, subintegris vel plerumque praesertim
ad apicem remote 3-5-dentatis, laxe reticulato-venosis, utrinque
28 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
opacis subtus minutissime punctatis, 3-3.5 cm. latis, 8-10 cm. longis;
inflorescentiis solitariis, 2 vel 3 cm. longis, molliter puberulentis,
ramulis brevissimis, 3-5-floris; pedicellis ad 3 mm. longis; floribus
circa 4 mm. longis; sepalis minutissime pulverulentis exterioribus
1.5, interioribus 3 mm. longis; petalis ad basin pilosis. Peru:
Maquisapa on the upper Rio Nanay, Loreto, July, 1929, L. Williams
1196 (type, Field Museum).
In the absence of fruit the relationship of this vine is unknown,
but in aspect it resembles strikingly P. laeta and P. subauriculata,
from both of which it nevertheless is at once distinguishable because
of the acutely acuminate leaflets, the large stipules, and the tri-
angular stems.
Serjania elongata, spec, nov., fruticosa floribus exceptis glabra;
ramis teretibus obscure striatis nitidis; petiolo communi 6-8 cm.
longo; foliis 5-foliolato-pinnatis; foliolis integris vel interdum obscure
1-3-undulato-dentatis, late ellipticis, lateralibus breve petiolatis basi
apiceque obtusissimis vel rotundatis, terminalibus similibus sed ad
basin cuneato-contracto-petiplatis, circa 6 cm. latis et 10 cm. longis,
utrinque satis reticulatis nitidulis chartaceis, punctis vel lineolis
pellucidis minutissimis et vix notatis; petiolis omnibus nudis; inflores-
centiis racemiformibus solitariis tenuibus longe (6-8 cm.) peduncu-
latis, 2.5-3 dm. longis, laxifloris glabris vel fere glabris; ramulis
inflorescentiarum 2-4 mm. longis; floribus circa 4-fasciculatis, pedicel-
lis gracilibus vix 2 mm. longis; sepalis exterioribus glabris circa 2 mm.
longis, interioribus paullo tomentulosis, 2.5 mm. longis; petalis glabris
(intus ad basin paullo barbatis) obovatis circa 3 mm. longis; fila-
mentis breviter pilosis. Peru: In sunny brush ; flowers cream-colored;
La Merced, Junin, Aug., 1923, Macbride 5511 (type, Field Museum).
Without even young fruit, the generic position of the vine is open
to question. I think it is probably a Paullinia, but it resembles
not at all any species in the nearly complete collections assembled
at Munich by Radlkofer. The apparent absence of tendrils and the
subpinnate leaves suggest that genus more than Serjania, but there
is an almost exact simulation in the leaflets of S. foveolata and a close
approach to those of S. nutans, both species with ternate or biternate
leaves.
Gustavia caballoensis, spec, nov., arbor; foliis ut videtur ad
apicem ramulorum congestis subsessilibus (petiolis 1-3 mm. longis)
chartaceo-coriaceis glabris plus minusve praesertim ad apicem undu-
lato-crenulato-serratis oblongo-elliptico-oblanceolatis inferne arcua-
tim vel interdum cuneatim gradatim reductis, circa 3 dm. longis,
basi ipsa subrotundata circa 1.5-2 cm. latis, supra medium 8-12 cm.
latis, apice breviter acuminatis obtusis; nervis lateralibus circa 12
supra et subtus mediocriter prominentibus ut etiam venulis tenuis-
sime reticulatis; floribus solitariis axillaribus; pedicellis demum 2.5
cm. longis ad basin 1-3-bracteolatis, ad apicem bibracteolatis;
PERUVIAN PLANTS 29
bracteolis rotundatis 2-4 mm. longis; calycis pulverulent! margine
integro vel obscure crenulato; petalis ut videtur subaequalibus late
obovatis vel obovato-oblongis 3-3.5 cm. longis pulverulentis; antheris
vix 3 mm. longis; fructus subglobularis apice truncatus circa 4 cm.
diam. Peru: Caballo-cocha, Loreto, Aug., 1929, L. Williams 2240
(type, Field Museum), 2152, 2236(1}.
The last collection, in bud and with a section of an immature
fruit (from which the description is drawn), has longer leaves with
about 20 lateral nerves.
This material very doubtfully represents a new species, but
rather than modify the character of any one of the several described
forms to which it could thereby be referred, and so perhaps cause
confusion, I give it a name. It bears a striking resemblance to G.
Marcgraviana, G. Ulei, and (presumably from description) G. Poep-
pigiana. These three species, however, all have terminal flowers,
more or less evidently racemose. The first, with exactly the foliage
of this, grows in eastern Brazil, and its terminal flower or flowers
seem to have slightly longer anthers. The second and third, which
are not clearly distinct from each other, have subsessile or shortly
pediceled, terminal flowers. In this connection G. insignis Linden
ex Hook. Bot. Mag. 84: pi. 5069. 1858 may be mentioned. Miers in
his treatment of the genus, Trans. Linn. Soc. 30: 177. 1874, refers it
to G. superba Berg. The latter, however, has long-petioled leaves.
The former has the subsessile leaves of our plant but they are
"spinulose-serrate," and the calyx is somewhat lobed; otherwise it
approaches G. caballoensis. No Gustavia already collected in Peru
is comparable to the one here described, with the possible exception
of G. longifolia, which has the pedicels bracteate below the middle.
Gustavia mangua, spec, nov., arbor(?); ramulis ignotis; petiolis
5-10 mm. longis; foliis oblongo-lanceolato-oblanceolatis in petiolum
longe sensim angustatis basi ipsa acutis, supra medium gradatim
latioribus plerumque 6-8 cm. latis, sursum angustioribus et in acumen
acutum anguste productis, 3-4 dm. longis, rigide chartaceis, margine
subintegris vel distanter crenulato-serrulatis, valde crispato-plicatis;
nervis tenuibus supra vix prominulis subtus stramineis et prominenti-
bus glabris, venis transversis tenuissime reticulatis; ut videtur
cauliflora; racemis circa 2.5 cm. longis multifloris; bracteis numerosis
ovato-acutis 3 mm. longis; pedicellis gracilibus minute pulverulentis
paullo sub medium bibracteolatis, 3-3.5 cm. longis; calycis margine
integro vel obsolete 4-lobato; petalis circa 8, oblongo-obovatis circa
' 2.5 cm. longis 7-10 mm. latis, pulverulentis; antheris 2 mm. longis.
Peru: Lower Rio Nanay, Loreto, May-June, 1929, L. Williams 281
(type, Field Museum).
30 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
"Mangua," the native name, may serve also as the specific one
for this tree which, among Peruvian species at least, resembles most
G. augusta, from which its almost caudate-acuminate leaves and
small flowers readily distinguish it. Its character seems to be
definitely at variance with that of any species as described by Miers
in his treatment of the genus, Trans. Linn. Soc. 30: 175-188, 1874,
as well as with any of the few proposed since that date. Miers' work,
however, is not entirely intelligible to me, and I do not know
how much reliance can be placed in his alignment.
Grias Neuberthii, spec, nov., arbor(?) ; ramis ramulisque ignotis;
foliis ut videtur sessilibus oblongo-spathulatis maximis breyiter
acuminatis, sensim ad basin angustatis, supra medium latissimis,
majoribus ad 8 dm. longis et 1.5 dm. latis, integris vel leviter undu-
lato-crispatis glabris paullo nitidulis subcoriaceis; nervis lateralibus
circa 25 plerumque alternantibus supra vix notatis subtus valde
prominentibus, venis utrinque mediocriter prominulo-reticulatis;
racemis (e trunco nascentibus?) ut videtur ad 1 dm. longis cinereo-
tomentulqsis; bracteis 1 cm. longis; pedicellis subangulatis crassis
bracteis vix longioribus; bracteolis calycibusque sericeo-tomentulosis;
calycibus bibracteolatis (bracteolis oblongis pbtusis 8 mm. longis)
lobis 4 subrotundatis circa 7 mm. longis; petalis 4 crassis subovalibus
circa 3 cm. longis; filamentis crasso-angulatis; antheris oblongis,
circa 1 mm. longis; thecis connatis longitrorsum dehiscentibus.
Peru: Caballo-cocha, Loreto, Aug., 1929, L. Williams 2337 (type,
Field Museum).
This tree is immediately distinct from all of the few species
described in the genus Grias by its inflorescence and connate anthers,
but because it is aberrant in character, it is possible that it has been
referred to Gustavia, most of the species of which are unknown to me.
The latter genus, however, has anthers dehiscent by terminal pores,
but the cells are connate in the manner of those of this plant. As the
androecium is equally expanded, there is no choice of genera, granted
that the plant has been placed in the right family, except these two,
and as the method of anther dehiscence is the diagnostic character for
the genera, there appears to be no reason for including it in Gustavia.
The material, unfortunately, is poor, the inflorescence being badly
broken, and its attachment not indicated. The leaves, broken off
above the base, are similar, except for being subopaque, to those of
Grias grandifolia, a species with glabrate, small-bracted inflorescence
and subglobose, separate anthers, as typical for the genus. The
native name, "chope," has been recorded also for species of Gustavia.
In naming this interesting species for Carl Neuberth, Custodian
of the Herbarium of Field Museum for more than thirty years, well-
merited recognition is given a conscientious botanical worker.
PERUVIAN PLANTS 31
Combretum Llewelyni, spec, nov., fruticosum (scandens?);
ramulis foliis subtus petiolisque parce rufo-lepidotis; petiolis circa
5 mm. longis; foliis supra glabris nitidis subchartaceis late ellipticis
vel fere subrotundatis, apice abrupte caudato-acuminatis, 8-10 cm.
longis, 5 cm. latis; nervis lateralibus 6-^-7, supra mediocriter subtus
alte prominentibus ad marginem curvatis et cum venis plus minusve
definite reticulatis; spicis terminalibus circa 2 dm. longis; calyce
17 mm. longo rubidp-lepidoto e basi anguste campanulata vix yel
haud constricto, lobis anguste triangularibus basi vix 3 mm. latis,
5-6 mm. longis, acuminatis intus parce lanato-villoso; petalis fere
linearibus circa 2 mm. longis; staminibus calycem circa duplo
superantibus; samara ignota. Peru: Along Rio Itaya, Loreto, Llew-
elyn Williams 143 (type, Field Museum).
Only C. Loeflingii Eichl. and C. Baslerianum Mildbr. have calyces
that even approach the exceptionally elongate, narrowly toothed
ones of this attractive shrub. C. Llewelyni also has narrower petals
and much broader leaves.
Gunnera peruviana, spec, nov., maxima; ligulis linearibus,
remote spinuloseque dentatis; petiolis circa 3 dm. longis foliisque
subtus haud spinescentibus sed plus minusve cum pilis spinuliformi-
bus pilosis demum glabris; foliis supra conspicue bullato-rugpsis
valde scabris vix tuberculatis rotundato-cordatis minoribus circa
3 dm. latis et plerumque prostratis vel adscendentibus, majoribus
erectis magnificisque perbreve 5-7-lobatis, lobis late rotundp-ovatis
fere integris solum repando-undulatis et minutissime denticulatis;
inflorescentiis sessilibus dense spinulpso-pilpsis circa 3.5 dm. longis
fere oblongis, ramis remotis (rare attingentibus) patentibus plerum-
que 5-6 cm. longis; bracteis persistentibus conspicuis 1-2 cm. longis
irregulariter spinulpso-dentatis glabris anguste oyato-lanceolatis
longe acuminatis apice haud vel vix dilatatis; sepalis haud notatis;
ovario subgloboso. Peru: Tambillo, Dept. of Huanuco, May 8,
1923, Macbride 3583 (type, Field Museum).
So many Gunneras have been described which, at least from
description, appear to be very similar that one would suppose there
must already be a name for every form. But this seems not to be
the case and, if there is any constancy in the described characters,
it becomes necessary to give new names to the above and following
plants. G. peruviana is more than likely G. scabra R. & P. as to the
Muna, Peru, specimen cited by the authors but not as to description,
plate, or the Chilean material, which is G. chilensis Lam. The
latter differs from this plant in its dense inflorescence and deeply,
narrowly lobed and acutely serrate leaves. Also, the related G.
Berteroi Phil. andG. manicata Lind. differ fromG. peruviana decidedly
in their leaf dentation. Yet other forms, as G. rheifolia Schindl. and
G. commutata Blume, are otherwise distinct enough, as, for instance,
32 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
in their dense inflorescences. If the sepals could be determined as
persistent, a comparison could be made with G. pilosa Kunth, G.
boliviano, Morong, and G. apiculata Schindl., but the spines and pilos-
ity of the first are lacking, as are the smooth leaves of the second,
and the acutely lobed foliage and dense inflorescence of the third.
Since Schindler monographed the family for Pflanzenreich he has
described G. Margaretae and G. Annae, both from Peru but both
ebracteate or nearly so and with narrowly or acutely lobed leaves.
Gunnera Bolivari, spec, nov., ut videtur maxima; ligulis igno-
tis; petiolis inflorescentiisque ubique rubro-purpureis glabratis
vel glabris; foliis haud profunde obtuseque 5-7-1 obatis et obscure
repando-dentatis supra viridibus glabris et laevigatis subtus palli-
dioribus solum ad nervos parce tenuissime pilosis, ut videtur magnis;
inflorescentiis ut videtur ebracteatis (bracteis minutis) paullo attenu-
atis ad apicem longe (fere 1 dm.) pedunculatis circa 3.5 dm. longis,
8-10 cm. latis; ramis yalde patentibus numerosis sed haud congestis
gracilibus fructiferis vix vel paullo incrassatis, plerumque 5 cm. longis;
sepalis aliquid persistentibus; ovario ovoideo. Peru: Cerro de Cusi-
lluyoc, Dept. of Cuzco, Pennell 13989 (type, Field Museum).
Apparently very distinct from known species because of its
glabrous, essentially ebracteate inflorescence and glabrate leaves,
with the possible exceptions of G. pyramidalis Schindl. of Juan
Fernandez, which is more pubescent and has a very pyramidal
inflorescence, and G. Margaretae Schindl., which has narrowly lobed
leaves.
Bowlesia flabilis, spec, noy., annua suberecta glabrata 2-2.5
dm. alta; caulibus foliisque viridibus sed (praecipue foliis subtus)
parce cum pilis plus minusve stellato-ramosis pubescentibus; petiolis
(inferioribus) circa 2 cm. longis; foliis circa 1.5 cm. longis, late ovatis
vel subrotundatis, basi cprdatis, late 5-lobatis, lobis integris mucro-
nulato-acutis; foliis superioribus valde sed gradatim reductis, 3-lobatis
vel subintegris, circa 1 cm. longis vel bracteiformibus et 3-5 mm.
longis; inflorescentiis sessilibus simplicibus 1-4-floris; fructu nudo
sed valde glochidiato-ciliato fere 2.5 mm. longo et circa 2 mm. lato.
Peru: Wood-clearing, Mito, Huanuco, July 23-Aug. 14, 1922, Mac-
bride & Feather stone 1718 (type, Field Museum).
Among Bowlesias with glochidiate fruits this resembles most B.
tenella, from which its 3-lobed or subentire and greatly reduced upper
leaves distinguish it. Its name refers to its habit of growth, to its
habitat, and to its specific standing. The last characteristic, however,
is shared with most species of Bowlesia; quite possibly they are all
reducible to about two.
Hydrpcotyle Dahlgreni Rose & Macbr., spec, nov., caulibus
pedunculis petiolisque plus minusve rufo-hirsutis; caulibus 1-2 mm.
PERUVIAN PLANTS 33
crassis; petiolis 2-3 cm. longis, lamina reniformi-cordata vel suborbi-
culata usque ad 5 cm. lata lobulato-crenata, lobulis truncatis baud
prominentibus, vel manifeste lobata, supra parce adpresseque hir-
sutula, subtus tenuiter strigillosa; pedunculis circa 1 dm. longis;
inflorescentiis simpliciter umbellatis multi (60-80) -floris, circa 2 cm.
diam.; pedicellis filiformibus 6-8 mm. longis; petalis pvatis, 1 mm.
longis; fructu circa 1.5 mm. crasso orbiculato-reniformi. Peru:
On moist sunny banks at Cueva Grande near Pozuzo, Huanuco, June
23, 1923, Macbride 4776 (type, Field Museum).
Because of the conspicuous, long, coarse pubescence this plant
at once suggests H. Barbarossa Cham., which, however, has peltate
and lobed leaves. Among species with the leaves not peltate few
seem near enough in character to merit comparison, but perhaps H.
callicephala Cham, with much longer petioles and lobed leaves and the
recently discovered H. tambolomensis Wolff with much finer pubes-
cence and lobate leaves are related.
Dr. Rose proposed naming this plant for the collector. I have,
therefore, felt it my privilege to change the name and share in the
publication. Apparently an unusually distinct species, it gives me
pleasure to name it for a botanist who is first a scientist.
Weigeltia nanayensis, spec, nov., fruticosa pumila; ramulis
glabris praeter apicem minute parceque rufo-lepidotis; petiolis 1.5-2
cm. longis; foliis elliptico-lanceolatis basin versus gradatim attenua-
tis, apice plerumque acuminatis, circa 1.5 dm. longis, 5 cm. latis,
chartaceis utrinque prominulo-reticulatis; punctulis paucis subtus
vix notatis; racemis densifloris circa 2 cm. longis; pedicellis bracteolis
multo brevioribus circa 2 mm. longis; floribus 4-meris vix 2.5 mm.
latis; lobis calycinis ovato-acutis plus minusve dense nigro-punctatis
ut etiam petalis duplo longioribus obtusis; staminibus petalis sub-
aequalibus, filamentis alte insertis gracilibus, antheris subrotundatis
haud emarginatis vel punctulatis dorso medio affixis. Peru: Lower
Rio Nanay, Williams 658 (type, Field Museum).
A specimen in bud from the same locality (657) is probably the
same. It has broader leaves (to 6.5 cm.) and erect inflorescences
about 5 cm. long. The species is rather aberrant in Weigeltia, but its
regularly lobed calyx and its roundish, dorsally fixed anthers seem
to forbid referring it to any other genus, though the superficial
resemblance to Cybianthus psychotrifolius Rusby is striking. Perhaps
only one genus is concerned, as at one time considered. Even so,
this shrub is distinguishable from Rusby's by its very thin, reticulate,
sparsely punctate foliage.
Stylogyne amplifolia, spec, nov., arbuscula glabra; foliis ellip-
ticis vel oblongo-ellipticis breviter acuminatis basi late cuneato-
acutis plerumque 1 dm. latis et 2.5-3 dm. longis, integerrimis vel
34 FIELD MUSEUM OF NATURAL HISTORY BOTANY, VOL. XI
leviter undulato-crenulatis, chartaceo-coriaceis supra aliquid nitidis
venis paullo notatis, subtus densissime reticulato-venosis, utrinque
plus minusve dense punctatis; petiolis circa 1 cm. longis antice valde
alatis; inflorescentiis axillaribus et lateralibus solitariis vel fascicu-
latis paucifloris vix 1 cm. longis; floribus ignotis ut videtur racemoso-
umbellatis; bracteolis persistentibus anguste oyatis acutis circa 1.5
mm. longis; pedicellis fructiferis ad 5 mm. longis; sepalis plerumque
5 breviter connatis vix 1.5 mm. longis qvatis obtusis dense glandulo-
lineatis, punctis valde elongatis, margine fere nudis membranaceis
baud ciliolatis; drupis globosis fere 6 mm. crassis valde cum glandulis
rubris elongatis rugulosis. Peru: Caballo-cocha, Loreto, L. Williams
2117 (type, Field Museum), 2081. Lower Rio Nanay, Williams
587(7).
Apparently this small tree is referable to Conomorpha or to
Stylogyne, and I have placed it in the latter genus largely because
its sepals are more typical for that group. It is only on the inflores-
cences (as interpreted) that it has been eliminated from other genera.
These are in mature fruit and broken, but seemingly the flowers were
pseudo-racemose or subumbellate. The large leaves, glabrousness,
persistent bracts, and lineate (not punctate) glandulosity constitute
a combination of characters apparently not described for any species
in either of the genera considered.
Ditassa Schlechteri, nom. nov. D. montana Schlechter, Medel.
Herb. Leid. 29: 13. 1916, not Decaisne in DC. Prodr. 8: 576. 1844.
The earlier use of the name montana in this genus precludes its
acceptance for this recently discovered Bolivian species.
Blepharodon nitidum (Veil.), comb. nov. Cynanchum nitidum
Veil. Fl. Flum. 3: pi. 74- 1827; text 121. 1825. B. diffusum Dene, in
DC. Prodr. 8: 603. 1844.
Decaisne has referred the plant of Vellozo unquestionably to his
own B. diffusum.
Metastelma Schlechteri, nom. nov. M. ditassoides Schlechter,
Medel. Herb. Leid. 29: 13. 1916, not Schlechter, Notizbl. 6: 175.
1914.
In naming this Bolivian plant Schlechter evidently overlooked
his own use of the same species name for a Brazilian Metastelma.
Tassadia Rusbyi, nom. nov. T. Sprucei Rusby, Bull. Torr.
Club 25: 498. 1898, not Fourn. in Mart. Fl. Bras. 6 4 : 229. 1885.
It is not obvious that this Rusbyan species is referable to one
previously described.
Prestonia amazonica (Benth.), comb. nov. Haemadictyon
amazonicum Benth. in Mart. Fl. Bras. 6 2 : 166. 1860.
PERUVIAN PLANTS 35
The union of Haemadictyon and Prestonia seems to be universally
accepted, but I do not find that this excellent Amazonian species
has been transferred.
Odontadenia anoinala (Heurck & M. Arg.), comb. nov. Aniso-
lobus anomalus Heurck & M. Arg. Obs. Bot. 160. 1870.
Odontadenia Benth. antedates Anisolobus A. DC. by three years,
and this Peruvian species does not seem to have been named under
the former genus.
%
Of
a
ION 291
VOL. XI, No. 2
)STLY PERUVIAN-IV
lACBRIDE
)F TAXONOMY
THE
AUG 26 1931
UNIVERSITY nr |LU(V()JS
IT OF BOTANY
CHICAGO, U. S. A.
JULY 29, 1931
