TER VERKRIJGING VAN DEN GRAAD VAN
DOCTOR IN DE WIS- EN NATUURKUNDE AAN
DE RIJKSUNIVERSITEIT TE UTRECHT, OP
GEZAG VAN DEN RECTOR MAGNIFICUS
Dr. TH. M. VAN LEEUWEN, HOOGLEERAAR
IN DE FACULTEIT DER GENEESKUNDE, VOL-
GENS BESLUIT VAN DEN SENAAT DER UNI-
VERSITEIT TEGEN DE BEDENKINGEN VAN DE
FACULTEIT DER WIS- EN NATUURKUNDE TE
VERDEDIGEN OP MAANDAG 3 APRIL 1939,
DES NAMIDDAGS TE VIER UXJR

Bij het voltooien van dit proefschrijft wensch ik alle hoogleeraren
en docenten, die tot mijn wetenschappelijke vorming hebben bij-
gedragen, in het bijzonder mijn promotor, prof. dr. A. A. Pulle,
hartelijk dank te zeggen.

The present paper has been written in connection with the ac-
count of the Papilionaceae for P u 11 e's Flora of Suriname. The
investigations were chiefly carried on in the herbarium of Utrecht;
I also spent some time in the herbaria of Kew, Leiden and Paris and
of the British Museum of Natural History in London. I wish to ten-
der my best thanks to the directors and staffs of these institutions
for their hospitality and assistance and also to the „Miquelfondsquot;
which enabled me to go to London and Paris. Further I am in-
debted to the directors of the herbaria of Berlin-Dahlem, Brus-
sel, Geneva and Leiden for lending specimens.

Miss A. Kleinhoonte, who first was to write the account of
the Papilionaceae and had already determinated a large part of the
material, could, owing to lack of time, not finish the work. Some
new species and critical remarks were published by her in Rec.
Trav. bot. neerl. XXV and XXX. On the suggestion of Prof.
A. A. P u 11 e I have taken over her work. I wish to thank here
Prof. P u 11 e for his advice and interest.

Literature, The standard work for the Papilionaceae of tropical
South America is still B e n t h a m's treatment in M a r t i u s Flo-
ra Brasiliensis XV, though of course so many new species have
been described that it has become very incomplete and its nomen-
clature is no longer up to date. B e n t h a m's work is especially
important because B e n t h a m was well acquainted with the Pa-
piHonaceae of other parts of the world also. After B e n t h a m mo-
nographs even of genera are rare, and most of them deal with
restricted areas only. It is a pity that in E n g 1 e r Das Pflanzen-
reich no part of the Papilionaceae has as yet appeared. A revision of

many genera appears to be desirable, especially in connection
with the question, whether B e n t h a m's large conception of the
genera is justified, or whether some recent authors are right in
segregrating some genera.

A. D u c k e, the student of the flora of the Amazonian district,
has especially paid interest to the PapiUonaceae (and Mimosa-
ceae); his various publications, chiefly in Arch. Jard. Bot. Rio
de Janeiro, are also important for the study of the PapiUonaceae
of Guiana.

Delimination of genera. In the delimination of the genera I
have chiefly followed B e n t h a m. For many of the large genera,
as Cassia L, Caesalpinia L, Bauhinia L (f.e. B r i 11 o n and Rose
in N. Am. Flora!), Desmodium Desv. (Schindler) a far
going segregration has been proposed. A general objection to these
treatments is that they take into account the species of a restricted
area only (with exception of S c h i n d 1 e r's still little surveyable
division of Desmodium). There are also some particular objections.
For example, in my opinion Herpetica Raf. (Cassia alata L) and Cha-
maesenrui (Raf.) Pittier (a.o. Cassia reticulata WiUd.) cannot be
considered as distinct genera. Ta über t was certainly not justi-
fied in considering Schnella Raf. (as genus) and Tylotea Vog. (as
section) as one section of the genus Bauhinia L. The distmguishing
characters are given by V o g e 1 in Linnaea XIII and by B e n-
tham. The species enumerated byBrittonandRose under
the genus Schnella Raf. belong all to the section Tylotea Vog.,
Schnella Raf. being restricted to Brazil and Guiana.

In several smaller genera, as Canavalia Adans., Clitoria L,
Macrolohium Schreb.(?), Mucuna L, the sections distinguished
by B e n t h a m appear to be so natural that they are considered by
some authors as distinct genera. With as much reason this could
be done in the genera Centrosema D.C., Dimorphandra Schott,
Dioclea H.B.K., Ormosia Jacks., Peltogyne Vog. etc.

The deliminations between the genera Vigna Savi, Phaseolus L
and Dolichos L and between the genera Lonchocarpus H.B.K. and

Derris Lour, are stiU arbitrary. The problem can perhaps best
be solved by distinguishing more genera.

Unidentified species. I have tried to identify as much as pos-
sible the species described from Guiana. In the following cases
this was not possible, either because the type specimen was too
incomplete or because I could not obtain the type specimen. The
species enumerated under the nrs. 3, 4, 9, 10 and 13 are at any
rate not known from Suriname.

2.nbsp;Bauhinia Richardiana D.C. 1825. Fr. Guiana.
Described from leaves only. Type specimen not seen.

Type specimen not seen. Cited by B e n t h a m, who also did
not see the type, as synonym of Cassia glandulosa L. sensu
Benth.

4.nbsp;Cynometra racemosa Benth. 1840. Fr. Guiana.
Only once collected? Fruit not known.

Type specimen could not be traced. Judging from the des-
cription, identical with Dioclea glabra Benth.

8.nbsp;Lonchocarpus chrysophyllus Kleinh. 1930. Suriname.
Type specimen incomplete.

9. Machaerium polyphyllum (Poir. 1816) Benth. 1838. Fr. Guiana.
B e n t h a m described a duplicate of the type specimen, Patris
s.n. [G DC]; this specimen could not be traced in Geneva.

10.nbsp;Melanoxylon speciosum R. Ben. 1920. Fr. Guiana (Marowijne
Riv). According to D u c k e perhaps a species of Recordoxy-
lon Ducke.

A monotypic(?) genus of the Sophoreae, of which the pod is
still unknown. Once collected by Richard.

The type specimen consists of imdeveloped leaves only. Pro-
bably a species of Peltogyne Vog.

A comparison with the Papilionaceae of the neighbouring coim-
tries shows that our knowledge of the Flora of Suriname is still
very incomplete. The PapiHonaceae of Para are best known. In
Arch. Jard. Bot. Rio de Janeiro Ducke gives an enimieration of
the Papilionaceae of Para. According to him there are ±: 160
species of the Papilionaceae-Caesalpinoideae and =t 220 species
of the Papilionaceae-Papilionateae. For Siuriname these numbers
are respectively ±: 85 and ± 130. (The cultivated and introdu-
ced species are not taken into account). Though probably the
flora of Para, which is much larger, is richer than that of Suri-
name, it is not to be expected that the difference is so great.
Several of the species enimierated by D u c k e (especially of the
Papilionatae, many species of the Caesalpinoideae having perhaps
a restricted area) are probably merely not collected in Suriname.

In the first place this must be supposed of those species, which
are also known from Br. and Fr. Guiana (Cassia Apoucouita
Aubl.; Cassia praetexta Vog.; Machaerium floribundum Benth.;
Ormosia Coutinhoi Ducke) or also from Fr. Guiana (f.e. Cassia
calycoides D.C.; Cassia Spruceana Benth.; Crudia bracteata
Benth.; Crudia tomentosa (Aubl.) Macbr.; Machaerium altiscan-
dens Ducke; Platymiscium filipes Benth.) or also from Br. Guiana
(f.e. Bowdichia virgiloides H.B.K.; Cynometra bauhiniaefolia
Benth.; Etaballia guianensis Benth.; Hymenaea palustris Ducke;
Lonchocarpus rariflorus Benth.; Mucuna rostrata Benth.). Of
course there are also species known from Suriname and Para,
which are not yet collected in Br. or Fr. Guiana, (f.e. Ormosia
fastigiata Tul.; Ormosiopsis flava Ducke; Poecilanthe effusa
(Hub.) Ducke).

Berlin-Dahlem	B.
British Museum	BM.
Göttingen	GÖTT.
Kew	K.
Leiden	L.
Paris	P.

Dimorphandra conjugata (Splitg.) Sandw. in Kew Bull. 1932.
406; — Mora conjugata Splitg. in Tijdschr. Nat. Gesch. en Phys. IX
(1842) 109; — Dimorphandra latifolia Tul. in Arch. Mus. Par. IV
(1844) 189; Benth. in Benth. et Hook. f. Gen. PI. I. 2 (1865) 587;
Ducke in Arch. Jard. Bot. Rio de Janeiro III (1922), IV (1925) 44.
The description of the hitherto unknown pod can now be given:
Pod dehiscent, linear-oblong, straight or sUghtly falcate, shortly
acuminate or apiculate, with thickened margins, subglabrate,
12—20 cm long and 2—4 cm broad. Seeds flat, obliquely ovate,
up to 1 cm long, with a thick albumen. Embryo with thin foUa-
ceous cotyledons and a straight more or less fusiform radicula.

Fructiferous specimens: Suriname, Nickerie Riv. (Stahel and Gong-
grijp 3576 [U]); Br. Guiana, Demerara Riv. (F.D. 2479 [K]).

Tulasne distinguished in the genus Dimorphandra Schott
three sections: Pocillum Tul., Eudimorphandra Tul. and Phanerop-
sia Tul. The last section, represented by D. latifolia Tul. (= D. con-
jugata (Splitg.) Sandw.) only, wag characterized by its mcluded,
villose petals and conspicuous persistent staminodes. The other
character given by T u 1 a s n e: leaves simply pinnate, is incorrect,
as already pointed out by B e n t h a m. T u 1 a s n e's mistake is due
to incomplete material, in reality the leaves of D. conjugate are
1-2-pinnate with (for this genus) very large leaflets.

Bentham referred T u 1 a s n e's section Phaneropsia to the
genus Mora Schomb. ex Benth. and reduced the latter to a section

of the genus Dimorphandra Schott. The genera Mora and Dimor-
phandra were again separated by D u e k e, the main differences
being that in Mora the leaves are simply pimiate and the seeds
large and exalbuminous, while in Dimorphandra the leaves are
bipinnate and the seeds small and albuminous. D. conjugata was
retained by him in the genus Dimorphandra.

While the two remaining sections, Pocillum and Evdimorphan^
dra, were distinguished by Tulasne and Bent ham chiefly
on accomit of their staminodes — (staminodes broad and oonnivent
in Pocillum, narrow and free in Eudimorphandra) — other distin-
guishing characters (in the form of inflorescence and pod) were
given by D u c k e, who moreover described several species of the
section Pocillum with aberrant staminodes. The two sections can
now be characterized as follows:

Flores sessiles vel breviter pedicellati in racemis v. spicis pau-
cis elongatis. Calyx campanulatus. Petala exserta glabra vel parce
puberula. Staminodia decidua, interdum dilatata et conniventia,
antherifera vel anantherifera, saepius angusta et hbera, anthera
rudimentari praedita. Legiunen late falciforme, dehiscens, ligno-
sum. Semina plana, ovata. Foliola parva numerosa.

A survey of the species belonging to this section is given by
D u c k e in Journ. Wash. Ac. Sc. 25 (1935) 193—198.

Flores sessiles in spicis dense paniculatis. Calyx campanula-
tus. Petala exserta glabra. Staminodia decidua, angusta, libera,
anantherifera. Legumen lineari-oblongimi, crassum, coriaceiun,
indehiscens. Semina cylindrica.

D. conjugata has been plaeed-on accoumt of the inflorescence-
by D u c k e as well as by S p r a g u e and Sandwith in the
section Eudimorphandra; the pod however proves thatT u 1 a s n e's

section Phaneropsia has to be reestablished. The flowers also are
very characteristic.

Flores sessiles in spiels dense paniculatis. Calyx cylindricus.
Petala inclusa pilosa. Staminodia subpersistentia, libera, angus-
ta, anantherifera, lamina crassa carnosa a stipite abrupte distinc-
ta. Legumen lineari-oblongum, planum, rectum vel leviter fal-
catum, dehiscens. Semina oblique ovata, plana. Foliola pauca
magna.

In general the 3 sections are sharply distinguished, so that
D u c k e is inclined to speak of subgenera, but the place of D. Da-
visii Sprague et Sandwith, placed by the authors under Pocillum,
is still doubtful. At present only infertile pods are known, which
agree best with those of Phaneropsia. The form of the stamino-
des, the sericeous (though exserted !) petals and the few and
large leaflets also point in this direction. The form of the inflo-
rescence and calyx is as in Pocillum.

Arbor excelsa, usque ad 50 m altus (teste Stahel et Gonggrijp).
Ramuli petioli inflorescentiae rubiginoso-pubescentes. Folia 20-35
cm longa; pinnae 7-9-jugae, 5-15 cm longae; foliola altemata,
petiolata,6-12-juga, oblongo-lanceolata, apice acuta vel breviter
acute acuminata, basi obliqua, rotundata vel obtusa, 2,5—5 cm
longa, 1-1,5 cm lata, supra nitidula glabra, subtus praesertim ad
costam minute stellato-pubescentia; costa supra impressa subtus
prominente, nervis utrinque tenuissime impressis. Inflorescen-
tia spicata spicis corymboso-paniculatis; spicae tenuae 2-5 cm
longae. Calyx campanulatus 1,5 mm longus, extus sparse pu-
bescens, lobis brevibus imbricatis. Petala glabra, obovato-spathu-
lata, incurva, 2,5 mm longa, 1,5 mm lata. Staminodia decidua,
libera, apice anguste ovoideo-clavata. Ovarium sparse pubescens
fere glabrum, subsessile, stylo brevissimo. Legumen lineari-ob-

„Branches and petioles rubiginous-pilose; flower-bud pinkish-
brown. Tree aibout 30 m high.quot; (L a n j o u w).

Allied to D. exaltata Benth. and D. multiflora Ducke. Both
those species have only 4—5-pinnate leaves; D. exaltata differs
moreover by the prominulous venation of the leaflets and larger
calyx-lobes; D. multiflora by the densely hirsute ovary.

Arbor. Foha T-juga, 15-20 cm longa; foliola alternantia, bre-
viter saepe late acimiinata, obliqua, coriacea, glabra, crebre ve-
nulosa, ne quidem nova pellucido-punctata, terminalia 4—6 cm
longa et —2cm lata, inferiora saepe minora. Flores sessiles albidi.
Spicae 10 cm longae, paniculatae. Bracteas caducissimas non vidi.
Sepala ± 4 mm longa, extus rufo-tomentosa vel glabrata, intus albi-
do-vHlosa. Ovarium ad suturas villosum. Legxmien fere orbicula-
tum, brevissime stipitatum, 3 cm longum et latum, glabmjm. Semen
unicum, ab arillo semicinctum.

Intermediate between C. reticulata Ducke and C. Langsdorfii
Desf., differing from both species by its (constantly?) 7-jugate
leaves and epimctate leaflets. The flowering plant much resem-
bles C. reticulata Ducke, with leaflets of the same form and size.
The fruit of C. reticulata is however quite different, being ovoid,
often 2- or more-seeded, and with a slender about 5 mm long
stipe. According to Ducke in Arch. Jard. Bot. Rio de Janeiro
IV (1925) 46, the leaflets of C. Langsdorfii are always smaller

and (except in a Rio de Janeiro variety) always obtuse. The color
of the arillus is said to be yellow in C. reticulata, and red in C.
Langsdorfii; in C. epunctata it is unknown. The imperfectly
known C. venezuelana Harms et Pittier, with fewer and larger
leaflets is probably also a nearly allied species.

— Apalatoa spicata Aubl. PL Guiane fr. I (1775) 383 1.147 (des-
cr. et ill. leguminis ad Pterocarpum Rohrii Vahl pertinentis ex-
ceptis); — Crudia spicata Willd. Sp. PI. II (1799) 539; Urban in
Symb. Ant. VI (1909) 11 in obs.; PuUe in Rec. Trav. bot. neerl.
VI (1909) 269; — non Crudia spicata (Aubl.) Willd. sensu Benth. in
Fl. Bras. XV. 2 (1870) 238 in obs.; Huber in Bol. Mus. Goeldi
V (1909) 385; Ducke in Arch. Jard. Bot. Rio de Janeiro I (1915)
23; IV (1925) 262; (= Crudia bracteata Benth.); — non Crudia
spicata (Aubl.) Willd. sensu Grisebach Fl. Br. W. Ind. (1860)
216; Fawcett et Rendle Fl. Jamaica IV (1920) 21 (= Crudia
antillana Urban).

Arbor. Ramuli petioli pubescentes, rarius glabrati. Stipulae
fohaceae, acuminatae, obUquae, —3 cm longae, —12 mm latae,
deciduae. Folia 6—15-foliata, saepe 13-foliata, —35 cm longa;
fohola oblonga, apice acuminata vel caudato-acuminata, membra-
nacea, utrinque parce pubescentia vel rarius glabrata, 6—9 cm
longa, 2—2,5 cm lata. Racemi densiflori; rachis bracteae bracteolae
pedicelli pubescentes. Bracteae ovatae, obtusae, 8—15 mm longae
4—8 mm latae. Sepala ovata, puberula, ± 5 mm longa; recepta-
culimi oblique turbinatum, 3 mm longum. Ovarium tomentosum;
stylus basi excepta glaber 8 mm longus. Legumina stipitata, ju-
niora tomentosa, adulta non vidi.

The type specimen of Aublet has glabrate leaves. Its inflores-
cences are not too weU preserved, but by its 13-foliate leaves,
pubescent racemes and tomentose ovary it quite agrees with the
Suriname specimens. C. bracteata Benth. in Hook. Journ. Bot. II
(1840) 101 (type Martin s.n. [K; P]) was afterwards in Fl. Bras
XV. 2 (1870) 238 in obs. identified by B e n t h a m with C. spicata
(Aubl.) Willd., but is a quite distinct species.

The chief distinguishing characters are: Leaves 5—7-foliate,
glabrous; racemes glabrous; receptacle campanulate; ovary pilose
at the sutures; pod glabrous. It has been collected in Fr. Guiana
and Para.

A third si)ecies, C. antUlana Urb. from Jamaica, has also been
confused with C. spicata (Aubl.) Willd., lastly by F a w c e 11 and
R e n d 1 e, Fl. of Jamaica IV. 2 (1929) 121. According to a letter of
Urban in the Kew herbarium Fawcett and Ren die, as
they had no flowering Jamaican plants, based their flower-descrip-
tion on A u b 1 e t's plant. Even at present there are only fruiting
specimens in the herbaria of Kew and Paris. It seems that Brit-
ton and Rose also saw no flowers, for in N. Am. Fl. 23. 4 (1930)
223, under the name of Apalatoa antillana (Urb.) Standley, they
clearly follow the description given by Fawcett and R e n d 1 e.
It is improbable, according to U r b a n, that the bracts of C. antil-
lana are really large, like those of C. spicata (Aubl.) Willd. and
C. bracteata Benth. At any rate it differs from the first species by
its glabrous racemes and by its small and linear stipules, from the
second species by its tomentose pod. The distinguishing charac-
ters given by Urban himself in Symb. Ant. VI are somewhat
confusing, as he misquotes B e n t h a m.

Crudia aromatica (Aubl.) WUld. Sp. PL II (1799) 540; — Tou-
chiroa aromatica Aubl. PL Guiane fr. I (1775) 385 t. 148; — Crudia
unifoliata Kleinh. in Rec. Trav. bot. neerl. XXX (1933) 170.

I have compared the Suriname material with Aublet's specimen
in the British Museum, which seems to be the plant figured on
A u b 1 e t's tab. 148.

Crudia oblonga Benth. Bot. Sulph. (1844) 89 in obs., Fl. Bras.
XV. 2 (1870) 238 in obs.; — Crudia pubescens Benth. in Fl. Bras.
I.e. 240.

Tj^e specimen of C. oblonga Benth. is Martin s.n. [K] from
Fr. Guiana, though Bentham afterwards in Fl. Bras. I.e. gives
as type locality: America centralis and the Index Kewensis: India
occidentalis. Martin's specimen has shortly acimiinate leaflets,
while C. pubescens Benth. (type Spruce [K] from the Rio Ne-
gro) is in Fr. Guiana also a more common form; it has obtuse and
somewhat broader leaflets.

Since the name of the genus Crudia Schreb., has to be conser-
ved against Apalatoa Aubl., the correct name for the species
usually named C. obliquu Griseb. is Crudia glaberrima (Steud.)
Macbr. (Hirtella glaberrima Steud., Apalatoa glaberrimu Taub.),
and for Crudia Parivoa D. C.: Crudia tomentosa (Aubl.) Macbr.
(Parivoa tomentosa Aubl., Apalatoa tomentosa Taub.).

Peltogyne paniculata Benth. in Hook. Journ. Bot. II (1840) 96,
Fl. Bras. XV. 2 (1870) 231 t. 60 fig. 1; Ducke in Arch. Jard.
Bot. Rio de Janeiro HI (1922) 94, t. 19 fig. 5; IV (1925) 265,
Trop. Woods. 54 (1938) 3.

Peltogyne pubescens Benth. in Hook. Joum. Bot. II (1840) 96,
Fl. Bras. XV. 2 (1870) 231, 232 in obs.; Ducke in Arch. Jard. Bot.
Rio de Janeiro III (1922) 95, Trop. Woods 54 (1938) 4; Sandwitli
in Kew Bull. 1931. 366; Britton et Rose in Aim. N. Y. Acad. Sc.
35 (1936) 65; — Peltogyne paniculata Benth. sensu Pulle Enum.
(1906) 210; Pittier in Trab. Mus. Com. Ven. III (1928) 69; Benoist

m Arch. Bot. V. 1 (1931) 108 passim; - Peltogyne venosa Benth.
m Fl. Bras. I.e. p.p. (quoad legumen tantum); - Peltogyne amplis-
sima Pittier ex Knuth in Fedde Rep. XLIII (1928) 370 n.n.

1.nbsp;Leaflets oblong, acuminate; indumentum of the inflores-
cence shortly adpressed-pubescent; flowers ± 5 mm long with
short stipe; petals white; style as long as the ovary; pod at the
upper suture narrowly marginate.........P. paniculata Benth.

2.nbsp;Leaflets ovate-oblong, obtuse; indumentum of the inflo-
rescence loosely pubescent; flowers ± 7 mm long with longer
stipe; petals pink; style twice as long as the ovary (and stamens
in accordance); pod not marginate....... P. pubescens Benth.

Of these characters, some have proved inconstant. The form
of the leaflets in the two species for example is more variable
and therefore less characteristic than many authors have reali-
zed. Specimens of P. pubescens from Suriname and Fr. Guiana
provided with oblong and acuminate leaflets have consequently
been determinated as P. paniculata, and Ducke has even been
inchned to consider P. pubescens as a variety of the latter. In
Tropical Woods I.e. Ducke however treats them as two distinct
(though not sharply distinct) species.

The bark of P. pubescens is said to be greyish (Ducke, Rio
Branco), nearly black, rust-brown on cross section (G on g g r ij p
Suriname) or reddish-brown (S a n d w i t h, Br. Guiana); its heart
wood bright-violet (Ducke, Rio Branco) or violaceous-brown to
brown-violaceous (Pfeiffer, Suriname). P. paniculata has ac-
cordmg to D u c k e a ferrugineous, smooth bark and a red-^brown
to brown-purple heartwood.

In two specimens of P. pubescens from Suriname (B.W. 6889
and 6899) the style is only slightly longer than in P. paniculata.

The leaflets of these specimens are also aberrant; they are thick-
coriaceovis and covered with traces of a white wax. Such leaves
are also known in P. paniculata.

P. latifolia (Hayne) Benth. (Hymemtea latifolia Hayne), accor-
ding to H a y n e a specimen of an unJsnown collector from Bahia,
according to Bentham a plant of Sieber from Para, was only
known to Bentham from the description and figure given by
Hayne. The type was kindly lent to me by the Berlin Herbarium;
it is at any rate not identical with P. puhescens, as suggested by
Bentham, but appears to approach P. ilorihunda (H.B.K.) Benth.

Peltogyne venosa (Vahl.) Benth. in Fl. Bras. XV. 2 (1870) 233
in obs. (descr. leguminis excepta); Ducke in Arch. Jard. Bot. Rio
de Janeiro III (1922) 98 (descr. leguminis excepta); Benoist in
Arch. Bot. V. 1 (1931); — Peltogyne confertiflora Benth. sensu
Pulle Enum. (1906) 210; Pfeiffer Houts. v. Suriname I (1926) 259
fig. 39 non Benth. 1870.

B e n t h a m's fruit description is incorrect; the fructiferous spe-
cimen of Martin [K] cited by him belonging to P. puhescens
Benth. Bentham remarks that P. venosa resembles P. densiflora
Spruce ex Benth. except for the pod and the glabrous ovary. The
pod however, already correctly described by B e no i st I.e., is quite
similar to that of P. densiflora. The specimen B.W. 5852 from
Suriname is a smaU-flowered form which on account of its to-
mentose ovary must be reckoned to P. densiflora, but agrees
otherwise perfectly with the specimens of P. venosa from Br.
Guiana (which I compared at Kew). P. densiflora can therefore
best the treated as a variety of P. venosa.

Peltogyne venosa (Vahl) Benth. var. densiflora (Benth.) Amsh.
nov. comb.; — Peltogyne densiflora Spruce ex Benth. in Fl. Bras.

XV 2 (1870) 232 t. 60 fig. 2; Ducke in Arch. Jard. Bat. Rio de
Janeiro I (1915) 25, IH (1922) 99 t. 60 fig. 2, Trop. Woods 54 (1938)
5; — Peltogyne paraensis Huber teste Ducke.

Possibly Vouapa Simira Aubl. belongs here. The vernacular
name given by Aublet (Simira) points either to Hymeimea Cour-
haril L. (Suniri of the Caraibs) or to a Peltogyne species (Simi-
rang of the Caraibs), the wood described by Aublet as violaceous,
to a species of Peltogyne. Aublet's type specimen in the British
Museum consists of a sterile branch with three very young and
imdeveloped 2-foliate leaves (exactly as described by Aublet). As
the fruits described by Aublet can hardly belong to a member
of this family, the species remains doubtful.

According to the form of the pod and the development of the
receptacle, in Peltogyne four groups can be distinguished. Of three
of these groups the pods have been described and figured by Ducke
in Arch. Jard. Bot. Rio de Janeiro III (1922) 98 pi. 19.

I.nbsp;Receptaculum campanulattmi. Legumen plantun, oblique
rhomboidxmi vel triangulare, dehiscens, coriaceum, venosum,
sutura superiore anguste marginata (rarius emarginata). Semina
arillo parvo praedita. (Group of P. paniculata Benth.).

II.nbsp;Receptaculum campanulattmi. Legumen planum, plus mi-
nusve orbiculatum, indehiscens, coriaceum, venostmi, sutura su-
periore anguste sed distincte alata. Semina arillo minuto prae-
dita. (Group of P. venosa (Vahl) Benth.).

III.nbsp;Receptaculvim brevissimum. Legumen plantam, orbicula-
tum, indehiscens, exalatum, coriaceum, venosum. (Group of the
type species, P. discolor Vog.).

rV. Receptactilum brevissimum. Legttmen plano-compressimi,
dehiscens, exalatimi, sublignosimi. Semina exarillata.

florihunda (H.B.K.) Benth. (Orinoco, and Rio Branco when
Ducke's identification of H.J.B.R. 3860 is correct) and pro-
bably p. latifolia (Hayne) Benth. (Bahia?). According to Ben-
th a m, P. porphyrocardia Grisebach n.n. is probably synonymous
with P. florihunda (H.B.K.) Bth., but it has been described
by Williams in Flora of Trinidad and Tobago as a distinct
species.

Eperua stipulata Kleinh. in Ree. Trav. bot. neerl. XXX (1933)
171; — Eperua Schomburgkiana Benth. aff. Pfeiffer Houts. v.
Suriname I (1926) 248.

This species, described from leaves only, is characterized by
its large stipules and therefore in all probability identical wiÜi
E. Jenmani Oliv., known from Br. and Fr. Guiana. As a wood
sample has been collected and described by Pfeiffer I.e. it is
hoped that one day my identification can be verified.

Elisabetha coccinea Benth. in Hook. Joum. Bot. II (1840) 92;
Ducke in Trop. Woods 37 (1934) 19; — Elisabetha oxyphylla
Harms in Notizblatt 59 (1915) 316i; Ducke ,in Trop. Woods. I.e.

The type specimen of E. coccinea Schomburgk s.n. from Br.
Guiana has retuse leaflets. AH other specimens seen from Br.
Guiana, (Myers 5906 [K]; Im Thum [K], Appun [K; B.M],
Rich. Schomb. [B]) as well the Suriname specimens collected
along the Corantijne River, have acute or obtuse leaflets. They
(especially the narrow-leaved forms) agree therefore with Ule
8146 (compared at Kew) from an affluent of the Rio Branco,
the type of E. oxyphylla, distinguished from E. coccinea on ac-
count of this leaf-character. As no other differences could be

seen it is probable that the type of E. coccinea (though badly pre-
served) is merely a specimen with abnormal leaflets.

Elisabetha coccinea differs from E. princeps Benth. and its
alhes by its relatively few and large leaflets and early deciduous,
inconspicuous (at least not known) stipules and above all by
its pod. The upper suture of the pod is in E. princeps and allies
incrassate and dilated, in E. coccinea narrow and margin-like.

Bauhinia cinnamonea D. C. Prod. II (1825) 517; Benth. in Fl.
Bras. XV. 2 (1870) in obs.; Sagot in Ann. Sc. Nat. (1882) 317 p.p.
(descr. floriim excepta); — Bauhinia Versteegii Pulle Enum.
(1906) 213 t. XI.

Distribution: Fr. Guiana (Martin s.n. fr. [P], type).
Suriname: Upper Gonini Riv. (Versteeg 163 fl. and. fr. Aug., type
of B. Versteegii Pulle; Gonggrijp 3699 fl. Feb.); Upper Suriname Riv. near
Goddo (Stahel 119 fl. and fr. Jan.).

B. cinnamonea was placed by de Candolle (to whom the
flowers were unknown) in the section Caulotretus Rich. (^
Schnella Raddi) and is mentioned iby Bentham, who did not
see the plant, as possibly identical with B. smilacina (Schott)
Steud. Sagot even ascribed to it a detached flower belonging
to some Bauhinia species of the section Tylotea Vog. (probably
B. rubiginosa Bong.). B. cinnamonea however belongs to the group
of B. holophylla (Bong.) Steud. in the section Pauletia D. C.

Bauhinia rubiginosa Bong. in Mem. Acad. Petrogr. VI (1836)
4; Benth. m Fl. Bras. XV. 2 (1870) 208; Ducke in Arch. Jard.
Bot. Rio de Janeiro IH (1922) 109; — Schnella rubiginosa Benth.
in Hook. Joum. Bot. H (1840) 97; — Bauhinia dubia Vog. in

Linnaea XIII (1839) 314 non G. Don 1832; — Bauhinia coronata
Benth. in Fl. Bras. I.e. 209; PuUe Enum. (1906) 214; — Bau-
hinia speciosa Vog. I.e. non Roxb. 1825 n.n.; — Bauhinia superha
Steud. Nom. ed. 2 (1841) 192; — Bauhinia riparia Splitg. ex
Benth. in Fl. Bras. I.e. 208 in obs. p.p. (legumine exoepto); —
Bauhinia marowijnensis Kleinh. in Rec. Trav. bot. neerl XXX
(1933) 72.

In the typical B. ruhiginosa the leaves are bifid. Specimens
in which the leaves are bipartite or even bifoliate have, on ac-
count of this character and of the greater or lesser development
of bracts and bractlets, often been described as distinct species.
The division of the leaves and the size of bracts and bractlets
is however very variable, often even in the same specimen (see
also Ducke I.e.).

Bauhinia dubia Vog., named B. coronata by B e n t h a m, type
specimen Poiteau [B] from Fr. Guiana, is a form with 2-foliate
leaves and relatively small bracts and bractlets (± 3 mm long).
The specimen Versteeg 241, eniamerated by Pulle as B. coro-
nata, agrees better with the following form.

Bauhinia speciosa Vog., named B. superha by S t e u d e 1, type
specimen Poiteau s.n. [B] from Fr. Guiana, entirely agrees with
B. marowijnensis Kleinh., type specimen Kappler s.n. [L; U].
The leaves in this form are bipartite.

Bauhinia riparia Splitg. ex Benth., type specimen Splitgerber
548 [K; L; P] from Suriname, has completely bifoliate leaves,
but otherwise agrees with B. speciosa. The pod Bent ham
describes as jjlegumen multo majusquot; (tlian in jB. T^higiTtosa) is
probably the detached pod found on a sheet of Martin s.n. [K];
this pod however does not belong to it. Curiously enough the
sheet of Splitgerber 548 in the herbarium of Leiden does not
bear any name, while on the duplicates sent to Kew and Paris
the name B. riparia Splitg. was written.

In general it can be said that the specimens from N. Brazil and
Guiana have deeper divided leaves and smaller flowers than
those found in S. Brazil, but that those characters are in both
regions variable.

Possibly B. Outimouta Aubl. is identical with B. ruhiginosa.
The type specimen of Aublet [B.M.], consists of very large,
membranaceous, 2-foUate leaves only. Similar leaves have been
collected in Suriname, but only on sterile specimens. Perhaps
they are leaves from young plants or from coppice shoots.

—nbsp;Bauhinia splendens H.B.K. var. laüfolia Benth. in Fl. Bras.
XV. 2 (1870) 209; — Bauhinia chrysophylla Vog. in Linnaea
XIII (1839) 21 teste Benth. I.e.

The type specimen of Aublet [B.M.] consists of a leafy branch
with a deflorated raceme and some traces of a pod. As other
(flowering) specimens have been collected in Fr. Guiana (Mar-
tin s.n. [B.M.]) and in Suriname (Stahel 129 [U]) it can now
be identified with certainty. The leaves of the type specimen
are not completely glabrous, as described bij Aublet, but minu-
tely pubescent beneath. Such feeble pubescence can often be
found in the broad-leaved (4—5-nerved) specimens distinguished
by Bentham as B. splendens var. latifolia.

—nbsp;Bauhinia splendens H.B.K. Nov. Gen. Sp. VI (1824) 321;
Benth. in FL Bras. XV. 2 (1870) 208; — Schnella splendens Benth.
m Hook. Journ. Bot. II (1840) 97; — Bauhinia guianensis PuUe
Enum. (1906) 214.

— Bauhinia angulosa Vog. sensu Pulle in Ree. Trav. bot. neerl.
IX (1912) 139 non Vog. 1839.

Frutex scandensj cirrhifer trunco complanato. Ramuli juniores
inflorescentiae breviter ferrugineo-pilosi. Folia basi cordata vel
rotimdata, ad ^/a usque ad ^h biloba lobis acuminatis, coriacea,
supra glabra nitida, subtus pubescentia, 7—9-nervia; 7—10 cm

longa 5—7 cm lata. Racemi laxi. Bracteae —2 mm longae, deci-
duae. Pedicelli 8—12 mm longi, graciles, minute bibracteolati.
Alabastra urceolato-globosa, umbone nainute 5-dentato coronata,
ferrugineo-pubescentia. Calycis tubus campanulatus, interdum
irregulariter breviter fissus, 5—7 mm longus. Petala unguiculata,
auriculata, obovata, extus villosa, 1 cm longa, summum complica-
tum. Stamina 10 fertilia antheris parvis ovatis. Ovarium subsessile
villosum stylo glabro aequilongimi. Legimien non visum.

Nearly allied to B. guianensis Aubl. (B. splcndens H.B.K.) and
differing chiefly by its much longer pedicels. Moreover the leaves
in B. guianensis seem to be constantly 2-foliate or nearly so and
the indxamentum of the inflorescence much shorter. B. angulosa
Vog. has the calyx-lobes oblong as in B. ruhiginosa.

Bauhinia cumanensis H.B.K. Nov. Gen. et Sp. VI (1824) 321;
Benth. in Fl. Bras. XV. 2 (1870) 212; — Schnella cumanensis
Britton ©t Rose in N. Am. Fl. 23.4 (1930) 206; — Bauhinia co-
lumbiensis Vog. in Linnaea XIII (1839) 313; — Schnella colum-
biensis Benth. in Bot. Voy. Sulph. (1844) 89; Britton and Rose
in N. Am. Fl. I.e. and in Ann. N. Y. Acad. Sc. 35 (1936) 163;
— Schnella brachystachya Benth. in Hook. Joum. Bot. II (1840) 98.

The Suriname specimens as well as Schomburgk, type of Schnella
brachystachya from Br. Guiana, belong to the form with roiuided
leaflets described by V o g e 1 as B. columbiensis, which B e n t h a m
already considered as synonymous with B. cumanensis H.B.K.,
but which Britton and Rose reinstate as a distinct species.

Fawcett and Rendle Fl. Jamaica IV (1920) 103; - Chamaefistula
frutwosa Pittier in Trab. Mus. Com. Ven. Ill (1928) 152- Britton
and Rose in N. Am. Fl. 23.4 (1930) 237 quoad nomen tanWm non
quoad descr.; — Cassia bacillaris L. f. Suppl. (1781) 231- Benth
in Fl. Bras. XV. 2 (1870) 98 t. 31, Trans. Linn. Soc. 27 (1871)
521, aliis auctoribus.

Britton and Rose l.c. identify _ without commentary -
a Panama species with C. fruticosa, and cite the locality Panama
as well as the figure in Rel. Houst. with an interrogation sign
C. fruticosa Mill, was cited by Benth am as a synonym
under C. bacillaris. After studying the type specimen, Hous-
ton fr. in the British Museum, I see no reason for doubting
the correctness of Bentham's identification. The Panama spe-
cies has according to B r itton and Ro s e suborbiculate leaf-
l^s and a turgid, veiy broad (-2% cm) pod and will resemble
^erefore, better than the type specimen itself, the figure in Rel
Houst., of which B entham says: „The artist has so altered the
proportions, shortening the leaflets and pod, and increasing the

curvatoe of the latter, as to make it (the plant) quite unrecog-
nizable.quot;nbsp;^

Cassia nitida Rich, in Act. Soc. Hist. Nat. Par. I (1792) 451-
- Cassia viminea L. sensu D. C. Prod. II (1825) 494 p.p. (quoad
specmima Portoricensa); - Cassia quinquangulata Rich, sensu
Urban in Symb. Ant. IV (1905) 272 non Rich. 1792; - Chamae-
fistula antUUina Britton et Rose Sc. Surv. Porto Rico V (1924)
369, N. Am. Fl. 23.4 (1930) 233.

T^e type specimen, Leblond s.n. [P], from the „Antillesquot; is one
of the plants alluded to by R i c h a r d in a note at the end of

Le Blond: Pleraeque plantae GaUo-guianenses, nonnullae Marti-
mcences. De Candolle and Bentham who did not see the

plant tried to identify it with a Guiana Cassia species (D e C a n-
dolle with C. Apoucouita Aubl., Bentham with C. viminea L.
sensu Benth.).

Cassia lucens Vog. Syn. Cass. (1837) 46, Linnaea XI (1837) 687;
— Cassia racemosa Mül. sensu Benth. in Fl. Bras. XV. 2 (1870)
126, Trans. Linn. Soc. 27 (1871) 549; Pulle Enum. (1906) 216 aliis
auctoribus; non C. racemosa Mill. Diet. ed. 8 (1768); Vogel in
Linnaea XV (1841) 170; Craib in Kew Bull. 1912. 151 passim.

I failed to identify the type specimen of Cassiu racemosa ([B.M]
from Colombia) with any Cassia species known to me. At any
rate it can not be identical with C. lucens Vog., as supposed
by Bentham, the form of the inflorescence and the number,
form and venation of the leaflets being different (the flowers are
poorly developed). A description of Miller's plant is given by
Vogel in Linnaea XV. I.e., who considered it a poor specimen
of a doubtful species.

Cassia viscosa H.B.K. Nov. Gen. et Sp. VI (1824) 360; Benth. in
Fl. Bras. XV. 2 (1870) 132, Trans. Linn. Soc. 27 (1871) 559; —
Grimaldia viscosa Britton et Rose in Ann. N. Y. Acad. Sc. 35
(1936) 187; — Cassia cuneifolia Vog. Syn. Cass. (1837) 51, in Lin-
naea XI (1837) 695; — Grimaldia cuneifolia Britton et Rose in
Ann. N. Y. Acad. Sc. I.e.; — Cassia viscoso-pilosa Steud. in Flora
1843. 760.

var. acutifolia Ducke in Arch. Jard. Bot. Ill (1922) 116; —
Grimaldia columhiana Britton et Rose in Ann. N. Y. Acad. Sc. I.e.

Britton and Rose distinguish a form with ovate acute leaf-
lets as G. columhiana and a form with obcordate leaflets as G.
cuneifolia (they consider C. viscosa as a doubtful species perhaps
synonymous with C. hispidula Vahl). The type specimen of C.
viscosa H.B.K. ([P] from Colombia) shows obcordate. ovate and

mtermediate leaflets. All other specimens seen from Colombia
(several specimens m the Paris herbarium and in the Kew her-
barium André 867 and 2922 cited by Br it ton and Rose
under G. columbiana) have ovate leaflets, which may be either
rounded or obtuse and aristellate at the apex or acute. This form
occurs also, though less common than the form with obcordate
leaflets, in Brazil and has been described by D u c ke as C vis-
cose var. acutifolia. The type specimen itself proves by the va-
riability of its leaflets that it is not possible to distinguish two
species on accoimt of leaf characters only.

The pubescence of the pod in the Colombian specimens is some-
what longer than in the Brazilian specimens.

Cassia faginoides Vog. Syn. Cass. (1837) 50, - Cassia hispidula
Vahl var. faginoides Benth. in Fl. Bras. XV. 2 (1870) 131- —
Cassia Killipii Rose in Journ. Wash. Acad. Sc. 17 (1927) 167;

Considered byBenthamasa variety of C. hispidula Vahl
differs from that species by its rounded flower-buds and pubes-
cent leaflets only (in C. hispidula the buds are acuminate and
the leaflets glabrous). The type specimen (Sellow, duplicate seen
m Paris, and several other specimens from S. Brazil in Paris
and Kew) agrees weU with G. decora Britton et Rose (type
Palmer 501, duplicate seen in Kew) except for the somewhat
more strongly nerved leaflets. I consider G. Killipii as a small-
leaved, few-flowered form. The Surmame specimen, Frickers and
Muller 19, belongs to this form.

Cassia tetraphyUa Desv. s.l.; - Cassia Desvauxii Coll.
Of the section Chamaecrista, to which this and the following

„(The section is) an exceedingly puzzling one to botanists. The
nicest shades by which the majority of forms pass into each
other make it impossible to settle what is to be regarded as spe-
cies with any satisfaction.quot; This citation is especially applicable to
the 4-foliate Cassiae of the series Xerocalyx Vog. The apparently
quite independently varying characters axe: form and size of the
leaflets, length of the pedicels, size of the flowers, gland (stipi-
tate-sessile) etc. (especially if one takes into account specimens
from different regions).

Bent ham distinguishes in this group 8 species („species om-
nes vix inter se distinctaequot;). In Surtname 3 forms can be dis-
tinguished; according to B e n t h a m's treatment in the Flora
Bras, one must be reckoned to C. Desvauxii CoU., one to Cassia
uniflora Spreng., while the third is a mountain form apparently
not described before.

As probably synonymous with C. imijlora Spreng. B e n t h a m
cites C. Persoonii Coll. This name was given by Col laden in
1816 to C. laticeolata Pers. 1806 (non Forsk. 1775) and has been
accepted by most authors, because it is the oldest (though doubt-
ful) name and because C. uniflora Spreng. is a later homonym
of C. uniflora Mill. 1768. Probably a specimen named C. lanceolata
in the herbarium Persoon [L] has to be regarded as the type
specimen. It has linear-oblong, 12—14 mm long and 3—5 mm
broad leaflets, lanceolate-cordate stipules of nearly the same
length, solitary sessile petiolar glands and one slender flowering
pedicel about 3,5 cm long, bearing neither flowers nor fruits.
Perhaps a better duplicate may be found in the herbarium La-
marck [P], from where Persoon's specimen must have come, but
I could not trace it. It is however already evident that C. lan-
ceolata Pers. cannot be identified with C. uniflora Spreng., which
(according to B e n t h a m) has oblong leaflets, a stipitate gland,
and shorter and thicker pedicels.

—nbsp;1 cm long. Petiolar gland sessile, depressed. Pedicels slender^
during flowering 1—2 times as long as the leaves. Largest sepals
± 1 cm long. Petals somewhat longer than the sepals. ^Ovary
villose. Pod oblong, - 3 cm long and 6 mm broad, adpressed
pubescent.

To this species is also reckoned by Bentham an Amazonian
form with ± oblong, larger leaflets, larger flowers and pedicels
shorter than the leaves. Some Suriname specimens agree with
this form; it is probably the same aa the form identified by B r i t-
ton and Rose with C. pulchra H.B.K. (But this species agrees
perfectly with C. tetraphylla Desv.; compared in Paris).

In order to add as little as possible to the confusion, I have
treated the 3 Suriname forms as varieties of the first legitimately
published species, C. tetraphylh, Desv. It is probable that the other
species admitted by Bentham may also be considered as va-
rieties, but the delimination of those varieties in the different
regions will need a special study.

Cassia tetraphylla Desv. Journ. Bot. Ill (1814) 72; — Chamae-
crista tetraphylla Britton et Rose lin Ann. N. Y. Acad. Sc. 35
(1936) 183; — Cassia Desvauxii Coll. Hist. Cass. (1816) 131-
Benth. in Fl. Bras. XV. 2 (1870) 157, Trans. Linn. Soc. 27 (1871)
568 p.p.; — Cassia pulchra H.B.K. Nov. Gen. et Sp. VI (1824)
362; — Chamaecrista pulchra Britton et Rose l.c. quoad nomen.

Distribution: Amazonian district, Suriname (the Suriname
specimens with subobtuse stipules) ......

— Cassia ramosa Vog. Syn. Cass. (1837) 55 and in Linnaea XI
(1837) 704; — Cassia uniflora Spreng. Neue Entd. I (1820) 291;
Benth. in Fl. Bras. XV. 2 (1870) 157 t. 43 fig. 1; non MiU. 1768;
— Cassia uniflora var. ramosa Benth. in Fl. Bras. I.e.; — Cassia
uniflora var. parvifolia Benth. in Trans. Linn. Soc. 27 (1871) 568;
Pulle Enum (1906) 217; — Cassia savannensis Miq. in Ann. Nat
Hist. 1843, 15.
A specie praesertim differt glandula stipitata.

A specie differt statura parva, foliolis oblongis, glandula majore
scutellata brevissime stipitata, floribus minoribus.

Suriname. Upper Litanie Riv. (mount Knopaiamoi, Rombouts 809
type [U]; mount Teeboe, Versteeg 775 named C. uniflora Spreng, by
Pulle i.e.); Voltzberg (Pulle 267; Lanjouw 871).

In recent floras (Fawcett and Rendle Fl. of Jamaica; N.
Am. Fl.) this species is restricted to Jamaica. Cassia virgata
Swartz, treated by Bentham as a distinct species, has proved
to be identical with C. glandulosa L. The position of C. glandulosa
L. sensu Benth. has consequently become doubtful. Its distribu-
tion is according to Bentham: Brazil, Guiana, Colombia and
Peru. Bentham cites 7 synonyms the oldest of which, C. Otter-
heinii Mey. 1818, is known from the description only. It is also
evident that Bent ha m's conception of the species is much
larger than will be tolerated by many authors.

Cassia disadena Steud. (type Hostmann 1179 from Suriname)
also cited by B e n t h a m as synonym of C. glandulosa L. sensu
Bentham, agrees very well with a W. Indian variety of C. glan-
dulosa L, the var. Sumrtzii (Wikstr.) Macbr. (Cassia or Chamae-
crista Swartzii of other authors.) The only difference is that in

the two Suriname specimens most leaves have two petiolar glands,
while in the W. Indian specimens two petiolar glands are an
exception.

In the Suriname specimens referred hy me to this species the
pubescence of the pod is longer than in the typical form; they
agree however m this character with the description of C. steno-
carpa given in the N. Am. Fl. (and with C. Broughtonii Fawcett
and Rendle, considered m the N. Am. FL as a synonym).

Cassia Pennelliana Amsh. nom. nov.; — Chamaecrista Brownie
ana Britton and Rose in N. Am. Fl. 23.4 (1930) 293, Ann. N. Y.
Acad. Sc. 35 (1936) 117 in key, non Cassia Browniana Kunth 1824.

Of the two specimens collected by Rombouts under nr. 420, one
agrees well with Ch. Brovmiana (compared with a specimen of
Fennel in Kew from Colombia); in the other specimen coUected
under this nr. and in Rombouts 356, the pubescence and the gland
are less developed and the leaflets more obtuse.

— Dicorynia paraensis Benth. in FL Bras. XV. 2 (1870) 81 p.p.
(quoad specimen citatum Sagot tantum); Pulle in Rec. Trav. bot.
neerl. IV (1907) 131; Jansonius in Verh. Kon. Acad. Wet. Am-
sterd. 18.2 (1914) 35; Record Timbers of Trop. America (1924)
242; Pfeiffer Houts. v. Suriname I (1926) 262 pi. XI fig. 41; Be-
noist in Arch. Bot. V. 1 (1931) 117 pl. IV et XXIII.

Arbor excelsa cortice cinereo-brunnea. Stipulas non vidi. Ra-
muli juniores petioli inflorescentiae aureo-tomentosi. FoUa 5-7-
foliata; fohola ovata vel oblongo-ovata, basi obtusa vel rotundata.

Fig 2L Dicorynia guianensis Amsh. a. A 3-seededpod. b. An 1-seeded pod.
d. Flower, petals and part of the sepals removed (B.W. 452). e. Diagram.
Dicorynia paraensis Benth. (forma? uaupensis Spruce), c. Flower. (H.J.B.R.
23319). Marttusia, parviflora Amsh. f. Flower (B.W. 22).

apice acuminata coriacea, supra glabra, subtus pubescentia gla-
Wentia, 7-15 cm longa et 3-6 cm lata; nervi primariL ve-
^que supra inconspicuis, nervis subtus prominentibus venis laxe
reüculatis subtus plusminusve prominulis (vel in aliis specif!

Bracteae bracteolaeque ovatae, -6 mm longae, caducissimae. Pe^
dicelli - 9 mm longi, basi articulati. Sepala inaequalia extus
sericeo-t^tosa, 1 cm longa, 2 exteriora coriacea, latiora 2

Peta a teste coll. alba ungue nigro, breviter (3-4 mm) unguicu-
lata lamine suborbiculato 1 cm longo, extus pubescentia. Stam^aa

(superior paullulum compressior) 5 mm longae 2 mm crassae
utraque 8-locularis. Ovarium sessüe, velutinum stylo glabroTuW

sail 5 mm late alata subcoriaceum, parum venosum, diu tomen-
tosum adultum plus minusve glabratum, 5-7 cm longum etquot;quot;
cm latum. Semina 1-3, suborbiculata, circiter 1,5 cm longa, al-

^nth., It differs however of all hitherto described species of
^coryu^ by Its anthers, which ar. nearly equal and both 8-cel-
led. In o^her species the anther of the superior stamen is 4-celled

a paraen.« is a very variable species or perhaps some of those
vane^s have to be regarded as distinct species (according to
Taubert and Ducke). The following varieties and fJrms
have been distmguished by Bentham:

D. paraensis Benth. Type specimen from „Paraquot;. [P]. Rio
Negro (Spruce 1918 and 3501 [K]; H.J.B.R. 35072).

Forma parvifolia Benth. Manaos (Spruce s.n. anno 1835;
H.J.B.R. 20337 and 24184).

A small flowered and slender form with small narrow leaflets.
Considered by D u c k e as a good species. The small, glabrous,
narrowly winged pod (quite different from that of D. guianensis)
described and figured in the Flora Bras, belongs to Spruce 2135
cited above. The fruits of the other „formsquot; of D. paraensis are
not known.

D. uaupensis Spruce ex Benth. (Rio Uaupes, Spruce 2772 type,
[K]; Manaos, H.J.B.R. 23319; specimen in hb. Paris from „Paraquot;

Leaflets more distinctly reticulate and shining and sepals and
especially petals more pubescent than in other „formsquot;.

According to Ducke in Arch. Jard. Bot. Rio de Janeiro IV
(1925) 58 D. paraensis does not occur in Para in its present limits,
but is replaced there by D. ingens Ducke, easily recognizable
by the dark color of the indumentum. All specimens of D. para-
ensis Benth. s.l. of which the locality is known as well as all
specimens of D. breviflora Benth. and D. macrophylla Ducke have
been collected along (the Rio Negro or its affluents.

G1 e a s o n recently, in Phytologia I (1935) 141, proposed to
replace the name Martiusia Benth. 1840 by the name Martioden-
dron Gleason, because the name Martiusia (M. physalodes) was
already used by S c h u 11 e s in 1822 for a Clitoria species of Cli-
toria section Neurocarpum Benth. Some authors consider this sec-

Martiusia parviflora Amsh. n. sp.; _ Martiusia parvifolia Benth.
sensu Pulle Enum. (1906) 218 non Benth. 1840; Pfeiffer Houts
V. Suriname I (1926) 266 pl. XI fig. 44; - Martiusia excelsa

Arbor excelsa, 30-35 m alta, cortice cinereo-brunnea (BW)
Süpulas non vidi. PetioH inflorescentiae aureo-rufo-tomentosi.'
Folia 7-jugata vel rarius 5-jugata; foliola ovata vel ovato-oblonga
apice breviter acuminata, basi obtusa rotundata vel subcordata
^lacea, supra glabra, subtus sparse pubescentia, flavescentia,'
»-14 cm longa 4-5 cm lata, costa supra impressa subtus promi-
nente, nervis primariis venisque supra inconspicuis subimpressis
nervis subtus parum prominentibus venis , dense reticulatis subtus
vix prominulis. Panicula ampla. Flores flavi (B.W.). Alabastra
ac^mata, incurva, extus dense pubescentia, -1,5 cm longa Se-
pala lanceolata, 1,5 cm longa. Petala obovata, basi attenuata gla-
bra maequaha, 16 mm longa et 6-10 mm lata. Stamina 4 filamen-
tis 1,5 cm longis antheris parum inaequahbus 11 et 13 mm longis
pilosis. Ovarium sessile, tomentosum, stylo glabro. Legumen (in-
fertile t^tum vidi) planum, oblongum, minute tomentosum, ve-
nosum, 10-15 cm longum 5 cm latum, sutura dorsah -2 cm
sutura ventraU —1 cm late alatum.

The flowers are twice as small as in M. exeelsa Benth. and
M. parvifolia Benth. and even somewhat smaller than in M, elata
Ducke (Martiodendron macrocarpon Gleason is identical with M.
elata var. occidentalis Ducke). M. parviflora is the only one of the
four hitherto described species iin which a tomentose ovary and
pilose anthers go together. By its large braa'dly winged pod it is
nearly allied to M. elata and M. parvifolia, both species with
7—9-foliate leaves and narrower leaflets.

Alcohol material of the fruits of this species has been col-
lected in Suriname. The fruit is oblong, 1—2-seeded, 5—7 cm long
and 2,5—3 cm broad. The seeds show a iJiin albimien and an
embryo with thin foliaceous cotyledons.

In B e n t h. et H o o k. f. Gen. PI. I. 2 (1865) 562 and in Fl. Bras.
XV. 2 (1870) 46, it is stated by Bentham that the seeds of
Sclerolobium arei exalbiiminous. On t. XII fig. 1 in Fl. Bras, the
seeds of S. paniculatum Vog. are figured, and in this figure in-
deed no albumen is visible, but in dried seeds the albimien, may
ibe inconspicuous.

Tulasne in Arch. Bot. Mus. Par. IV (1844) 125 describes
for the seeds of S. sericeum Txil. (= S. chrysophyllum Poepp. et
End.) an „iategtamentimi interior (vel perispermum) crassimum
comeo albeoquot; also apparently an albumen.

The presence of albumen in the seeds of Sclerolobium would
be another argument for the near alliance of the genera Sclero-

lohium Vog. and Tachigalia. Aubl., placed usually in different
groups, but whose fruits too resemble each other closely. Tachi-
galia Aubl. was placed by B e n t h a m under the Caesalpinoideae
— Amherstiae on account of the character: stipe of the ovary ad-
nate to the waU of the receptacle; it can further be easily
distinguished because of its obliquous receptacle, but in other
flower-characters the two genera show much resemblance.

Swartzia apetala Raddi Quar. Piant. Nuov. (1819) 19; Benth. in
Fl. Bras. XV. 2 (1870) 30; — Swartzia glabra Vog. in Linnaea XI
(1837) 175.

A specie differt foHolis distincte acuminatis venis utrinque pro-
minulis paullulum laxius reticulatis.

Swartzia apetala itself is not known from the Amazonian district.
There are some small differences in the leaflets, but the flowers of
the Suriname spechnen agree perfectly with those of the species.
In Salzmann s.n. from Bahia, the leaflets are also acuminate.

As this species has been confused with an Amazonian species a
detailed description is given here.

Swartzia Benthamiana Miq. in Stirp. Stu-. Sel. (1850) 15; Ben-
tham in Fl. Bras. XV. 2 (1870) p.p. (quoad spechnina citata'Sagot
et Kappler tantum); Pulle Enum. (1906) 220; Benoist in Arch. Bot.
V. 1 (1931) 127;? Sandw. in Kew Bulletin 1934, 362.

Arbor. Ramuli petioli petioluli tomentosi vel glabrati. Stipulas
non vidi. Folia 5—7-foliolata rachi subterete; foliola ovata vel
oblonga apice acuminata basi obtusa vel rotundata, coriacea supra
glabra nitidula subtus minute cinereo-pubescentia 10—16 cm
longa et 5—7 cm lata costa supra impressa subtus valde promi-
nente nervis primariis venisque supra inconspicuis subtus plus
minusve prominulis. Racemi axillares terminalesque interdum pro
parte laterales, multiflori, tomentosi, —20 cm longi. Bracteae mi-
nutae 1 mm longae; bracteolae nullae. Pedicelli robustiores 4—6
mm longi. Alabastra globosa, dura, nitidula, rufo-tomentosa, adul-
ta —7 mm longa. Calyx coriaceus valde, irregulariter in segmenta
4—5 dehiscens. Petalum xmguiculatum orbiculatum, deciduum,
6—9 mm longum, ex coll. album. Stamina majora 4 antheris
oblongis 2 mm longis filamentis basi pilosulis; stamina minora
numerosa antheris parvis 1 mm longis. Ovarium stipitatum to-
mentosum 4—6-ovulatum 3—4 mm longum 2—2,5 mm latum
stipite tomentoso ± 4 mm longo stylo 1 mm longo uncinato gla-
bro. Legumen ovatum paullum compressum stipite brevi crasso
glabro 5—6 mm longo 4 mm in diametro, lignosum, tomentosum,
reticulato-lamellatum, circiter 6 cm longum 4 cm latum 3,5 cm
crassum. Semen unicum ovatum compressum 5 cm longum 3 cm
latum —1,5 cm crassum arillo parvo margine crenulato funiculo
dilatato 1,5 cm longo.

The type specimen, Kappler 1929 [U] has lost aU its flowers;
a duplicate in Paris is better preserved.

The described fruits were preserved in alcohol. Fructiferous
material has also been collected in Fr. Guiana (see Benoist I.e.).
The ribs of the pod are already conspicuous in very young stades.

In the Amazonian specimens hitherto identified with S.
Benthamuma the pod is smooth; they have therefore to be
regarded as a distinct species:

— Swartzia Benthamiana Miq. sensu Benith. in Fl. Bras.
XV. 2. (1870) I.e. p.p. (quoad specimen citatum Spruce 1843) et
in obs.; Ducke in Arch. Jard. Bot. Rio de Janeiro IV (1925) 290;
VI (1933) 28.

Swartzia lamellata Ducke which was distinguished by the author
from S. Benthamiana on account of its lamellate pod, is pro-
bably a variety of S. Benthamiana, but differs in having smaller
leaflets, longer bracts and bibracteolate pedicels. S. laevicarpa
shows some variability in leaf-characters, but S. Benthamiana is
rather uniform, at least in Fr. Guiana and Sioriname. Flowering
specimens are characterized by their hard, globose, rugulose flo-
wer-buds, coriaceoxxs, very irregularly splitting calyx and the
colour of the indumentum of the inflorescence. Only in Lanjouw
908 (with the same kind of leaves) the calyx is less coriaceous,
so that the flower-buds are compressed and indented (in sicco)
resembling those of S. laevicarpa. Specimens from Br. Guiana
according to Sandwith I.e. resemble the Amazonian species;
as no fructiferous specimens have been collected in Br. Guiana
their identity is still doubtful.

Arbor. Ramuli novelli ferrugineo- vel albido-tomentosi. Stipu-
las non vidi. Folia 7—15-foliolata rachi subterete; foliola oblonga
apice acimainata basi obtusa vel rotundata, glabra, 7—12 cm longa
3—5 cm lata, coriacea, nervis primariis venisque utrinque inconspi-
cuis. Racemi laterales rufo-^omentosi. Bracteae oblongae, concavae,
—6 mm longae. Pedicelli sub alabastro —5 mm longi, bibracteolati.
Alabastra conoidea-globosa, tomentosa, —9 mm longa. Flores aper-
tos non vidi. Stamina majora 5 filamentis glabris; stamina minora
mmierosa. Ovarium glabrum stylo filiforme in alabastro —6 mm

longo. Legumen stipitatum, oompressum, oblongum, sublaeve, cir-
citer 8 cm longum 5 cm latum 2,5 cm crassum stipite crasso 2 cm
longo. Semen unicum, compressum, 5 cm longum 2,5 cm latum
1,5 cm crassum, arillo parvo crenulato, funiculo paullulum dilatato
3 cm longo.

Suriname: Sektie O (tree n. 534, B.W. 1320 ster., 2303 fl. (buds) Aug.,
2534 ster.; B.W. 5381 ster.); Brownsberg (tree n. 1011, B.W. 1790 ster '
3308 fl. (buds) Sept.).

The tree n 1011 is described byGonggrijpas follows-
Tree, — 60 cm in diameter. Stem crooked, form very irregular,
with broad deep furrows and slanting frames, and with very high
spurs. Head irregular with crooked steeply slanting branches.

Nearly allied to S. polyphyUa D. C. from Fr. Guiana and men-
tioned under that name in Pfeiffer Houts. v. Suriname. S. poly-
phylla (type specimen kindly lent by Geneva; other specimens
seen: Sagot fl. [P]; Melinon fr. [P]) has the flower-buds —5 mm
long, globose and with darker indumentum, the bracts and bract-
lets minute (— 1 mm long), the leaflets smaUer (— 9 cm long,
usually shorter) and relatively broader. Its pod is rather similar.

The stem of S. remigifer and of other Swartzia species with
the same type of stem is used by the Indians for the making of
paddles.

Arbor. Ramuli novelli petioli petioluli tomentosi. Stipulae an-
guste lanceolatae, —1 cm longae. Folia 11—13-foliolata rachi an-
guste alata; foliola breviter petiolulata, stipellata, oblonga, apice
acuminata basi obtusa, membranacea, supra glabra subtus puberula,
4—8 cm longa 1,5—2,5 cm lata; nervis primariis supra inconspicuis
subtus prominentibus venis laxe reticulatis supra inconspicuis
subtus prominulis. Racemi laterales, tomentosi, 20—40 cm longi.
Bracteae ovatae 2 mm longae. Pedicelli robusti minute bibracte-
olati, apice valde dUatata, 2—2,5 cm longi. Alabastra ovoidea basi
inconspicue in pedicellum transeuntia apice obtusa, rugosa, sub-

tomentosa. Calyx crasse coriaceus in segmenta 4 aequalia dehis-
cens; segmenta oblonga 1,5 cm longa. Petalum magnum unguicu-
atum lamine obovato extus basi et secus venas pubescente, 4 cm
longum et 2,5—3,5 cm latum. Stamina majora circiter 15 filamentis
glabris antheris oblongis 2,5 mm longis; stamina minora numerosa
antheris 1 mm longis et latis. Ovarium stipitatum lineare tomen-
tosum, 2 cm longum 2 mm latum stipite 6 mn^ longo tomentoso
stylo fUiforme glabro (non ;bene vidi). Legumen subteres,
glabrum, 2—10-spermmn, inter semina constrictum, —20 cm
longum et circiter 1,5 cm latum, irregulariter rugosum.

Allied to S. Inxiflora Bong, and S. xanthopetala Sandw. From
those and other allied species it is readily distinguished by its
strongly dilated pedicels, ovoid flower-buds merging into the pe-
dicels and oblong sepals. The leaves resemble most those of S
Uxifhra Bong. (Dther aUied species are S. ohscura Huber and
S. ingifolia Ducke.

Swartzia prouacensis (Aubl.) Amsh. nov. comb.; - Bocoa
prouacenci^ Aubl. PI. Guiane fr. Suppl. (1775) 38 t. 391; Benoist
m Arch. Bot. V. 1 (1931) 132; - Swartzm minutiflora Kleinh. in
Ree. Trav. bot. neerl. XXII (1925) 408; Pfeiffer Houts. v. Suri-
name I (1926) 275 pi. XII fig. 47.

It was already suspected by Pfeiffer (on account of the
wood structure) and by Kleinhoonte that the Bocoa proua-
censis Aubl. might be identical with the Suriname specimens pro-
visionally described as S. minutiflora Kleinh. (and not identical
with Etaballia guianensis Benth. as often supposed, or with another
species of this genus probably to be united with the genus Inocar-
pus Forst.) Comparison with the type specimen of Aublet [B.M.]
— a sterile branch — showed that Pfeiffer and Klein-
h o o n t e's view is correct. The only fertile specimen which seems
to have been collected in Fr. Guiana is Sagot 1210 fr. [P, K]; its

Swartzia prouacensis (Aubl.) Amsh. belongs to the series Ste-
nantherae Benth., a name not very appropriate for S. prouacen-
sis in which the anthers are ovate-oblong. The group is a very
natural and distinct one, characterized as follows: Flowers small
in caulifloroiis racemes; calyx membranaceous splitting into 3
segments; petal wanting; stamens relatively few (—30), equal;
pod coriaceous, 1-seeded (as far as known).

Species belonging to this series are S. alterna Benth., S. mollis
Benth., S. racemulosa Huber, S viridiflora Ducke and judging
from the description S. cubensis (Britton et Wilson) Standley.

In the conception of the genus Diplotropis Benth. there have
been large variations, from the large conception of B e n t h a m in
Fl. Bras. XV. 1 (1859) 319 to its complete union by Ducke
with the gentis Bowdichia H.B.K.

In the Fl. Bras. Bentham reckons 5 species to Diplotropis,
in which he distinguishes 2 sections: Diplotropis Benth. and Cla-
throtropis Benth. The latter section has been elevated by H a r m s
inDallaTorre and Harms Gen. Siph. fasc. Ill (1901) 221,
to the rank of genus. A key to the Brazilian genera of the affi-
nity of Diplotropis is given by Ducke in Arch. Inst. Biol. Veg.
4.1 (1938) 18.

In the section Diplotropis the type species Diplotropis Martiusii
Benth. 1838 was united by Bentham with the genus Dibra-
chion Tul. (Dibrachion brasiliense Tul. and D. guianense Tul.)
and with Diplotropis ferrv^inea Benth. Of those plants, only the
pod of D. Martiusii and unripe pods of D. ferruginea were known
to hitn. In the generic description Bentham said that the pod
is thick-coriaceous or nearly woody, and tardily dehiscent. The

fruit of D. gumnensis was described by P u 11 e in 1907 and af-
terwards the fruit of D. brasiliensis by D u c k e. As those fruits
proved to be membranaceous and indehiscent, Ducke placed
D. guianensis, D. brasiliensis and D. ferruginea in the genus
Bowdichm H.B.K, retaining in the genus Diplotropis D Mar-
tiusu only, (in Arch. Jard. Bot. Rio de Janeiro I (1915) 32) Af
terwards (I.e. Ill (1922) 131, V (1930) 134) Ducke foundquot;that
the pod of D. MaHiusii is also indehiscent, though woody-coria-
ceous. For this reason he united the genera Diplotropis and
Bowdtchm completely, distinguishing in the genus Bowdichia a
section Eubowdichia Ducke (including Dihrachion) and a section
Diplotropis (Benth.) Ducke.

In Rec. Trav. bot. neerl. XXII (1925) 393 Bowdichia H.B.K
and Diplotropis Benth. were considered byKleinhoonteas
distmct genera, especially on account of the form of the petals
In Bowdichia the standard is broadly orbiculate and without la-
teral appendages, in Diplotropis the standard is oblong and biap-
pendiculate. This difference, to which little importance is atta-
ched by D u c k e, is very well illustrated in Fl. Bras. 1 c fig 123
(Bowdichia virgiloides H.B.K.) and fig. 127 (Diplotropis brasilien-
sis (Tul.) Benth.).

There is however another character, hitherto overlooked, which
m combination with the form of the standard, seems to me to
^stify a separation into two genera, Diplotropis Benth. (sensu
Klemh.) and Bowdichia H.B.K.

In two species which by the form of the petals must be recko-
ned to Diphtropis Benth. (sensu Kleinh.) and which I could

study in this regard, D. guianensis (Tul.) Benth. and D. racemosa
(Hoehne) Amsh. nov. comb. (Bowdichia racemosa Hoehne), the
seeds are exalbuminate. According to the figure and description
of Bentham, the seeds of Bowdichia virgiloides are provided with
an albumen. There is no albumen mentioned for the seeds of
D. Martiusii by Bentham. I could not study the seeds myself,
but their size and softness make the presence of an albumen
very improbable. In the following delimination of the genera
Bowdichia and Diplotropis, these genera are also sharply charac-
terized against the genus Clathrotropis.

a. Calyx incurved. Stcmdard oblong with 2 lateral basal appen-
dages. Wings and carinal petals free, long-unguiculate. Ova-
ry subsessile or nearly so. Pod indehiscent. Seeds soft,
exalbuminate.nbsp;Diplotropis Benth.

1.nbsp;Pod woody-coriaceous (adapted according to Ducke to
transport by water). Seed one, large, reniform.

,b. Calyx incurved. Standard broadly orbiculate, without lateral
appendages. Carinal petals free. Ovary distinctly stipitate.
Pod membranaceous, indehiscent. Seeds few, small, com-
pressed, hard, albuminous.nbsp;Bowdichia H.B.K.

c. Calyx straight. Standard orbiculate, without lateral appen-
dages. Carinal petals slightly cohaerent. Pod woody, dehis-
cent. Seeds few, large, compressed, without albumen.

Diplotropis purpurea (Ricsh.) Amsh. nov. comb.; — Tachigalia?
purpurea Rich, in Act. Soc. Nat. Hist. Nat. Par. I (1792) 108;

The specimen in the Paris herbarium from the herbarium
Kichard is not named Tachigulm purpurea, but a long des-
mption (partly cited by T u 1 a s n e in Arch. Mus.. Par. 1 c ) has
b^n added by R^hard, in which it is said that the plL is
„affims Tassmequot; (Tassui is Richard's name for Tachigaliu Aubl)
Tulasne gives Tachigaliu? purpurea Rich, as a synonym of his
D^brach^on guianense. Richard's description in Act. Soc. Hist
Nat. I^ar l.c. is very short („Tachigalia? purpurea petiolis tereti-
bus foholis ovatis panicula decompositaquot;). A duplicate of the type
specimen (Fr. Guiana, Leblond, equally not named by R i c h a r d)
was kindly lent to me by the Geneva Herbarium; this duplicate
IS at once the type specimen of D. guianensis Tul

I quite agree with Duck e's suggestion that D. purpurea (D
gu^ne^ (Tul.) Benth.) and D. hrasiliensis are not specifically
distmct, and I regard D. hrasiliensis therefore as a variety of D
purpurea. Other varieties have been described by Ducke in
Arch Jard. Bot Rio de Janeiro V (1930) 132 (under Bomdichia
hras^l^ens^s (Tul.) Ducke). The species and the varieties can be
characterized as follows:

Leaflets ovate, obtuse or retuse, rarely shortly acuminate at
the apex, coriaceous, glabrous, the veins prominulous above and
less so beneath. Indumentum of the inflorescence greyish-rufous-
tomentose.

var. leptophylla (KlehA.) Amsh. nov. comb.;- Diplotropis
leptopkylla Kleinh. in Rec. Trav. bot. neerl. XXII (1925) 392.

Leaflets shortly acximinate, subcoriaceous, with a few scattered
hairs beneath. Otherwise as in the species.

var. brasiliensis (Tul.) Amsh. nov. comb.; — Dibrachion brasi-
liense Tul. in Ann. Sc. Hist. Nat. 2. 20 (1843) 139, Arch. Mus. Par.
rV (1844) 103 t. 7; — Diplotropis brasiliensis Benth. in Fl. Bras.
XV 1 (1862) 32 t. 1267; — Bowdichia brasiliensis Ducke in Arch.
Jard. Bot. Rio de Janeiro I (1915) 32, ffl (1922) 132, IV (1925)
pi. 25 fig. f., g., V (1930) 132.

Leaflets thin-coriaceous, acunünate, veins equally prominulous
on both faces. Indumentum of the inflorescence greyish-ferrugi-
neous.

Leaflets thin-coriaceous, acuminate, the veins prominulous above
and less so beneath. Indumentum of the inflorescence canescent.

Leaflets acuminate, coriaceous, glabrous; veins prominulous
above and less so beneath. Indumentum of the inflorescence dark-
rufous-pubescent. Flowers somewhat larger than in the species.

Para near Faro (H.A.M.P. 15686 [P]). type; N. Matto Grosso
(Krukoff 1562, type of D. triloba Gleason); Bahia (Martius s.n. [P]).

The other specimen cited by G1 e a s o n l.c., Krukoff 1308 from
the same locality, is intermediate between the var. leptophylla
and the var. belemnensis.

The transversal position of the cotyledons in Ormosia and Or-
mosu,pszs appears to be due to a growth process, the cotyledons
m unnpe seeds being obliquous. Comparison wi«x a nearly allied
joup ^ferr^^to below al«gt; shows that the position of ^he rl.
dicula (m relation to the pod) remains unchanged, but that the
e^yledons have ultimately 90° diverged from thei^ orig^ t
sition (in relation to radicula and pod)nbsp;s ^ po

Only in 2 s^i« of Ormosia, O. melanocarpa Kleinh. and
a holerythra Ducke, the cotyledons are parallel with the valves.
The flowers of O. holerythra Ducke are still unknown, the spe-
^es differs o^erwise from O. melanocarpa by the much larL
dmiensions of p^ and seeds only. Provisionally this rather dis-
tmct group can best be considered as a fourth section of Ormosia
(.^erican species). Three other sections have been described by

66. The genus Ormosia seems to be absent in Africa (according
to Harms who described 3 nearly allied African genera) The
Asiatic species have been arranged by Prain; I do not know
whether his section Ormosia proper is quite identical with the
section Bico^es Ducke, to which the type species belongs. The
division of Ormosia (American species) can be given as follows-

Standard reflexed, mostly bicallous at the base. Ovary sub-
sessile, densely pubescent. Pod dehiscent. Seed red, with black
spot (the black spot in some species not constant), moderately
compressed. Hilus small. Cotyledons transversal.

Ovary subsessile, densely pubescent. Pod indehiscent, opening
by putrefaction. Seeds orange-yellow, with small hilus. Cotyle-
dons transversal.

Ovary shortly stipitate, glabrous or nearly so. Pod indehiscent.
Seeds brown-red, one-coloured, slightly compressed, with linear
hilus. Cotyledons transversal.

Ovarium breviter stipitatimi, ad suturas tantum pilosum. Se-
mina unicolora, rubra, valde compressa, hilo brevi, cotyledonibus
valvis paralelis.

The 3 latter sections differ as much from the section Bicolores
(Ormosia s.s.) as the genus Ormosiopsis Ducke, which is dis-
tinguished on account of its terminal stigma and globose, one-
coloured, black or red seeds.

Ormosia coccinea Jacks, in Trans. Linn. Soc. X (1810) 360 t.
25; Benth. in Fl. Bras. XV. 1 (1862) 317; — Ormosia subsimplex
Spruce ex Benth. in Fl. Bras. XV. 1 (1862) 316 t. 125; Ducke in
Arch. Jard. Bot. Rio de J^eiro IV (1925) 68.

Bent ham distinguishes an Ormosia coccinea with oblong
leaflets which are minutely pubescent beneath, and an O. sub-
simplex with ovate or broadly oblong leaflets, nearly glabrate
beneath, somewhat smaller flowers and a darker indumentum
of the inflorescence. Those differences are however not constant,

bpruce 2955 [K], xs a plant with old and therefore very rigidlv
coriaceous, shining and nearly glabrate leaflets. Such leLes

cited by Jackson; the other specimen was seen by Jackson
m the herb^Lambert. The flower description id figi^^of
Jack s o n have apparently been made after the latter spSfmen
which could not be traced; Judging from the figure i! amies'
with the narrow-leaved form to which by BenthaJthe
name O. coccinea Jacks, was restricted.nbsp;^ quot; a m the

Distribution: Suriname (o.a. Hostmami 1299 [U; BM.; K- P] type)-
Br. Guiana (Jenman 6569 [K]).

^ber m Bol. Mus. Goeldi V (1907) 398; Ducke in Arch. Jard
Bot. Rio de Janeiro IH (1922) 138, IV (1925) 67

Distribution: Suriname (Pulle 473, Corantijne Riv.); Br Guiana
(Jenman 4171 and 6299 [K]); Para; Amazonas.nbsp;Guiana

Arch. Jard. Bot. Rio de Janeiro IV (1925) 65; - ? Orrnosia
coarctata Jackson m Trans. Linn. Soc. X (1810) 363 fig. 27.

The type specimen of O. coarctata, Anderson from Br. Guiana,
could not be traced in the Br. Mxis. or in Geneva. Possibly it is
identical with O. fastigiata, with a wide distribution throughout
Brazil and the most collected Ormosia species in Suriname. The
description given by Williams in Fl. Trinidad and Tobago
I, 4 (1931) of a fruiting specimen identified by him with O. coarc-
tata also agrees well with O. fastigiata, except for the somewhat
smaller fruits. The inflorescence of O. coarctata however is said
by Jackson to be short and compact (hence the name), while the
inflorescence of O. fastigiata is on the contrary very large. More-
over, O. fastigiata is not yet known from Br. Guiana, nor any
other Ormosia species agreeing with J a c k s o n's description of
O. coarctata, so that O. coarctata Jacks, is still a doubtful spe-
cies.

O. fastigiata is characterized by its thick, sulcate, densely to-
mentose branchlets, relatively long, linear, subpersistent stipules
and tomentose pod. In the type specimen (Claussen [P]) the
stipules are already thrown off, and were therefore not men-
tioned by Tulasne and Bentham. Ducke distinguished
his O. stipularis only on accoxmt of the stipules.

Ormosiopsis flava Ducke in Arch. Jard. Bot. Rio de Janeiro
IV (1925) 61 pi. 25 fig. a, b, Arch. Inst. Biol. Veg. 4,1 (1938) 20;
— Clathrotropis? flava Ducke in Arch. Jard. Bot. Rio de Janeiro
ni (1922) 134; — Clathrotropis? surinamensis Kleinh. in Rec.
Trav. bot. neerl. XXII (1925) 61 fig. 11.

The type specimen of C. surinamensis Kleinh. differs from the
type specimen of O. flava Ducke by its more numerotis, narrower
leaflets and smaller flowers. Subsequent collections of O. flava
by Ducke in Para have shoAvn however that the number as
well as the form of the leaflets are variable. The leaves of some
of these specimens agree entirely with those of C. surinamensis,
as already remarked by Ducke l.c. (1938); they differ by the
somewhat larger flowers only.

Dussia discolor (Benth.) Amsh. nov. comb.; — Geoffroya dis-
color Benth. in Hook. Joum. Bot. II (1840) 91, Joum. Lüxn. Soc.
IV Suppl. (1860) 124; — Dussia cayennensis Harms in Fedde's
Rep. 19 (1924) 293; — Vexillifera micranthera Ducke in Arch.
Jard. Bot. Rio de Janeiro III (1922) 140 with fig.; — Dussia mi-
cranthera Harms I.e. 291.

Type is Martin s.n. from Cayenne. The best specimen is in
Paris, a rather bad duplicate in Kew, while the type of D. cayen-
nensis Harms is apparently another badly preserved duplicate in
Berlin. In Bentham's time the geniis Dussia was stiU unde-
scribed.

According to Harms, D. cayennensis should differ from D. mi-
cranthera in the smaller size of the flowers; the size of the flo-
wers is however too variable as to be of specifical value. More-
over, the type specimen of D. micranthera itself is a small flo-
wered form (Calix sub anthesi circa 8 mm longus, according to
Ducke), while according to Harms in D. cayennensis: Calyx
usque 8 mm vel ultra longus.

Dalbergia glauca (Desv.) Amsh. nov. comb, (non D. glauca
Wallich Cat. (1828) 862 n.n.; Benth. in Joum. Lhm. Soc. sub D.
ovatam Grah. pro syn.); — Ecastophyllum glaucum Desv. in Ann.
Sc. Hist. Nat. Par. 1.9 (1826) 423; Bentham m Joum. Lmn. Soc.
rV Suppl. (1860) 51; — Ecastophyllum foliosum Benth. in Hook.
Joum. Bot. II (1840) 64; — Drepanocarpus falcatus Miq. in Lin-
naea XVm (1844) 476; Benth. in Joum. Linn. Soc. I.e. 71; Pulle
Enum. (1906) 228; — Dalbergia Spruceana Benth. sensu Pulle
in Rec. Trav. bot. neerl. IX (1912) 140 non Benth. 1860; — Dal-
bergia atropurpurea Ducke in Arch. Jard. Bot. Rio de Janeiro

The type, a specimen from the herb. Desv. in herb. gen. Paris,
— from Porto Rico according to D e s v a u x, but probably from
Fr. Guiana, the species is not known from Porto Rico — bears
only one falcate-oblong pod, though D e s v a u x described the
pod as suborbiculate. The type specimen of E. joliosum Benth.
has obliquous-ovate fruits, so that Bentham didl not recognize
it as a member of the section Selenolohium Benth. s.s. (near D.
inundata; with falcate-oblong thick fruits), but placed it imder
Ecastophyllum, treated by Bentham as a distinot genus.
Bentham even says: „Flores E. monetariaequot;, which is not true,
the calyx, the color of the petals (dark violaceous in D. glauca,
white in D. monetaria) and the number of the stamens (resp. 10
and 9) being different. T a u b e r t, in E. P. Nat. Pflanzenfam. Ill,
3 (1894) 385, considered Ecastophyllum as a section of Dalhergia,
and even placed it under the section Selenolohium Benth.

Bentham thought that Drepanocarpus jalcatus Miq. was a
mixtum of D. inundata Benth. (the fruits) and Dr. lunatus (L. f.)
Mey. (the leaves); this is at least not true of the type specimen
in Utrecht, as already remarked by Pulle. I did not see a du-
plicate (according to Bentham transmitted by M i q u e 1) in
Kew. Though the leaflets resemble in form and seize those of
Dr. lunatus, the nervature is much less crebrous.

Owing to the partly incorrect description of Bentham, the
species was again described as D. atropurpurea Ducke. D. revo-
luta Ducke was distinguished by Ducke on account of the
shorter calyx and the coriaceous leaflets. These differences how-
ever do not hold true.

The calyx of the Guiana specimens is about as long as in
D. revoluta, but shorter than in specimens from Para (D. atro-
purpurea). In both regions the length of the calyx is variable.

The leaves appear usually together with the flowers, so that
only fruiting specimens have adult leaflets. In some Guiana spe-
cimens (f. e. Gonggrijp 2237, Jenman 4351, Lanjouw 864 partly)

the old leaves have persisted in flowering specimens; the leaflets
are then rigid-coriaoeous as in D. revoluta. The position of the
leaves in D. revoluta shows that in this case also the leaves have
persisted.

Dalbergia subcymosa Ducke in Arch. Jard. Bot. Rio de Janeiro
III (1922) 144, IV (1925) 74; — Ecastophyllum pubescens D. C.
Prod. II (1825) 421; Benth. in Hook. Journ. Bot. II (1840) 64,
Journ. Linn. Soc. IV (1860) 51; Pulle Enum. (1906) 227.

Dalbergia Riedeli (Radlk.) Sandwith in Kew Bulletin 1931, 358,
non Dalbergia Riedeli (Benth.) Hoehne in Arq. Bot. Est. S. Paulo
I (1938) 27 t. 24; — Ecastophyllum Riedeli Radlk. in Koepf.
Anat. Char. Dalb. (1892) 41; — Ecastophyllum monetaria Pers.
var. Riedeli Benth. in Fl. Bras. XV. 1 (1862) 229 p.p. (quoad
specimen citatimi Spruce 1546 tantimi); — Dalbergia enneandra
Hoehne in An. Bot. Com. Lin. Tel. Mato Grosso Amaz. VIII (1919)
78; Ducke in Arch. Jard. Bot. Rio de Janeiro IV (1925) 74; —
Dalbergia pachycarpa Ducke in Arch. Jard. Bot. Rio de Janeiro
III (1922) 145.

The thick, corky pod and the leaflets which are ferrugineous
pubescent beneath prove that D. Riedeli (Radlk.) Sandw. is iden-
tical with D. pachycarpa. According to Ducke D. pachycarpa
is already a synonym of D. enneandra Hoehne; I have not seen
the type specimen of this latter species myself.

Judging from figure and description and because Hoehne
apparently considers the species as distinct from D. enneandra
Hoehne, it is very improbable that D. Riedeli (Benth.) Hoehne is
identical with D. Riedeli (Radlk.) Sandw. Type specimen of E.
Riedeli Radlk. is Spruce 1546, one of the specimens cited by

Bentham xmder E. monetaria var. Riedelt. The other specimen,
Riedel, type specimen of B e n t h a m's variety, is according to
Sand with l.c. quite distinct from E. Riedeli Radlk., while
Radlkofer, who did not see Riedel's specimen, conjectured
that B e n t h a m's identification of Spruce 1546 was correct. Per-
haps E. monetaria var. Riedeli Benth. is identical with D. Riedeli
Hoehne. The pod of this latter species is not known.

Machaerium isadelphitm (E. Mey.) Amsh. nov. comb.; — Dre-
panocurpus isadelphiis E. Mey. in Act. Nat. Cur. (1824) 807; —
Machaerium angustijolium Vog. in Linnaea XI (1837) 193; Benth.
m Joum. Linn. Soc. IV Suppl. (1860) 55, Fl. Bras. XV. 1 (1862)
236 t. 67; Pulle in Rec. Trav. bot. neerl. IX (1912) 141, aliis auc-
toribus.

Drepanocarpus isadelphus E. Mey. was already cited as a sy-
nonym of M. angustifolium Vog. by B e n t h a m („e descr.quot;) and
by Pulle l.c. Duplicates of the type specimen, Hostmann 629t,
are in Utrecht and Paris.

The Machaerium angustifolium of S a g o t in Ann. Sc. Nat. 6.
13 (1882) 30 is M. altiscavdena Ducke.

Machaerium Kegelii Meissn. in Linnaea XXI (1848) 257; —
Machaerium bracteatum Benth. var. Sagot in Ann. Sc. Nat, 6. 13
(1882) 303; Benoist in Arch. Bot. V. 1 (1931) 140 in key.

The species is easily recognizable by its large bractlets and
long inferior calyx tooth. By the venation of the leaflets it be-
longs to the artificial group Reticulata Benth.

The nearly allied M. bracteatum Benth. (M. marginatum
Standley) from Central America differs by the form of the leaf-
lets and especially by its much broader, very characteristic pod.

Pterocarpus santalinoides L'Hér. ex D. C. Prod. II (1825) 419-
Bak. f in Leg. Trop. Africa I (1926); Hutch, and Dak. Fl W

a second species, named by Huber P. amazorticus. Flowering
specmiens of this second species can only with difficulty be dis-
tmpiished by the longer bracts and bractlets and by the gene-
rally shorter pedicels; the pod however is quite distinct, being
^rky ^d attenuate at the margin only, while the pod of P. Rohrii
has a broad membranaceous wing all around the margin As al-
ready re^iarked by Ducke, the specimens of Spruce cited by
Benthanx under P. Rohrii in reality belong to this second
species. As B e n t h a m only knew the pod of P. Rohrii, the con-
fusion m Bentham's description is visible in ,the words:
Bracteae Uinceolatae-setaceae, caducissimae; pediceUil vel fere 2
^ longi; bracteolae subulatae, calyce paullo vel duplo breviores
W^t .s i^mted here in itaücs refers to P. amazcmicus Hub. only.quot;

in tig. 92 the flowermg specimen belongs to P. amazonicus, the
smgle flower and the fruit to P. Rohrii VaU.

P. amazonicus is however identical with the W. African P. san-
talinoides L'Her.! The name santalinoides was reestablished by
Baker I.e.; Bentham mentions the species still as P. escu-
lentus.

That P. santalinoides occurs in South-America was known to
Bentham, who cites a fructiferous specimen of Martin from
Fr. Guiana, and writes: „Perhaps introduced there by the ne-
groes, who eat the seedsquot;. The range of the species in South
America is however much larger than Bentham suspected, so
that an mtroduction is not very probable. Apparently it is, like
Andira inermis (Sw.) H.B.K., Dalhergia ecastophyllum (L) Taub.
and Machaerium (Drepanocarpus) lunatum (L. f.) Ducke, one of
the species of the Dalbergieae, which are common to tropical W.
Africa and South-America.

Fr. Guiana (o.a. Martin fr. [K], cited by Bentham l.c. under P.escu-
lentus; Sagot 123 fl.; Melinon 2(47 fl. [P], named P. violaceus Vog. by Be-
noist in Arch. Bot. V. 1(1931) 139); Suriname (a.o. Versteeg 232 named
P. Rohrii Vahl by Pulle l.c.; Tresling 472 cited by Ducke under P. amazo-
nicus); Br. Guiana (a.o. Jenman 7260 fl.; Persaud 171 fr. [K]; Im
Thum anno 1879 fl. [K]; Archer 2393 fr. [K]); Para andAmazonas
(Krukoff 5902, 5920, 5923; several specimens distributed by Rio de Janeiro
as P, amazonicus Huber); N. Maranhao (Froes 1948 fl.); N. Matte
Grosso (Krukoff 1622 fr.); Trinidad (Swabey 12607 fr. [K], 2547 fl.
[K].; the species is not mentioned in the Flora of Trinidad and Tobago);
St Vincent (a flowering specimen in Kew, named P. Rohrii Vahl by
Grisebach).

P. Rohrii Vahl has nearly the same distribution (Guiana, Para,
Amazonas, Peru, Trinidad), but seems to be less common, or has
been less often collected because it grows on dry, higher localities,
while P. santalinoides grows along rivers or in swamps.

Except on the characters named above, the two species can
also be distinguished by the following, perhaps not quite con-
stant characters: in P. Rohrii Vahl the leaves are (in sicco) darker
in color, more coriaceous and shining and often subcordate at the
base, with generally a smaller number of primary nerves; and
the indimaentum of tlie inflorescence is browner.

synonym of P. Rohrii Vahl, is not known to me; it is perhaps a
synonym of P. santalinoides.nbsp;pernaps a

P/^eUocarpu. amazcmum Mart. ex. Benth., cited by Bentham
m Fl. Bras Lc. as P. Rohrii Vahl var.? (v. status monstruosus?,
racemi rachide inflato-camosa, legumine incrassato-difformi)quot; A

Pterocarpus amazonum (Benth.) Amsh. nov. comb- _ Phel
locarpusarn^onum Mart. ex. Benth. ia Ann. Wien. Mus. II (1838)

According to Ducke, the deformation '(due to ants) of the
inflorescence is nearly constant in this species. There is how-
ever no reason to regard the pod as difformed.

b/p^^ the name (incorrectly formed) given
by Bentham to Ancylocalyx acuminata TuL in Ann Sc Hist
Nat. 2. 20 (1843) 137 t. 2. It appears from Tu 1 asWs desfrj.quot;

Marshall in Trees of Trinidad and Tobago (1934) 37 with fig.;
— Platymiscium nigrum Ducke in Arch. Jard. Bot. Rio de Janeiro
III (1922) 157; — Platymiscium Duckei Huber var. nigrum Ducke
l.c. rV (1925) 87.

In Arch. Jard. Bot. Rio de Janeiro IV l.c. Ducke writes: „P.
Duckei Huber doit être trés proche du P. trinitatis Benth., j'ignore
queUes sont les differences.quot; Comparing the two species in the
Kew herbarium I found that they agree very well and that on
account of its constantly 5-foliate leaves and nearly glabrous ca-
lyx P. trinitatis must be regarded as identical with the var. ni-
grum Ducke.

In Suriname the var. durum Ducke only has been collected;
the genus seems to be not known from Br. Guiana. This is pro-
bably due to the fact that in flowering specimens the leaves are
stiU undeveloped, so that species of this genus often remain un-
identified.

Lonchocarpus hedyosmus Miq. in Linnaea XVIII (1844) 564;
Benth. in Joum. Linn. Soc. IV Suppl. (1860) 101 passim; Kleinh.
in Rec. Trav. bot. neerl. XXX (1933) 173; — Lonchocarpus se-
riceus H.B.K. var. y? Benth. in Joum. Linn. Soc. Suppl. IV (1860)
89 p.p.; — Lonchocarpus sericeus H.B.K. sensu Pulle Enimi. (1906)
229 non H.B.K. 1824; — Lonchocarpus macrocarpus var. serico-
phyllus Benth. in Joum. Linn. Soc. IV Suppl. (1860) 91; — Lon-
chocarpus paniculatus Ducke in Arch. Jard. Bot. Rio de Janeiro
III (1922) 161, IV (1925) 88.

In his monograph of the genus Lonchocarpus in Joixm. Liim.
Soc. I.e. Bentham mentions this species under three different
names. Lonchocarpus hedyosmus Miq. (type Focke 895 fl. [U]),
was considered! by Bentham as probably not distinct from L.
sericeus H.B.K. Referring to another Siuriname specimen (Host-
mann 234), named by him L. sericeus var. y? (bracteolis parvis)
Bentham remarks: „Not having seen the pod of this plant,

The flower description was made by Kleinhoonte after
a very much insect-eaten specimen, B.W. 6802. This specimen
was not named as type specimen. In the two other specimens,
B. W. 6416, named as type specimen, and B. W. 6932, flower buds
only are present. Fruits are still unknown. It has therefore not
been possible for me to decide whether this is really a distinct
species or identical with L. Urucu Killip et Smilii (L. Nicou
(Aubl.) D. C. sensu Ducke 1922 non D. C.), certainly nearly al-
lied (the higher rotonone content mentioned by Krukoff and
Smith may be due to cultivation), or with L. rufescens Benth. as
suggested by Krukoff and Smith l.c.

It is stUl doubtful whether the three American species admitted
by Bentham and even placed by him in the section Euderris
Benth., really belong to the Asiatic genus Derris. (see Pittier
in Contr. U.S. Nat. Herb. 20 (1917) 41). In the present concep-
tion of the genus Derris in Asia it is however not possible to ex-
clude the American species.

Recently honchocarpus negrensis Benth. has been placed by
Killip in the gentis Derris, imder the name of D. amazonica
Killip. Though this species also has a (distinctly) winged pod, it
differs strikingly from the 3 other American species by the inflo-
rescence and by the form of the standard.

The differences between D. Umgijolia Benth. and D. negrensis
Benth. are not quite clear to me. Bentham distinguishes the
two species on leaf and on pod characters. The pods have the
S£ime dimjensions, but in D. negrensis the pod is coriaceous and
puberulous, and in D. longifolia membranaceous and rufous-velu-
tinous. But perhaps this difference is largely due to the fact, that
in the first case Bentham described an adult pod, and in the

Dipteryx punctata (Blake) Amsh. nov comb •
punctata Blake in Contr U S u L ' quot;quot; ^''^'^rouna
in Notizbl. 121 (19387123nbsp;^^^^^^

Ducke has shown, in various publications, that Cmimarouna
Aubl. 1775 and Tardea Aubl. 1775, considered by Bentham
as sections of one genus, possess so totally different pods that
they have to be regarded as distinct genera. Ducke keeps
A u b 1 e t's names, Coumarouna and Taralea, but Dipteryx Schreb.
1791 is one of the nomina conservanda, and must be kept for the
genus, which was first described by A u b 1 e t: Coumarouna Aubl.

The Suriname specimen agrees with the Amazonian specimens
named C. punctata by Ducke. Though I could not compare the
type specimen (Pittier 6464 cultivated in Venezuela) another cul-
tivated specimen from Venezuela seen in Kew agrees well. The
„Dipteryx odorata Willd.quot; cultivated on some of the W. Indian
islands is often this species.

Poecilanthe Hostmanni (Benth.) Amsh. nov. comb.; — Cyclo-
lobium Hostmanni Benth. in Journ. Linn. Soc. IV Suppl. (1860)
52; Sagot in Ann. Sc. Hist. Nat. VI. 13 (1882) 306.

Legumen oblongum, stipitatum, glabrum, dehiscens, 5-spermum
(teste Sagot), 14—15 cm longum 3—4 cm latum, valvis tordatis,
coriaceis.

The pod of Melinon s.n. was described by Sagot I.e., who al-
ready remarked that, when his identification was correct, the
species could not be retained in the genus Cyclolobium (with in-
dehiscent pod). The 4-fid calyx of P. Hostmanni is also charac-
teristic for the genus Poecilanthe; the species was placed by
Bentham in Cyclolobium on account of the 1-foliate leaves;
the pod was not known to him.

1 (1859) 129 p.p. (quoad descr. tantum, non quoad nomen); non
Centrosemu angystifolium Benth. in Ann. Wien. Mus. 11 (1838)
118; non Clitoria angustifolia H.B.K. Nov. Gen. et Sp. VI (1824)
417.

The flowers of the type specimen of CI. angustifolia (Venezuela,
Orinoco; Humboldt and Bonpland [P]), are, as already remarked
by Kunth himself, badly preserved. Still, it is clear that the
bractlets are falcate-oblong and that the inferior calyx tooth is
lanceolate and much longer than the calyx-tube. The species is
therefore nearly allied to C. pubescens Benth., from which it dif-
fers by its linear and glabrous leaflets only. It occurs in Venezuela,
Colombia and Brazil, and has often been confused with
C. virginianum (L) Benth. var. angustifolium Griseb. (C. pascuorum
Benth.), with subequal calyx teeth.

The form described by B e n t h a m in Fl. Bras. l.c. imider the
name C. angustifolium has large ovate bractlets and short calyx
teeth, differing from C. brasilianum (L) Benth. only by its smal-
ler and narrower leaflets. In Suriname it can not even be distin-
guished as a variety, many Surmame specimens showing both
forms of leaves. Evidently it is only a savannah form of C. bra-
silianum.

Centrosema capitatum (Rich.) Amsh. nov. comb.; — Clitoria
capitata Rich, in Act. Soc. Hist. Nat. Par. I (1792) 111; — Ccti-
trosema virginianum Benth. sensu Sagot in Ann. Sc. Nat. Hist.
VI. 13 (1882) 299; Pulle Enum. (1906) 231 p.p. non Benth. 1837.

Herbaceum. Ramuli volubiles, pubescentes vel demum glabrati.
Stipulae lanceolatae, parvae. Petioli 2—5 cm longi. Folia trifoliata;
foliola ovato-oblonga vel oblonga, apice acuminata, basi rotun-
data, utrinque glabra, rigidule membranacea, reticulata, 4—9 cm
longa 2—5 cm lata. Racemi 3—10-flori. Bracteolae falcato-lanceola^
tae, puberulae, 1—1,5 cm longae 3—4 nmi latae. Calyx pubescens,
tubo — 3 mm longo dentibus valde inequalibus superioribus —
2 mm longis, lateralibus circiter 4 mm longis inf^iore 1—1,5 cm
longo. Petala alba lineis purpureis notata. Vexillum 3—4 cm

longum extus sericeo-pubescens. Legumen lineare ^.U ■
vum, puberulum demum glabratum S^ _ Tw '
natum, 10-15 c^ longum et eirciter e'nL JuT '

®enth. in Seem Bot. Her
(1838) 109, FI. Bras. XV 1 nRi^q^ tAn- rlt; i
carnum TTrK qua ! ^ ' ~ ^alopogonium crrtho-
carpum Urb^ m Symb. Ant. I (1899) 327; Britton and Wilson in
Sc. Survey Porto Rico V (1924) 416

Urban d^inguishes a C. mucunoides Desv. with elongated
long-pedunculate racemes and falcate pods and a C. orthZ^^

Urb. with short, mostly sessile racemes and straight pods. U r-
ban himself gives no distinguishing characters (except in the
name), but no other differences are known to me or appear from
U r b a n's description.

Bentham distinguishes a Stenolobium hrachycarpum Benth.
with elongated racemes and a straight or falcate pod and a va-
riety brachystachyum Benth. (C. mucunoides Desv. cited as sy-
nonym) with short, often subsessile racemes. The two specimens
from Brazil cited by Bentham under the variety I have not
seen, the Central American specimens (according to Bentham
„praesertim in America centraliquot;) have to be reckoned to C. or-
thocarpum Urb.. The type specimen of S. hrachycarpum is from
Brazil.

In the specimens from Suriname the racemes are mostly short
(especially in the upper axils) as well as elongated in the same
specimen, or in some specimens short only or elongated only.
The pod is mostly falcate, but sometimes straight.

The differences seem also not to be sufficiently constant to
justify a separation into 2 species, though C. orthocarpum Urb.
might be distinguished as a variety distributed in Porto Rico,
Hispianola, Cuba, Central America and perhaps also in Colombia.

The type specimen [P] of C. mucunoides Desv. is too small to
be identified with either of the two forms. It bears only one short
raceme with straight pods, and seems therefore identical with the
form described by U r b a n as C. m-thocarpum (apparently B e n-
t ha m's opinion). But as type locality Fr. Guiana is given, and
it may be a part of a plant showing otherwise the characters of
S. hrachycarpum Benth. (C. mucunoides Desv. sensu Urb.).

C. mucunoides Desv. s.l. is a weed which has also been intro-
duced in tropical Asia and Africa.

Calopogonium velutinum (Benth.) Amsh. nov. comb. — Steno-
lobium velutinum Benth. in Tayl. Ann. Nat. Hist. Ill (1839) 437,
Fl. Bras. XV. 1 (1859) 141; — Rhynchosia Luschnathiana Walp!
in Linnaea XTV (1840) 295.

because tlie species is rather rare (only known from Bahia Es-
pirito S^to and Surmame) or because R. Luschmitkmna is given
m the Index Kewensis as the correct name.

moclea megacarpa Rolfe in Kew Bulletm 1901, 139; Wilhams
m Fl. Trinidad and Tobago I. 4 (1931) 238; - Dioclea reflexa
Wook. f. var.? gravdiflora Benth. in Fl. Bras. XV. 1 (1859) 163-
— Dioclea densiflora Huber in Bol. Mus. Goeldi V (1908) 406-
Ducke in Arch. Jard. Bot. Rio de Janeiro IV (1925) 96; - Dioclea
reflexa Hook. f. sensu Fawcett and Rendle in Fl. Jamaica IV 2
(1920) 59 p.p. (p.p. D. reflexa Hook, f.) fig. 18.

nora Benth. [K]), Colombia, Peru, Bolivia, Tobago Tri-
nidad (a.o. Hart 6406 type [K]), Paraguay.

Dioclea reflexa Hook. f. in Hook. Nig. Fl. (1849) 306; Benth in
M. Bras. XV 1 (1859) 162 (excl. var.); Fawcett and Rendle'fI
Jaimica IV. 2 (1920) 59 p.p. (p.p. D. megacarpa Rolfe); Britton
and Wilson in Sc. Surv. Porto Rico V (1924) 418 p.p.; Ducke in
Arch. Jard. Bot. Rio de Janeiro IV (1925) 92, 97 pi 4- Williams
m Fl Trinidad and Tobago IV (1931) 237 p.p. (p.p. D. „iolacea
Benth.).

Distribution: Tropical Asia and Africa, tropical America (Para
Guiana, Jamaica, Porto Rico).

Dioclea violacea Benth. in Ann. Wien Mus. II (1838) 132 • Fl
Bras. XV. 1 (1859) 162; Ducke in Arch. Jard. Bot. Rio de Janeiro
IV (1925) 97 pl. 4; - Dioclea reflexa Hook. f. sensu Williams in
Fl. Trinidad and Tobago I. 4 (1931) 237 p.p.

Distribution: Brazil, Guiana, Trinidad (a.o. Broadway 6448, 9339;
Fendler 315); Central America, also cited for Madagascar and
the Hawaian islands).

The species of Dioclea section Pachylobium are in reality sharp-
ly distinct in flower and fruit characters, as has been shown by
D u c k e m his treatment of the Para species of Dioclea in Arch
Jard. Bot. Rio de Janeiro IV (1925) 93 pi. 4-7. During my stay
at Kew I could study most of the Dioclea material which has
served for the Flora Brasiliensis and for the three recent W. In-
dian floras of Porto Rico, Jamaica and Trinidad, and saw that in
these works the species had mostly been confused.

Dioclea megacarpa Rolfe, described from Trinidad (Rolfe
also cites specimens from Brazil and Paraguay) was not recog-
nized by Huber and Ducke, who described this species as
D. detisiflora Huber.

The two varieties of D. reflexa Hook. f. cited by Bentham
are distinct species. The var. gravdiflora is identical with D. me-
gacarpa Rolfe. Of the specimens named var. glahrescens by B en-
t h a m, two (the Surmame specimen Hostmann 181 [K] and
Spruce 2153 from the Rio Negro) are identical with D. malaco-
carpa Ducke, while I could not identify the third, Gardner
5988, which at any rate belongs to another species.

In the flora of Jamaica D. reflexa and D. megacarpa have been
confused. In the Kew herbarium D. reflexa only is represented
from Jamaica. The figure in the Fl. of Jamaica however clearly
represents D. megacarpa Rolfe, as is shown by the linear pilose
bracts (lanceolate and adpressed sericeous in D. reflexa) and the
pod with straight upper suture (in D. reflexa the sutures of the
pod are both curved). In the description it is said: Branches,
petioles and mflorescences covered with brownish spreading hairs
or glabrate. Bracts long, lanceolate or linear. The words printed
here in italics refer to D. megacarpa Rolfe only.

This description and figure have been of influence on the de-
scription given in Sc. Survey of Porto Rico, though I have from
Porto Rico seen specimens of D. reflexa Hook. f. only.

From Trinidad I saw no specimens of D. reflexa (though it wUl
probably occur there), but several of D. violacea Benth. In the
Flora of Trmidad and Tobago D. megacarpa Rolfe (correctly des-
cribed) and D. reflexa only are mentioned, but D. reflexa is in-

cor^tly described, partly, it seems, because D. violacea has been
confused with it, and partly because of the figure in the Fl of
Jamaica (also confusion with D. megacarpa Rolfe).
The three species can be distinguished as follows:

Dioclea comosa (Mey.) Kuntze in Rev. Gen. (1891) 179; _
Dolichos comosus Meyer Fl. Esseq. (1818) 242.

Kuntze thought that Dolichos comosus Mey. was identical
with D. guianensis Benth., and Bentham supposed (in Hook.

The pod is described by Meyer as foUows: „Legumen sub-
h^osum, oblongum, compressiusculum, 3-4-spermum. Semina
orbiculata, compressa, hylo cincta.quot;

There are two sections of Dioclea in which the seeds are half
surrounded by a linear hilus, Eudioclea Benth. (to which D. gui-
anensis and D. lasiocarpa belong) and Pachylohium Benth. In the
first section the pods are flat-compressed and many-seeded, but
the pod of Paehyhbium agrees with Meyer's description. The

coma e foliolis lanceolatis aggregatis terminatiquot; suggest D. reflexa
Hook. f. 1849.

Stipulae parvae, baud productae. Carina subrostrata. Anthe-
rae omnes fertiles. Legumen oblongum, magnum, dehiscens, val-
vis lignoso-coriaceis convexis. Semina pauca, magna, compres-
siuscula, hilo brevi.

Most of the characters mumerated above are already men-
tioned by Ducke I.e., but the two species were retained by him
in the section Eudioclea Benth. The pod of the section Macro-
carpon is however distinct from that of any of the three hitherto
described sections (Eudioclea Benth., Pachylobium Benth. and
Platylobium Benth.). The anthers are all fertile as in the sec-
tion Eudioclea, which moreover differs by the nearly straight,
obtuse keel with crenulate or fimbriate upper mgrgi-n.

Dioclea virgata (Rich.) Amsh. nov. comb.; — Dolichos virgatus
Rich, in Act. Soc. Hist. Nat. Par. I (1792) 111; — Mucuna virgata
Desv. in Ann. Sc. Hist. Nat. I (1826) 423; — Dioclea lasiocarpa
Mart, ex Benth. in Ann. Wien. Mus. II (1838) 133, Fl. Bras XV
1 (1859) 166 t. 44.

cdli dT'J 21 g^d specimen of the species commonly
called D. la^carpa Benth. Richard's description is very short

C,^gnosus. foh. spicisque hirsutis, foliolis obovatis, abrupTe acu
«bTtlenbsp;^^^^^^^--Itiflora, llgumi^f p^^.

clea species of Fr. Guiana, refer to the common D. lasiocaroa
o^y, m which the pod is covered with short bristly hïrTX

R i c h a rd^s description of Dolichos scaher is as foUows- Sar-
mentis ta punctis elevatis exasperatis, foliolis ovatis

When this IS really a Dioclea species, the description agrees verv
well with D glabra Benth. The only other Diolea s^ecS JZ
glabrous, cor^ceous leaflets, D. macrocarpa Huber, fs not y^

D. glabra and not represented in the herb. Richard.
^iodeaap«ej.is H.B.K. Nov. Gen. et Sp. VI (1824) 438 emend.

^s kterdibus tubo brevioribus inferiore pauUo longiorf. VeS-
V quot;Tnbsp;^nbsp;^sue 5 mm longa lamine

oblonga, subrecta, obtusa, margine superiore crenulata. Antherae
uniformes. Legumen oblongtmi, pubescens, glabrescens, 7—10 cm
longum 2 cm latum, sutura vexillari leviter dilata. Semina oblonga
hilo lineari semicincta.

The type specimen of D. aparensis, in the Paris herbaritun,
from the Orinoco, bears only fruits. Those fruits agree with those
of Versteeg 797 fl. and fr. from Sxiriname. The flowers of Ver-
steeg 797 again agree with those of another specimen collected
at the Orinoco (Chaffanjon 916 [P]). The three specimens also
agree in the weakly developed pubescence of the leaves, but this
character is in Dioclea of minor importance.

The Guiana specimens of D. guiariensis Benth. differ in having
smaller flowers, a pubescent calyx and a densely rufotis, velvety
pubescent, narrower pod. D. lasiophylla Benth. has broadly roun-
ded wings (not narrowed at the apex), velvety pubescent leaflets
and a densely pubescent pod. The species are certainly nearly
allied and may prove to be identical, but the flowers of D. apuren-
sis resemble most those of D. sericea H.BJC., a species with rib-
bed pod.

Phaseolus trichocarpus Wright in Sauv. An. Ac. Habana 5
(1868) 337, Sauv. Fl. Cub. (1773) 30; Britton and Wilson in Sc.
Surv. Porto Rico V (1924) 420; — Phaseolus SchoUii Benth. var.
campestris (Benth.) Hassl. f. guianensis Hassl. in Candollea I
(1923) 463; — Phaseolus productus Ducke in Arch. Jard. Bot. Rio
de Janeiro IV (1925) 99.

Distribution: C uba (Wright [PD; Porto Rico; Suriname; Fr.
Guiana (Sagot 142 [P]; Perrottet s.n. [P], named P. campestris by
Piper; Richard s.n. [P], named Ph. longifalius by Piper); Para (H.J.B.R.
11876 type [U] and H.J.B.R. 17285 [U], cotype of Ph. productus Ducke).

Phaseolus campestris Mart, ex Benth. in Ann. Wien Mus II
(1838) 141, Fl. Bras. XV. 1 (1859) 188; Piper in Contr. U.S. Nat.
Herb. 22 (1926) 678; - Phaseolus Schottn Benth. var. campestris
(Benth.) Hassl. (f. brasiliensis Hassl.) in Candollea I (1923) 464;
Phaseolus juruanus Harms in Notizbl. 70 (1921) 506
Distribution: Suriname. Pard. Amazonas.

fc his revision of the Eastern South-American Phaseoli in Can-
dollea I (1925) Hassler treats Ph. campestris as a variety, dif-
fermg by its smaller flowers only, of Ph. Schottii. In Suriname
there are two distinct forms, of which one is identical with the
var. campestris (superfluously named var. campestris f. hrasi-
hensis by Hassler), tiie other with var. campestris f. guianen-
sis Hassl. Hassler distinguishes the f. guianensis from the va^

There are however other differences on which stress is laid by
D u c k e who described the f. guianensis as Ph. productus. In Ph
campestris the stipules are smaU, adnate or hardly produced
^d the pod is subcylindrical; in Ph. productus the stipules are
distinctly (2-3 mm) produced at the base and the pod is flat-
compressed, shorter and broader.

In Wien. Mus. l.c. B entham cites as type specimen of
Ph. S^ottU: Tejuco, Schott, and as type specimen of Ph. hmgijo-
hus: Brasilia, Schott. In th« Fl. Bras. l.c. Bentham unites Ph.
Schottii and Ph. longifolius under the younger name of Ph longi-
folius, and cites: Brasilia orientalis, Schuch, Tijuca in prov. Rio

de Janeiro, Schott, and Para, Obidos, Spraoe. In the Kew her-
ibarium are present: an incomplete specimen of Schiicht, from
Brasilia, i)erhaps the type of Ph. longijolius Benth., without fruits;
an incomplete specimen of Pohl'(?), from Tepuco, without fruits;
the specimen of Spruce from Obidos, bearing one subcylindrical,
exceptionally long (±9 cm) pod.

If Bentham was correct in identifying iSpruce's specimen
with Phaseolus Schottii and longijolius. Ph. trichocarpus must be
regarded as a distinct species or at any rate as a very distinct va-
riety. It is possible, the type specimen being very incomplete, that
Bent ham's identification was incorrect. Ph. trichocarpus is
not yet known south of Para. The position of Ph. Schottii Benth.
however is still doubtful to me, so that I prefer to treat Ph. cam-
pestris as a distinct species and not as a variety of Ph. Schottii.

Hoewel de genera Cassia L. en Bauhinia L. niet in de ruime
zin van Be n t h a m gehandhaafd kunnen worden, is de systema-
tiek van deze genera niet voldoende ontwikkeld om hierop een
bevredigende indeeling te kunnen baseeren.

Het is ongewenscht in genera, waarvan de zelfstandigheid nog
niet vast staat, dezelfde specifieke namen te gebruiken.

Linnaeus en andere oudere auteurs baseerden vaak soor-
ten op een beschrijving van praelinneaansche auteurs, welke ba-
schrijving zij uitsluitend min of meer volledig citeerden en waar-
naar zij overigens verwezen. Art. 37 van de nomenclatuurregels
kan aanleiding geven tot de opvatting, dat deze soorten niet gel-
dig gepubhceerd zijn of getypifieerd moeten worden met een
(meestal verkeerd gedetermineerd) exemplaar, onder dien naam
in het herbarium van Linnaeus enz. aanwezig.

De verklaring, dievanOverbeek geeft van het inhibitie-
verschijnsel van zijknoppen, is niet aannemelijk.

Strugger heeft waarschijnlijk gemaakt, dat een strooming
in den celwand plaats kan vinden.

Het staafjesepitheel in bepaalde inwendige organen van de Ar-
thropoden moet niet beschouwd worden als een gereduceerd
cilienepitheel; dit laatste ontbreekt geheel bij de Arthropoden.

Hoewel Gäumanner terecht op wijst, dat bij houtschimmels
de optimumtemperatuur voor groei en voor houtvermolming niet
dezelfde behoeft te zijn, heeft hy zulks niet aangetoond.

Er is geen reden Bacterium hegoniae Buchw., Phytomonas
flava Begoniae Wieringa en Bacterium flavozonatum McCuU. als
verschillende soorten te beschouwen.

	D. violacea	D. reflexa	D. megacarpa
Branches and petioles	sparsely pilose or glabrate.	glabrate.	densely pilose.
Bracts	linear-lanceolate, rigid, erect, with adpressed pubes- cence.	lanceolate, quite re- flexed, with ad- pressed pubescence.	densely pilose, li- near, herbaceous, spreading or recur- ved.
Indumentum of the in- florescence.	dark brown	rufous-ferrugi- neous.	ferrugineous.
Blower buds	straight	straight.	incurved.
Pod.	Adult pod nearly glabrate, with straight upper su- ture.	Adult pod nearly glabrate, the sutu- res both curved.	Adult pod still with much pubescence, the upper suture straight.