REINWARDTIA

Published by Herbarium Bogoriense, Bogor, Indonesia
Volume 7, Part 1, pp. 71 - 90 (1965)

REVISION OF

ALBIZIA Sect. PACHYSPERMA (Leguminosae-Mimosoideae)

by

F. R. FOSBERG *)

In the course of revising the genus Serianthes Benth. (Reinwardtia 5:
293—317. 1960) it was necessary to account for several names which had
been proposed in Serianthes but which applied to species belonging to a
group that seems better placed in the genus Albizia. **) After some further
study, the present treatment of this group, here regarded as constituting
Albizia section Pachysperma, was brought together. Owing to the scarcity
of material of most of the species concerned, as well as to the difficulty
in understanding inflorescence morphology in some of the species from
herbarium specimens alone, this paper is of necessity to be regarded as
only preliminary. It may, however, serve to direct attention to the fact
that the species involved do form a coherent group, with a number of
common characters as well as a most peculiar inflorescence that needs
much further study. Two new species and two new varieties, as well as
two transfers, are proposed.

Albizia is a large and complicated pantropical genus, the taxonomy
of which is not yet very firm. Like most genera of the Mimosoideae, it is
ill-separated from its relatives (see Fosberg, Reinwardtia 5: 295—296. 1960).
Its close relation to the much smaller and more coherent genus Serianthes
is through a small group of Australian-Malaysian-Papuasian species, the
best known of which are Albizia lophantha (Willd.) Benth. and Albizia
falcataria (L.) Fosb. (usually called A. jalcata (L.) Backer or A. moluccana
Miq.). The group also includes A. montana (Jungh.) Benth. ex Miq., A.
minahassae Koord., A. fulva White & Francis, and two new species describ-
ed below.

*) Pacific Vegetation Project, U.S. Geological Survey, Washington D.C.
**) This name has usually been spelled Albizzia, but it was originally published

Albizia Durazz. Mag. Tosc. 3, 4: 11. 1772. The wording of the current version of the
Code gives such enormous latitude in correcting orthographic "errors" that, if taken
at face value, a chaotic condition would result. Hence, it seems the part of wisdom
to adhere to strict priority unless an apparently erroneous spelling, or a spelling that
one dislikes, is clearly the result of a typographic error. In this case the spelling
with one z is clearly intentional, as this spelling is consistently used, although the
name of the man in whose honor it was proposed is consistently written with two z's,
"n Sig. Cavalier Pilippo degl' Albizzi".

S— 71 —



72 R E I N W A R D T I A [VOL. 7

These species are characterized by leaves with many very small leaf-
lets, and more especially by having the ultimate branchlets of the inflores-
cence spicate or racemose in the distal portion, rather than capitate or
glomerate as is the case with the majority of species of Albizia. The whole
inflorescence is commonly a panicle or reduced panicle, or solitary or
paired racemes. At least two of the species, possibly three, have a curious
tendency to produce a vegetative innovation at one or more of the principal
nodes of the inflorescence. The flowers are subtended by linear-lanceolate
to broadly ovate bracts, in some species resembling, somewhat, the concave
bracts that subtend the ultimate glomerules in Serianthes.

The spicate or racemose arrangement of the flowers is not unique to
this group in Albizia. A number of New Caledonian species have simple
spikes, often lax and elongate (sect. Spiciflora).

There is also a tendency in this group to have large, thick pods, tardily
dehiscent or even possibly indehiscent. The fruits of the group, except
those of A. falcataria, are not abundantly represented in the collections
available, so the full range of fruit characters is not well enough known.
However, enough is known so it can be said that if the commonly cited
"key" character of an indehiscent fruit were the only thing separating
Serianthes from Albizia the two genera would have to be combined. An
indehiscent fruit, however, is not unique in Albizia to this group, as the
fruit of A. salomonensis White is definitely indehiscent, having the margins
strongly thickened as in some specigs of Serianthes, but definitely not
woody. This species does not resemble the group here considered in any
other important respect, nor does it resemble Serianthes (specimens exa-
mined Walker and White 83 (BRI) (type), Walker and White B.S.I.P.. 11,
(BRI), and N.G.F. 588 (LAE, BRI)). This lack of correlation of fruit cha-
racters together, with combinations of other characters, is not only found
in this genus but seems all too frequent in the Mimosoideae. It makes fruit
characters rather unreliable for separation of groups above the rank of
species, a conclusion which was quite evidently held by Bentham and which
strongly influenced his work on the leguminosae.

The seeds are generally transversally oval or oblong, and thickish,
but they are not known for several of the taxa included in section Pachy-
sperma.

Bentham in 1844 (Lond, J. Bot 3: 85. 1844) placed all the species
Of Albizia known to him to have inflorescences with elongate axes in the
ultimate flower clusters in his section Spiciflorae. This included two new
Caledonian species, A. fulgens (Labill.) Benth. and A. granulosa (Labill.)
Benth. and an Australian species, A. lophantha (Willd.) Benth. Walpers,

1965] F. R. FOSBERG: Revision of Albizia sect. Pachysperma 73

in his Repertorium 5: 595. 1845—6, adopts this name and lists the same
three species. In Walpers' Ann. Bot. Syst. 4: 633. 1857, C. Mueller first
used the name Albizia sect. Lovhanta, but with no diagnosis, only including
in it A. montana Benth. and A. benthamiana Bl. This is, therefore, according
to Art. 34 (5) of the International Code, not considered validly published.
Fournier (Ann. Sci. Nat. 4, 15: 172. 1860) took up this name as Albizia
sect. Lophantha Walp., supplying a diagnosis and listing seven species,
including the three originally placed by Bentham in his sect. Spiciflorae.
Inclusion of these makes Albizia sect. Lophantha a superfluous name for
sect. Spiciflorae, thus preventing its further use.

In a much more comprehensive treatment of the Mimoseae in 1875
(Trans. Linn. Soc. 30: 558. 1875), Bentham put the same three species,
with a number of additional ones, in a new sect. Lophantha Benth., not
mentioning Fournier's use of the name. As this also included the type
of sect. Spiciflorae Benth. (as well as all of its other species), and as
it is a later homonym of Albizia sect. Lophantha Walp. ex Fournier, the
name is doubly illegitimate and may not be used. Included, however, were
two series, ser. Platyspermae Benth., including A. lophantha and A. mon-
tana (Jungh.) Benth. These names were validly published and are legitimate.
Taubert, in Engler & Prantl, Nat. Pfl. fam. 3(3): 106. 1894, adopts Bentham's
1875 arrangement, using the name Albizia sect. Lophantha Benth. essen-
tially without change except to list fewer species. As Bentham's two series
are not here regarded as particularly closely related to each other and
not close to any other section in the genus, they are here treated as
separate sections. For the New Caledonian species with spicate inflores-
cences it is convenient to use the name Albizia sect. Spiciflora *) Benth.,
which can be done by designating A. fulgens as the type of the section.

This section would include at least:
Albizia guillainii Guill., A. granulosa (Labill.) Benth., A. lentiscifolia

Benth., A. comptonii E. G. Baker, A. subfalcata Benth., A. fulgens (Labill.)
Benth., A. rivularis Fourn., A. glandulosa Guill., A. streptocarpa Fourn.,
A. callistemon Guill. & Beauv., A. obovata Benth., A. paivana Fourn.

These are the species with spicate inflorescences listed in Guillaumin's
Flore de la Nouvelle Caledonie 160, 1948, and two others, also from New
Caledonia, that seem to belong here. Their nomenclature and taxonomy
have not been checked. The other section, with a geographic range extend-

*) It seems preferable to use the singular substantive form for sectional names,
and there seems to be no prohibition in art. 21 of the Code against making the change
in ending of Bentham's epithets.



74 R E I N W A R D T I A [VOL. 7

ing from Sumatra to West Australia, New Guinea, and the Solomon Islands,
can be typified by Albizia lophantha (Willd.) Benth. As the sectional name
Albizia sect. Lophantha is illegitimate because superfluous when published,
the epithet Pachysperma *) is elavated from series to sectional rank for
this group.

ALBIZIA sect. Pachysperma (Benth.) Fosberg, stat. nov.

Albizia sect. Lophanta C. Muell. in Walp. Ann. Bot. Syst. 4: 633. 1857, nom.
nud. —- Albizia, sect. Lophantha Walp. ex Fourn., Ann. Sci. Nat. 4, 15: 172. 1860. —
Albizia sect. Lophantha series Pachyspermae Benth., loc. cit. — Type species Albizia
lophantha (Willd.) Benth.

Trees or shrubs, leaves bipinnate with small leaves; flowers in brac-
teate spikes or racemes, these paniculate umbellate, paired or solitary in
axils, the panicles or umbels often bearing a rudimentary vegetative branch;
seeds not strongly flattened.

Range: Indigenous from Sumatra to Western Australia, New Guinea,
and the Solomon Islands.

The seven species making up this section may be separated by the
following key:

A. Indument very close or very sparse, the plant appearing superficially almost
glabrous or very minutely and closely tomentose.
B. Leaflets pointed, ovate-oblong- to lanceolate; inflorescence a panicle of

spikes 7. A. falcataria
BB. Leaflets blunt, oblong; inflorescence of one to several pedunculate spikes

or spicate racemes at a node.
C. Pinnae usually 7—12; flowering racemes 3—6 cm long hairy

1. A. lophantha
CC. Pinnae about 14, flowering racemes 9—10 cm long, almost glabrous.

4. A. melanesica
AA. Indument on young parts abundant, often brownish, plant appearing definitely

pubescent.
D. Flowers in solitary or paired spicate racemes 2. A. montana
DD. Spikes paniculate or umbellate.

E. Calyx teeth unequal, about half the length of the tube
3. A. minahassae

EE. Calyx teeth equal or subequal, about onefourth the length of the tube.
F. Midrib of leaflet very close to distal margin, calyx 2—2.5 mm

long, corolla lobes 2—2.5 mm long 6. A. eymae
FF. Midrib of leaflet about % the way from distal to proximal

margin, calyx 3—4 mm long, corolla lobes 3—5 mm long . . .
5, A. fulva

*) Cf. note on page 73.

1965] F. R. FOSBERG : Revision of Albizia sect. Pachysperma 75

1. ALBIZIA LOPHANTHA (Willd.) Benth., Lond. J. Bot. 3: 86. 1844.

Mimosa distachya Vent., Descr. PI. Nouv. Jard. J. M. Cels 5.20. 1800. (non M.
distachya Cav. 1795). — Albizia distachya (Vent.) Macbr., Contr. Gray Herb. 59: 3.
1919.

Acacia lophantha Willd., Sp. PI. 4: 1070. 1806. — Mimosa lophantha Pers., Syn.
PI. 2: 264. 1807.

Mimosa elegans Andrews, Bot. Repos. 9: t. 563. 1809.
Acacia insignis Hoffmannsegg, Verz. Pflanzenkultur 159. 1824.
Albizia lophanthoides DC., Prodr. 2(2) : 457. 1825.?

Shrub or small tree, young growth very minutely and closely golden
sericeous tomentulose, branchlets prominently longitudinally striate, the
ridges flat-topped, mostly subglabrous, but some tomentulose, the grooves
mostly tomentulose, but occasionally glabrous; leaves bipinnate, pinnae
(6)7—12(13) pairs, leaflets (16)20—40, narrowly oblong, base strongly
oblique, apex slightly or scarcely mucronulate, with 3 strong nerves palmate
from base, the strongest close to but somewhat diverging from distal mar-
gin, petiole with an elongate gland 1—2 cm from base; inflorescence a
spicate raceme 3—6 cm long, paired, axillary, fairly densely flowered,
bracts minute, about 1 mm or less long, ovate, caducous while buds are
still very small, peduncle 10—12 mm long, pedicels very short, about
1 mm; flowers with calyx (1) 2—2.5 mm long, thinly sericeous, lobes tri-
angular to ovate, 1/3 to 1/2 the total length, corolla somewhat more sericeous
than calyx, (4) 6—7 mm long, lobes oblong-ovate, somewhat acute, with
a tiny dense tuft of wool at apex, thinly white-tomentulose within distally,
stamens to 15 mm long, yellowish or chartreuse green; mature pod oblong,
usually acuminate, stipitate, thin, dehiscent, 7—10 cm long, about 1.5 cm
wide, 3—8 seeds in each pod, causing swellings; seed transversely arranged
in pod, oval, swollen, black, faintly margined.

Native of western Australia, and widely cultivated in greenhouses and
in tropical and warm temperate areas of the world.

S. loc, s. coll., "Herb. Pers." (L), "h. Gron. v. H." (L), s. loc, Basil in 1865
(NY) ; Nova Hollandia, " Herb. Splitgerberianum" (L, 3 sheets); Western Australia:
s. loc, Kaspiciv 1 (NY) ; Albany, Meebold 11703 (NY) ; Horton's Siding, Denmark
Railway, White 5396 (NY) ; Swan District, Darling' Range, Pritzel 404 (L) ; Kelmecott,
Canning River, Morrison 104.35 (L).

In addition, Bentham, Fl. Austr. 2: 421—422. 1864, cites the following- collections
which I have not seen:

W. A u s t r a l i a : Goose Island Bay, R. Brown; King George's Sound,
Baxter; Geographic Bay, Fraser; Cape Naturaliste, Oldfield; Swan River?, Drum-
mond. F. von Mueller adds "S.W. Bay" Oldfield.

Cultivated specimens: C a l i f o r n i a : Berkeley, Bracelin 1822 (L) , 2267
(L) ; Ocean Beach, San Diego, Klawe 1591 (NY) ; H a w a i i a n I s . , Oahu,
Makiki, Heller 1949 (NY). M e x i c o : Distr. Federal: Mixcoac, Arsene 8816 (US) ;
City of Mexico, Rose & Hay 5281 (US); Valley of Mexico, Guadalupe, Rose & Hough



76 E E I N W A R D T I A [VOL. 7

4 5 3 8 ( U S ) ; G u a n a j u a t o , D u g d s i n 1 8 9 1 ( U S ) . V e n e z u e l a : C o l o n i a T o v a r ,
1 8 0 0 — 2 0 0 0 m , A l l a r t 487 ( N Y , U S ) , P i t t i e r 9 3 0 2 ( N Y , U S ) . C o l o m b i a : B o g o t a ,
Dugand 3687 (US), Guevara Amortegui 140 (US), Appollinaire s.n. (US); Antioquia,
S a n P e d r o , Daniel 1543 ( U S ) . E c u a d o r : P i c h i n c h a , C h i l l o g a l l o , Firmin 449 ( U S )

P e r u : C a j a m a r c a [ p r o b a b l y e r r o r f o r J u n i n ] , S a t i p o , [ e a s t o f ] H u a n c a y o , Soukup

3373 ( U S ) ; J u n i n , P u c a r a , Ferreyra 2821 ( U S ) . E n g l a n d : L o n d o n , s . coll.

" T o r r e y H e r b . " i n 1821 ( N Y ) . F r a n c e : H y e r e s , H o r t . H u b e r , s . coll. 4692

( N Y ) , 4691 ( N Y ) .

The two last specimens cited are given horticultural varietal names.
The differences between them and other material examined are not very
obvious. Two vaguely defined forms are apparent in the material examined,
one with leaflets 8—15 mm long and 1.5—3.5 mm wide (e.g., the Herb.
Persoon and two of the Herb. Splitgerberianum sheets at Leiden), and
one with leaflets mostly 4—6, or rarely 9, mm long, and 1—2 mm wide
(e.g., the Bracelin sheets and one of the Herb. Splitgerberianum ones at
Leiden). They run together and are probably of no importance. The Klawe
sheet at New York is peculiar. The branchlets are more closely striate than
usual, leaves reach 20 cm long, the 11—12 pairs of pinnae are almost
patent, to 8 cm long, the leaflets small, to 7 mm long, 1—1.5 mm wide,
slightly pointed, patent; pods 8 cm long, 16—18 mm wide, with 10—11
seeds, the base shortly stipitate, apex rounded, with a slight, reflexed
stylar mucro. If there were any evidence that this specimen represented
a population it would be amply distinct for varietal recognition. Arsene
8S16 from Mexico is similar.

Both Willdenow and Persoon cite the later homonym, Mimosa distachya
Vent., in synonymy, so that both must be regarded as simply renaming that
species, and the type of M. distachyum Vent., if it can be found, must be
considered to be that of both Acacia lophantha Willd. and Mimosa lophantha
Pets. Neither Willdenow nor Persoon mentioned the other's publication
of the epithet, so the names must be regarded as independently coined,
though if transferred to the same genus they could not be regarded as
homonyms because of the curious wording of the present International Code.

The plant has been in cultivation a long time but has apparently never
become very common. Edwards, Bot. Reg. 5: t. 361, 1819, says it was disco-
vered by Robert Brown and introduced in 1803 by Peter Good. That this
could not have been its first introduction into horticulture is shown by
the fact that Ventenat's description of it from Cels' garden is dated 1800.
Excellent plates of it have been published by Edwards, loc. cit, Loddiges,
Bot. Cab. t. 716, 1823, Curtis Bot. Mag. NS 5: t. 2108, 1820, and, as Mimosa
elegans Andrews, Bot. Rep. 9: t. 563, 1809. Acacia insignis Hoffmansegg,

1965] F. R. FOSBEKG: Revision of Albizia sect. Pachysperma 77

briefly described from Australia, seems to be the same. His specimen may
have been from a cultivated plant, but this is not clear.

The brief description of A. lophanthoides DC, from Jamaica, fits A.
lophantha so far as it goes, except for "foliolis 12-jugis." De Candolle
places it next to A. lophantha. Neither Fournier nor Bentham disposes of
it effectively. Perhaps it is a cultivated plant of A. lophantha.

2. ALBIZIA MONTANA (Jungh.) Benth. ex Miquel et al., PL Jungh. 1:
267. 1853.

Acacia montana Jungh., Tijdschr. Ned. Ind. 5: 626. 1842; 7(1): 163—164, 186.
1845; Natuur. & Geneesk. Arch. Ned. Ind. 2: 22—23, 35—36. 1845. — Inga montana
Jungh., Topog. et Naturw. Reisen 288, 305. 1845.

Acacia vulcanica Korth., Flora 5: 705. 1847.
Albizia benthamiana Bl. in Miq., Fl. Nederl. Ind. 1(1) : 30. 1855.

Small tree to 10 m, rarely 15 m, tall, to 20 cm, rarely 30 cm diameter,
bark gray, young parts densely, coarsely, and somewhat flocculently rusty
tomentose, pubescence tending to be lost in age, branchlets irregularly
longitudinally ridged, the ridges not conspicuously flat topped; leaves
up to 20 cm long, rachis rusty or grayish tomentose, pinnae 4—10 (12)
pairs, 8—10 cm long, leaflets (26) 32—37 pairs, narrowly oblong, very
oblique at base, apex rounded to somewhat acute, (6) 8—10 (11) mm long,
2—3 mm wide, very sparsely appressed puberulent beneath or completely
glabrous, palmately 3- or 4-nerved from base, main nerve or midrib1/3—1/4
the way from anterior margin; an elliptic raised gland near middle of
petiole; stipules triangular, scalelike; inflorescence a spicate raceme, axilla-
ry, 1—2 at a node, 7—9(12) cm long, peduncle 1—2(3) cm, bracts subtend-
ing flower buds ovate to linear-lanceolate, acuminate, densely golden
appressed pubescent and ciliate, very early caducous, pedicels short but
evident; flowers with calyx about 2.5 mm long, sparsely appressed golden
pubescent, glabrate, teeth 5, triangular,1/3 to 1/2 the length of calyx; corolla
5—6 mm long, appressed pubescent, tube somewhat exserted, lobes oblong,
obtuse to acutish, apex papillate within but without a tuft of wool; stamens
about 15 mm long, bright lemon yellow, anthers minute, orbicular or
even broader than long; pods stipitate, narrowly oblong, 8—11 cm long,
about 2 cm wide, slightly apiculate at apex, margin somewhat thickened,
constricted between seeds when poorly developed, 8—11 seeded; seeds
transversely arranged, oval, black, faintly margined, somewhat compressed,
4—6 mm long.

A common component of the low forest at higher elevations, usually
between 1500 and 3300 m, on mountains of Sumatra and especially Java,
with a variety on Flores. Junghuhn (Reisen, 1845 p. 139, 148, etc.) uses
the vernacular name ,,kamalandingang." *) Junghuhn cited no specimens,

*) Note of the Senior Editor: Kamlandingan is the common vernacular name
for Leucaena leueocephala; A. montana has similar leaves.



78 R E I N W A R D T I A [VOL. 7

but it seems practical to select his collection from Tjerimas in Leyden as
lectotype, as three duplicates of it exist.

This species has been reduced by some to A. lophantha, which, indeed,
it resembles very closely. The principal differences are in the abundant
flocculent rusty tomentum on all young parts, irregular, wrinklelike ridges,
rather than regular, flattish-topped ridges, on the twigs, different leaflet
venation with the midrib away from the margin, much longer racemes with
much larger and longer persistent bracts, corolla lobes without a tuft of
wool at apex but papillose within, pods broader, with 8—11 seeds, rather
than 3—8. The variety from Flores is in some respects intermediate, but
unquestionably belongs with A. montana.

2a. ALBIZIA MONTANA (Jungh.) Benth. var. montana

As described above.

S u m a t r a : s. loc. Korthals (L, 2 sheets); Pajakumbuh, Central Sumatra,
Mt. Merapi, near Bukittinggi, above "Amor Naturae," 1500—1000 m, Meijer 64.32 (L).
W e s t J a v a : Mt. Gedeh: La Riviere s.n.(L) ; Blume [ ?] (L, 4 sheets); 2135 m,
Hallier 513 (BO, L) ; 2135—2600 m, Hallier 535 (BO, L) ; 2900 m, Kuntze 4727 (NY) ;
top, Harreveld 3443 (L, 2 sheets) ; Sapiin in 1896 (BO); 2800 m, Van Steenis 1999 (BO) ;
2390 m, Bruggeman (2) 3717 (BO); Backer 31306 (BO); 2500 m, Kostermans in 1964
(BISH, US, Fo) ; 2900 m, Fosberg U635 (US, Fo); 2600—2800 m, Fosberg U627
(BISH, L, NY, US, Fo); M629 (BISH, NY, US, Fo) ; M63O (Fo). Mt. Pangarango,
Koorders 3865 (BO); Tjibodas: 1400—2900 m, Raap 82b (L); Koorders 766 (BO);
Koordes 2215b (BO). Mt. Wajang, Smith & Rant 589 (BO). Mt. Kantjana, Santosa
Tea Estate, s. coll. in 1953 (BO)."*Slt. Papandajan, Van Rijckevorsel 8 (BO); Koens
US (BO) ; 2650 m, Van Steenis 11670 (BO) ; ± 2650 m, Scheffer C. 29 (BO). Garut,
Koorders 3883 (BO); near Garut, s. coll. 3MS (L). Mt. Tjikuraj, 1500 m, Backer
5378 (BO). Mt. Tjeremai, Junghuhn (L, 3 sheets lectotype) ; 2600—3775 m, Backer
5127 (BO); 3600 m, Docters v. Leeuwen 2532 (BO); Houter 87 (BO); 2600—3000 m,
Backer 5049 (BO). Central Java: Mt. Slamet, 2800—3100 m, Van Steenis 11643
(BO, L); Koorders 3886 (BO); 2900 m, Koorders 3887 (BO); 2700—3100 m, Backer
500 (BO). Dieng Plateau: Junghuhn 80 (L); 2100 m, Brinkman 327 (BO) ; 2000 m, Back-
er 21677 (BO). Mt. Prahu, 1500 m, Koorders 13348 (BO, L). Batang Surdjo, Koorders
13388 (BO, L). Mt. Sumbing, 2100 m, Lbrzing 26 (BO). Mt. Telomojo, Koorders
27827 (BO). Mt. Merbabu: 1500 m, Koorders 3885 (BO); top, ± 3125 m, Docters
v. Leeuwen — Reynvaan 1165 (BO); 2800 m, Docters v. Leeuwen (BO); Docters v.
Leeuwen in 1912 (BO); 1800—2000 m, Backer 30258 (BO). Mt. Merapi: 1900 m, den
Berger 97 (BO); Hemken 13 (BO); Warburg 4642 (NY). Mt. Lawu: De Voogd 809
(L); top, 3200 m, "Alg" 3443 (L); 2900—3200 m, Elbert 155 (L) ; Coert 361; 3265
m, 998 (L) ; 3100 m, 17am Slooten 2558 (BO) ; 1650 m, Backer 6821 (BO); East Java:
Ponorogo, Mt. Manjutan, T. Ngebel, 1450 m, Koorders 3884 (BO, L); 1275 m, Koor-
ders 23154 (BO); 2400 m, Koorders 3888 (BO). Mt. Pitjis, Koorders 29280 (BO).
Mt. Sewu, Koorders 3868 (BO). Mt. Wilis, Backer 11396 (BO); 1500 m, Kuntze
5853 (NY, 3 sheets). Mt. Andjasmoro; Winckel 26 (L) ; 1980 m, Winkel 49 B. (BO).

1965] F. R. FOSBERG: Revision of Albizia sect. Pachysperma 79

Mt. Kawi, Junghuhn s.n. (L); 2850 m, Van Leeuwen 12232 (BO) ; Mt. Welirang, top,
± 3150 m, Posthumus 357 (BO). Mt. Ardjuno: top, ± 3300 m, Posthumus 331 (BO),
2800 m, Arens 28 (L) ; 2500 m, Rant in 1923 (BO); 3000 m, Koorders 43790 (BO).
Mungal-Penandjaan (Tengger Mts.), 2000—2500 m, Kobus in 1899 (BO); 2500 m, Popta
20 (BO); Roomo, Rant 925 (BO). Mt. Kukusan 1500—1700 m, Elbert 154 (L). Mt.
Semeru, 2500 m, Backer 3736 (BO, L). Tengger Mts., Koorders 13807 (BO) ; Koorders
13806 (BO; L) ; 2000 m, Koorders 37836 (BO); 2500 m, Koorders 37635 (BO); Tjema-
ralawang, 2225 m, Koorders 37633 (BO) ; Ajas Ajas, 2200 m, Gisius 45 (L). Mt. Batok,
2200 m, Koorders 37634 (BO) ; Widodaren, Keduwon, 2300 m, Koorders 37632 (BO) ;
Ngadisari, Koorders 37635 (L) ; Koorders 12821 (BO, L). Jang Plateau, Mt. Argapura,
2900m, V. Steenis 10954 (BO) ; 3020 m, Koorders 43540 (BO). Pantjur Idjen, Koorders
28562 (BO); Koorders 3889, 3890 et 3891 (BO); 2160 m, Koorders 3892 (BO). Idjen
complex, 1550 m, Backer 24875 [or 24175] (BO); Teijsmann in 1854 (BO). Idjen,
Mt. Sukat, 2500—2900 m, Van Steenis 12177 (BO). Mt. Merapi, 2600 m, Koorders
43061 (BO). Kawah Idjen, Clason-Laurman E. 58 (BO); Van Steenis 12116 (BO).
J a v a , location doubtful: Bangu, 2150 m, Coert 1020 (L) ; "Alp. orient." Waitz (L) ;
Tjipannas, a warm spring on slopes of Gedeh, Blume (L) ; s. loc. Backer s.n. (BO).
Completely without locality: s. loc. s. coll. (L); Herb. Reinwardtianum (L); "Herb,
variorum Bot." (L); s. coll. IBlumel] (NY); Junghuhn s.n. (L) ; "R or Z" (L).

2b. ALBIZIA MONTANA var. kostermansii Fosberg, n. var.

Ab var. montana differt in racemis longioribus, floribus minoribus,
leguminibus angustioribus.

Leaflets 4—8 mm long, 2 mm wide; racemes 8—14 cm long; flowers
with calyx 2—2.5 mm long, corolla 4—5 mm long, lobes ovate; pods narrow,
15 mm wide, with thick margins.

L e s s e r S u n d a I s l a n d s : F l o r e s , Keli M u t u , 1400 m , O . Jaag
1517 ( L , t y p e ) .

The differences shown by this specimen, together with its geographical
separation from the range of var. montana probably justify the assumption
that it represents a distinctive population.

An attempt to establish the exact place and date of publication of
Acacia montana Jungh. led us into a seldom-equalled bibliographic rat's
nest. This we were able to clear up partially largely through the efforts
of my associate, Dr. M.-H. Sachet. The places of publication indicated by
both Miquel and Bentham were completely erroneous. Examination of a
myriad of papers published by Junghuhn on his explorations of the volca-
noes of Java yielded many places where the name Acacia montana mihi
was used, mostly with no descriptive information at all. Most of these
uses were in articles dated 1845, as were two articles where the name
was accompanied by ample descriptive material in both Dutch and Latin.
There seems to be little chance of more precision as to the order in which



R E l N W A R B T I A [VOL. 7

these papers appeared during that year. To complicate matters Junghuhn
(or perhaps his editor, Nees), in at least four places, in one of these series
of papers, called the plant Inga montana or Inga (Acacia) montana, thus
effecting a transfer. However, it is not certain whether the transfer was
subsequent or prior to the publication of the name with a description. A
solution providing a definite date and place of publication may be reached
by really straining at a technicality and regarding as validly published an
earlier use of the name Acacia montana by Junghuhn, in 1842, accompanied
by the statement ,,10 hoog en hadden stammen van 4—5 duimen dikte."
Many botanists will refuse to regard this as a description, but perhaps if
they are faced with the problem of determining in what order the various
1845 uses of the name appeared they may be more inclined to be technical
and admit that for nomenclatural purposes a description does not have
to be diagnostic. The name Acacia vulcanica presents some of the same
sort of problems, but fortunately it cannot be regarded as anything but a
nomen nudum before 1847. The bibliography of these names, so far as
we have been able to verify it, is as follow:

Acacia montana Jungh., Tijdschr. Ned. Ind. 5(1) : 199 (n. mid.), 622 (n. raid.),
626 (accompanied by the "description" quoted above). 1842. — Acacia montana Jungh.,
Topogr. naturw. Reisen durch Java, 135, 139, 206, 345, 392, 484 (all nomina nuda),
Magdeburg, 1845. — Inga montana Jungh., Topogr. naturw. Reisen durch Java 288
(with a few words of incidental description), 289, 476, Magdeburg, 1845. — Inga (Aca-
cia) montana Jungh. ibid., p. 305. — Inga montana Jungh., Bot. Zeit. 3: 703. 1845 •—•
Acacia montana Jungh., Natuur. & Geneesk. Arch. Ned. Ind. 2: 22—23 (short but
adequate description in Dutch), 35—36 (ample diagnosis in Latin). 1845. Whether this or
the following appeared first is not known, but the editor of this journal says the
material in this article will be published in Tijdschrift van Neerlands Indie. The
time elapsed between writing and printing is not known; in Tijdschr. Ned. Ind. 7(1):
163—164. 186. 1842 (treatment identical with that in Natuur. & Geneesk. Arch.).
Neither of these gives any locality nor cites a specimen, but footnotes in both papers
indicate that where no locality is given, the plants grow on all of the volcanoes.

Acacia vulcanica Korth. ex Zollinger & Moritzi, Natuur. & Geneesk. Arch. Ned.
Ind. 3: 72. 1846 (n. nud., based on Zoll. & Mor. 1939, from Java). — Acacia vulcanica
Korth. ex Hassk., Flora 5: 705. 1847 (no description but the following reference: ,,Zoll.
Ill 72.5 Herb. 1939, Mortz. Verz. p.l. (A. montana Jungh. Arch. II, 22.4. Hasskl.)) •"
The Zoll. Ill reference is that quoted above, but the reference to Junghuhn's species
validates the publication of A. vulcanica and also makes it a superfluous name.)

Inga tengerensis Zoll. (Albizia tengerensis (Zoll.) Miq.) has been in-
cluded by some in the synonymy of A. montana, but its original description
(Natuur. & Geneesk. Arch. Ned. Ind. 4: 13, 81. 1845) indicates that it has
umbellate flowers on filiform pedicels. If the description is accurate the

1965] P. R. FOSBERG: Revision of Albizia sect. Pachysperma

plant must belong elsewhere *). Acacia saltuum Junghuhn (Bot. Zeit. 3: 703.
1845) has also generally been placed in A. montana. This plant was only
vaguely and incidentally described in a translation of one of Junghuhn's
accounts of his explorations. In the statement cited there seems to be
insufficient basis for identifying A. saltuum with Albizia montana, though
I can suggest no other disposition of it. Until a specimen can be found
labelled thus by Junghuhn, Acacia saltuum may best be regarded as a
nomen dubium.

3. ALBIZIA MINAHASSAE Koord. Flora N. O. Celebes, in Med. s'Lands
Pl.tuin 19: 417—419. 1893; Suppl. Flora N.O. Celebes 1: 13—15, pi. 4a, 4b.
1918. — Serianthes minahassae (Koord.) Merrill and Perry, J. Arn. Arb.
23: 393. 1942.

Serianthes ledermannii Harms, Bot. Jahrb. 55: 43. 1917.

Koorders in the two references cited above gives a very full description
of this species and an excellent illustration, drawn from the original mate-
rial. What little can be added to this, or is especially pertinent to the present
study, is given in the descriptions of the varieties below. Koorders had
available material only from Celebes. Subsequent collecting has shown that
the species is widely distributed in New Guinea. As might be expected in
such a widely distributed forest species there is considerable local variation.
It has been possible to characterize the known local populations as varieties,
though the paucity of material, the fact that parts of it were examined
before the whole series of varieties has become apparent, and the lack
of opportunity for direct comparison of some specimens with the others,
make the status of these varieties somewhat doubtful. Further collecting
will probably turn up many more such varieties, or it could show that
some or all of them could not be maintained. In any event, the proposed
arrangement will provide a frame of reference in which to study future
collections and will emphasize the characters and trends that seem signifi-
cant. The main trend seems to be toward reduction of the complex inflore-
scence and suppression of the curious vegetative branchlets in the
inflorescence as one goes eastward in the range of the species. This
reduction culminates in the umbellate to solitary spikes of the related A.
melanesica of the. Solomon Islands, which seems distinct enough to be
considered a separate species. It is realized that these varieties are very

*) Note by the Senior Editor. This species has been included in Abarema sapin-
doides (A. Cunn.) Kosterm.; the type specimen is Zollinger 2521 (P) (cf Kostermans,
Bull 20, Organ. Sci. Res. Indon. 39. 1954).



82 R E I N W A R D T I A [VOL. 7

closely related and that some students may prefer to ignore them and
merely use the binomial Albizia minahassae for the entire assemblage.

The species as a whole may be briefly characterized as follows:

Trees, said to reach 60 m tall, young growth brownish pubescent,
leaves bipinnate 12—16 jugate, pinnules multijugate, leaflets small, more
or less oblique at base, petioles and rachises with disklike glands; inflores-
cence an umbel, or a proliferating panicle, of pedunculate spikes, with
a tendency to produce a rudimentary vegetative branchlet at one or more
ramifications; flowers sessile or subsessile, subtended by caducous bracts;
calyx campanulate, corolla with petals united into a tube below, up to 8
mm long, stamens many, conspicuous, up to 25 mm long; pod very large,
up to 25 cm long, 4 cm wide, subligneous, splitting along margins when
mature but valves not separating.

Koorders, in both of his descriptions, emphasizes the relationship of
A. minahassae to A. moluccana (A. falcataria) as does Harms in his discus-
sion of Serianthes ledermannii. As noted above, they are similar in the
general type of inflorescence, not duplicated in other parts of the genus.
However, A. minahassae differs greatly from A. moluccana in the pubescen-
ce, the umbelloid branching of the inflorescence, the vegetative branchlet in
the inflorescence, the more deeply and irregularly lobed calyces, the thick-
er, more densely sericeous corollas, and in the larger, thicker, incompletely
dehiscent pods which lack a dorsal keel. It is the character of these pods,
which approach the completely indehiscent woody ones of Serianthes, as
well as a similarity in habit, that has led some workers to place this species
in Serianthes.

The following artificial key may aid in separating the varieties of
Albizia minahassae as they are at present understood:

, A. Inflorescence a proliferating panicle
B. Leaflets with midrib equidistant or nearly so from both margins, about

9. mm long and 4 mm wide 3a. var. minahassae
BB. Leaflets with midrib about twice as far from lower margin as from upper,

5—7 mm long, 1.5—2 mm wide 3b. var. proliferata
AA. Inflorescence not conspicuously proliferating, spikes usually umbelloid

C. Inflorescence with a rudimentary vegetative branch at the point of origin
o f the rays o f the umbel; midrib o f leaflet close t o upper margin . . . .

3c. var. umbellata
CC. Inflorescence umbelloid but lacking vegetative branchlet; midrib of leaflet

subcentral or slightly nearer upper margin . . . 3d. var ledermannii

3a. ALBIZIA MINAHASSAE Koord. var. minahassae

Leaf up to 16—18 cm long, one or two disklike glands near base of
petiole, also on rachis, one at base of each pair of pinnae, pinnae 13 pairs,

1965]
F.R. FOSBERG: Revision of AMzia sect. Pachysperma

83

each with up to 20 pairs of leaflets, these about 9 mm long, 4 mm wide
sericeous beneath, tending to be glabrate in age, very slightly hirtellous
above, somewhat oblique, slightly pointed, midrib subcentral or slightly
nearer the upper margin. Inflorescence a proliferating panicle of pedunculate
spikes, varying m degree of development but usually 2—3 times branched
rarely reduced to a single umbel of spikes, normally 2—5 spikes at each
ramification along with a vegetative branch, this usually rudimentary occa-
sionally with a fully developed leaf; flowers sessile or subsessile
disposed along the ultimate 15 mm of the spikes; calyx, 4—5 mm long'
5—6 mm wide, broadly campanulate, 2—3 (rarely 4—5) dentate, appearing
rather bilabiate, teeth strongly unequal; corolla white, 7—8 mm long, tube
not exserted, lobes ovatetriangular, 3—4 mm long; stamens 20—23 mm
long; fruit straight, 11—25 cm long, 23—40 mm wide, subligneous, with
somewhat thickened margins, subindehiscent, splitting along margins when
very mature, but valves not separating, endocarp parchmentlike, tending to
separate from mesocarp when pod is opened, cross veining on sides of
pod visible, veins 3—8 mm apart, tending to anastomose in central region
of pod.

Endemic to Celebes, apparently known only from Menado, in the
Minahassa region on the northeast paninsula, and from Malili, on the
southeast peninsula, ranging from 5 to 500 m elevation. Vernacular names
recorded: ,,paka," ,,teduhu mopute," ,,teduhu pute," and ,,teduhu puti."
Koorders cites two specimens, his nos. 17650 and 17654 with his original
description. Of these the first has been indicated as lectotype.

C e l e b e s : Minahassa: Menado, Koorders 17542, 17552, 22617, 294X2,
(all L) ; 50 m, 17541 (L), 200 m, 17654 (L), 317 m, 22620 (L), 500 m, 17548 (L),
17650 (L, isolectotype); Malili, 450 m, Cel./V-225 (BO, 9 sheets, L, 4 sheets); 5 m,
6.6. 32805 (BO, L) ; Malili, Kampong Angkona, 10—50 m, 6.6. 23972 ( L ) ; 32808 ( L ) ;
32809 ( L ) ; 32810 (L, S I N G ) ; 32811 ( L ) ; 32814 (L); 20806 (BO, L, S I N G ) ; 30 m,
32807 (L, 2 sheets); Malili, Kawata, 400 m, Cel./V-S70 ( L ) ; Cel./V-371 (BRI, SING).
J a v a : Cultivated in Hortus Bogoriensis: s. coll. I.K. 24 (SING); I.B. 50a
( S I N G ) ; Koorders 42809 (L, SING, 2 sheets) ; Koorders in 1915 (L); de Wit in 1941
(L) ; s. coll., s.n. (L, 2 sheets).

3b. ALBIZIA MINAHASSAE var. proliferata Fosberg, n. var.

Ramuli, foliorum rhachides paniculaeque ferrugineo-pubescentes, folio
orum laminae subtus ferrugineo-pubescentes, costae propius marginibus
anterioribus; svicae paniculatae umbellatae cum ramulis sterilibus dispositae.

Tree 20—30 m tall; branches eventually horizontally spreading from
a short gray trunk; young parts and leaf rachises densely rusty-brown
pubescent; leaf rachises 17—18 cm long, pinnae 12—15 or more pairs,
6.5—9 cm long, with disklike glands on rachis between most pairs of pinnae,
leaflets 24—28 pairs, oblong, slightly curved upward, 5—7 mm long, 1.5—2



84 R E l N W A R D T l A [VOL. 7

mm wide, apex rounded, upper surface glabrous, lower densely brown seri-
ceous-tomentose, midrib 1/3 the distance from distal margin; inflorescence
a pedunculate umbellate panicle, peduncle about 35 mm long, several pe-
dunculate spicate branches arising at summit along with a sterile vegetative
branch which originates at the same point, and develops, usually giving
rise to a further umbel of pedunculate spikes, often with a vegetative branch;
fertile part of spikes about 20 mm long; flowers with calyx campanulate,
densely sericeous, 4 mm long, lobes 3—4, ovate, about 2 mm long but
unequal, corolla white, densely sericeous without, glabrous within, about
8 mm long; lobes oblong, 5 mm long; fruit (almost mature) 15—18 cm
long, 3.5—4 cm wide, apparently indehiscent, not woolly, margin somewhat
thickened.

N e w G u i n e a : Papua: Lower Fly River, east bank, opp. Sturt Island,
in rain forest, Brass 8076 (A, BM, BO, type, L, LAE, US); Ihu-Vaitata River, Brass
1107 (A, BRI); Palmer River, 2 m. below junction of Black River, 100 m, Brass 7349
(A, L).

3c. ALBIZIA MINAHASSAE var. umbellata Fosberg, n. var.

Ramuli, rhachides, paniculaeque ferrugineo-tomentosi, foliola 4—6 mm
long 1—2 mm lataque vel parviora, subtus sericea leviter obliqua, costa
propius margini anteriori, inflorescentia umbellata vix proliferata ramulis
sterilibus ferme rudimentaribus; calycis lobi rotundi.

Tree 16—32 m tall, flat-topped; young parts, leaf rachises and panicles
closely brown pubescent or tomentulose, rachises up to 21.5 cm long,
with few disklike glands, only 1-2 near apex and a large one near middle
of petiole, pinnae 12—16 pairs, 5—6 cm long; leaflets about 22 pairs, stiff,
glabrous above, densely sericeous beneath, 4—6 mm long, 1—2 mm wide,
oblong, straight, but the blunt apex toward distal side, midrib close to
distal margin; inflorescence usually umbelloid, a peduncle 6 cm high bear-
ing 3—6 branches up to 7 cm long, the ultimate 1—2 cm spicately florifer-
ous, a small more or less abortive vegetative branchlet originating at the
same point as fertile rays of umbel (this not evident on nearly mature
fruiting inflorescence; on others it is well developed and leafy, making
the pods appear to be on simple axillary spikes; on Brass & Versteegh 13182
and Koster BW 1449 the inflorescences are well developed and show 2
clusters of 3—4 spikes arising one just above the other, the vegetative
branchlet arising between them but from the adaxial side of the upper
cluster); flower with calyx about 4 mm long, divided more than halfway
into 3—4 rounded unequal thinly sericeous lobes; corolla 6—8 mm long,
lobes oblong-ovate, densely sericeous without, glabrous within; fruit straight,
thinly tomentose, 9—17 cm long, 2.7—3.5 cm wide, rather more closely
cross-veined than in var. minahassae, with deep transverse grooves between
seed cavities; seeds oblong, 12—13 mm long, 6—7 mm wide, faintly margin-
ed, rarely 2 in a cavity, cavities otherwise definitely separated.

1965] F.R. FOSBERG: Revision of Albizia sect. Pachysperma 85

Found in West Irian in primary rain forest on flood plains and lower
mountain slopes from sea level to 750 m.

W e s t I r i a n : Idenburg R., Bernhard Camp, 100 m, Brass & Versteegh
1351,6 (BO, type, L, LAE) ; 50 m, Brass 13970 (BO, L, LAE) ; 50 m, 6.6. 25681 (BO,
L, SING), 2 km s.w. of Bernhard Camp, 750 m, Brass & Versteegh 13182 (BO L'
LAE). Mimika, Siere (Uta) , 5 m, 6.6. 32841 (L). Manokwari, Oransbari, 50 m, Mangold
BW 2086 (L, LAE); Momi, 25 m, Kostermans 322 (L, 2 sheets, SING). Salawati
Island: 6 m, Koster BW 1U9 (BO, L, LAE) ; Saloal, 0 m, Versteegh BW 4670 (L, LAE).

3d. ALBIZIA MINAHASSAE var. ledermannii (Harms) Fosberg, n. comb.

Serianthes ledermannii Harms in Bot. Jahrb. 55: 43. 1917,

Tree 20—25 m tall; leaves with glands on rachis between upper pinnae
and at middle or above base of petiole; leaflets oblong to oblong-ovate or
oblong-lanceolate, 5—7 mm long, 1.5—2.5 mm wide, glabrous or almost
so above, puberulent and pale beneath, midrib subcentral or a little nearer
the anterior margin; inflorescence ferruginous villous, umbelloid(?), bran-
ches 2—5, 6—7 cm long; flowers white, subtended by small broadly ovate
acute early caducous bracts. (Description extracted from that of Harms,
no specimens available).

"Nordostl. Neu-Guinea: Pionierlager am Sepik, Sumpfwald (Ledermann
7311)" type, not seen.

This description agrees in every important respect with that of Albizia
minahassae, and I have no hesitation in including Serianthes ledermannii
in that species, even though I have not seen specimens of it. In the absence
fruit, Harms expressed some doubt whether this plant should be in Albizia
or Serianthes and remarked on its resemblance to Albizia moluccana Miq.
(A. falcataria (L.) Fosb.)

4. Albizia melanesica Fosberg, n. sp.

Foliorum rhachides spicaeque subglabrae, rhachides 15—21 cm longae,
foliola maxime 8 mm long a 2.5 mm lataque infra minute sericea; spicae
solitariae vel umbellatae; calyx 3—5 mm longus, lobi rotundi; corolla 8 mm
longa; fructus maxime 21 cm longus 7 cm latusque.

Giant tree 35 m tall, with straight trunk; leaf rachis 15—21 cm long,
almost glabrous, with circular glands at base of distal pinnae, pinnae about
14 pairs, leaflets about 20 pairs on a pinna, oblong, up to 8—10 mm long,
2.5—3 mm wide, minutely sericeous beneath and on midrib above, midrib
1/3 distance from anterior margin; spikes solitary to umbellate, up to 5



86 R E I N W A R D T I A [VOL. 7

originally at a node, to 9—10 cm long, almost glabrous; flowers with calyx
3.5 mm long, deeply divided into 2—3 (?) rounded lobes, thinly sericeous,
corrolla up to 8 mm long, sericeous lobes linear-oblong (?), fruit about 21
cm long, 7 cm wide, distinctly veined, margins thickened, probably indehis-
cent.

The material on which this species is based is inadequate for full
characterization, but its subglabrous rachises and spikes and the reduced
inflorescence suggest that it should be distinguished. Unfortunately when
the specimens were examined, the arrangement of the inflorescence of
Albizia minahassae was not understood, and the notes recorded were not
as full as they should have been.

S o l o m o n I s l a n d s : Bougainville: Koniguru, Buin, 900 m i n rain
forest, Kajewski 2143 (BISH, type, BRI, L, SING). New Georgia Group, Vanganu
I., Keli River, in riverine rain, forest, Walker & White 147 (BRI, K). San Christoval
I., Logie 355 (LAE), 351 (LAE). Malaita I. near Buma Mission, Walker & White
BSIP 72 (BRI).

F.S. Walker, in "Forests of the British Solomon Islands Protectorate,"
p. 136, 1948, has given a description of this plant under the name Serianthes
minahassae (Koord.) Merr. & Perry, and Merrill and Perry have referred it
to Albizia minahassae Koord. which they transferred to Serianthes, incor-
rectly in my opinion.

Two sterile (one with detached pod) specimens from New Britain
possibly belong here, though their leaves are larger, reaching 30 cm long,
up to 20-jugate. *) A detached^'pod on the sheet with White 10964 in Leiden
is somewhat smaller, 15 x 5.5 cm. The spacimens cannot be assigned defi-
nitely until re-collected with flowers and fruit.

N e w B r i t a i n : near Urin, 50 ft. White 10016 (LAE); ridge top east
of Old Mahararu Village, 3 mi. s. of Mt. Otto, Talasea Subdistr. 1000 ft., White 10964
(L, LAE).

5. ALBIZIA FULVA White and Francis, Proc. Roy. Soc. Queensl. 38:
250. 1927.

Large tree, to 43 m tall; branchlets conspicuously zig-zag, angled, leaf
scars prominent, lenticels noticeable, young parts strongly brown-tomentose;
leaves up to 25 cm long, rather stiff, pinnae 10 to 13-jugate, densely fulvous-
tomentose, except on upper surface of leaflets, these up to 22-jugate, oblong
or oblong-ovate" straight or rarely slightly falcate, slightly or not at all,

*) Note of Senior Editor: Leaves of pole stage of A. minahassae have leaves
of 60 cm long, and are up to 24-jugate.

1965] F. R. FOSBEKG: Revision of Albizia sect. Pachysperma, 87

cuspidate up to about 12 mm wide 3/4 mm wide, midrib very near distal
margin, blade sparsely appressed hirtellous above, densely pubescent be-
neath; panicle axillary, heavy, up to 16 cm long, irregularly or racemosely
twice-branched, no trace of vegetative branch (except in Hoogland & Schod-
de 6771), ultimate ramifications shortly spicate, pedunculate, spicate portion
elongating to at least 2.5 cm in fruit, whole panicle pedunculate or sessile
(in type one panicle has a spike at base, other has not), densely fulvous
tomentose, bracteoles ovate, slightly acuminate, 3—4 mm long, very early
caducous; flowers with calyx 3—4 mm long, sparsely to densely sericeous,
subequally toothed to 1/4 total length; corolla about 7—8 mm long, lobes
oblong-lanceolate to ovate, 3—5 mm long, brownish green outside, pale
green within; stamens about 15 mm long, white, style about 20 mm long;
fruit straight, linear-oblong, 12—13 cm long, 21—27 mm wide (no fruit
on type, but a note says pods 1/2 inch by 5 inches), sides very thinly brown
tomentose, margins somewhat thickened, surrounded by a wing less than
1 mm wide on one side, 2—3 mm on other; seeds elliptic, 6 by 3 mm, up
to 14 in pod.

N e w G u i n e a : 3000—5000 ft. in Owen Stanley Range, Trail from
Kokoda to the Gap, Lane-Poole 263 (BRI, type); Papua: Boridi, 5000 ft., Carr 14580
(CANB, NY, SING); Lala River, 5500 ft., Carr 15838 (SING); Territory of New
Guinea: Morobe Distr., Bulolo Valley, 300O ft., McVeigh & Ridgwell 7343 (CANB,
EAE); Western Highlands Distr., Upper Wahgi Valley, Komum Valley, Kulikenda
Village, Hoogland & Pullen 6271 (CANB, LAE).

A specimen from Western Highlands Distr., Wahbag Subdistr., Upper
Tar Valley, near Wahpanamanda, 6800 ft., Hoogland & Schodde 6771
(CANB) differs in having the panicle only once ramified and in having a
small, scarcely developed vegetative branch in the panicle, the spicate
portions of the ultimate ramifications up to 3.5 cm long. It probably should
be included in A. fulva, at least until enough material is collected to show
that it is consistently different. The original drawing of A. fulva appears
to show a vegetative branchlet in the inflorescence, but examination of
the type, from which the drawing was obviously made, shows that this is
simply due to the rachis of the panicle being pressed against an ordinary
vegetative branchlet. Dr. A. Kostermans has suggested in conversation and
on several herbarium sheets that this species be reduced to Albizia falcata
(=A. falcataria). This is certainly the correct disposition of the specimens
which he has so annotated, but they are not identical with the material cited
above as A. fulva.

6. Albizia eymae Fosberg, n. sp.

Ramuli, rhachides vaniculaeque, dense ferrugineo-tomentoso-pilosi,
folia bipinnata 13—14-juga, foliola 22—24-juga mavcime 8—9 mm longa



R E I N W A R D T I A [VOL. 7

3 mm lataque sericea; vanicula 13—20 cm longa biramificata cum ramulo
sterile valde reducto: calyx 2—2.5 mm, longus 5-dentatus dentes minuti
subaequales, corolla extus valde sericea 5—7 mm longa, lobi 2—2.5 mm
longi ovato-lanceolati acuti.

Young parts, leaf rachises, secondary rachises and inflorescence den-
sely brown tomestose-pilose; leaves bipinnate, rachises 21—23 cm long,
pinnae 13—14 pairs, 8—9 cm long, widely separated, with disklike glands
on rachis at bases of most pairs of pinnae, leaflets up to 22—24 pairs, to
8—9 mm long, 3 mm wide, sparsely sericeous above, more densely so
beneath, midrib Vs the distance from the anterior to posterior margins,
tomentose-pilose beneath; inflorescence definitely paniculate, twice ramified,
thyrsoid in appearance but ultimate ramifications umbelloid, each umbel
bearing a very much reduced vegetative branch; flowers with calyx 2—2.5
mm long, thinly tomentose, divided about % to base into 5 subequal teeth;
corolla 5—7 mm long, densely sericeous without, glabrous within, lobes
2—5 mm long, ovate-lanceolate, acute; stamens about 1 cm long.

This species seems about intermediate between A. falcataria and A.
minahassae. It differs from A. falcataria in the umbelloid ramifications and
vegetative branchlets in the inflorescence and from A. minahassae in the
greater ramification and thyrsoid appearance of the panicle, the more
remote pinnae and the form of the corolla, and from both of them in the
almost regular dentation of the calyx. It is closer to A. minahassae but
seems sufficiently^ distinct. Unfortunately the fruit is unavailable. Named
for the collector, the late Pierre Joseph Eyma, who died in a prisoner-of-war
camp in Singapore during World War II.

W. I r i a n: Wissel Lake region, environs of Enarotallf, 23-11-1939, Eyma
54.38 (BO, type, duplicates said to be in K. and L.).

7. Albizia falcataria (L.) Fosberg, comb. nov.

Adenanthera falcataria L., Sp. PL, ed. 2: 550. 1762 (based on Clypearia alba
Rumph. Herb. Amb. 3: 176., t. 111. 1743.)

Albizia'! rnoluccana Miq., Fl. Nederl. Ind. 1: 26—27. 1855.
Albizia falcata (L.) Backer, Voorl. Schoolfl. J a v a 109. 1908; Schoolfl. J a v a

437. 1911; Merr., Int. Rumph. Herb, Amb. 249. 1917. — Adenanthera falcata L. in
Stickm., Herb. Amb. 14. 1754; Amoen. Acad. 4: 124. 1759 (ed. 2, 1788); Syst. Nat.,
ed. 10: 1020. 1759. (based on Clypearia rubra Rumph. Herb. 3: 176, t. 112. 1743).

Large, fast-growing tree, youngest tips appearing tomentulose, when
older, puberulent, twigs irregularly angled, with conspicuous white lenti-
cels; leaves to 40 cm long, petiole with (rarely without) an oval or elongate
diskshaped gland 2—3 cm "from base, pinnae (4—7) 8—10 (11—12) pairs,

1965] F. R. FOSBERG : Revision of Albizia sect. Pachysperma 89

rachises of both orders more or less rusty tomentulose, leaflets 15—25 pairs,
or less, obliquely oblong tending to be slightly falcate, midrib near distal
margin, blade slightly puberulent above, more densely so beneath, acute
to slightly acuminate or subcuspidate at apex, appearing paler beneath:
inflorescence a closely pubescent to glabrate panicle 10—15 cm long, 3—5
times irregularly branched, ultimate branchlets spicate, the spikes elongat-
ing, as flowers mature and drop, to 1—2 cm long, or even longer; flowers
with calyx sericeous, turbinate, 1—1.5 mm long, teeth triangular, 0.5 mm
long; corolla sericeous, 3—4.5 mm long, lobes oblong-ovate, acute; stamens
to at least 1 cm long; flower colour variously described as white, cream, pale
yellow or light green; fruits straight, to 12 cm long and 2.5 cm wide, usually
smaller, with a wing 3 mm wide on ventral margin, shortly stipitate, slightly
beaked, with up to 1.5 seeds.

A common, widespread tree of the Moluccas, New Guinea, New Britain,
and the Solomon Islands, widely planted elsewhere as a very fast-growing
tree used in reforestation.

I n d o n e s i a : Moluccas: Ternate I.: Toramadiahi 300 m, Beguin 1228 (L).;
Molulu U t a r a , Dessa, Fola-Madiahe, 150 m, Bish 6 (L, SING). Ambon I.: s. loc.
Oldenburg 21 (L) ; de Vriese in 1859—1860 (L). Morotai I.: Mangowo R., Kostermans
1372 (LAE). Ceram I.: Oost Ceram, Artafela, 60 m, 6.6. 25812 (L, SING). Banda I . :
G. Api 5 m, 6.6. 13444 (L) ; Pohon Siekat, s. coll. (L), Buru I.: Wae Mole 4 m, 6.6.
24452 ( L ) ; Balos 50 m, 6.6. 31356 (L) ; Hat 5 m, 6.6. 2U56 (L). Halmaheira I . :
Raruba, Djailolo, 100 m, b.b. 23729 (L). Sula I.: Sanana, Tandjung Baleha, 175 m,
6.6. 28767 (L, SING). Mangoli, n. Mangoli, 75 m, 6.6. 29793 (L, S I N G ) , 29863 (L).
West I r i a n : Lorentz R. behind Kloof Bivouac, 30 m, Pulle 95 (L) ; Bendawaja 2,
I. Japen, 10 m, Malinka BW 7023 (L) ; Jappen-Biak, Mt. Hirong near Serui, 10—150
m, Oet & Idjan 755 (L) ; Taniba b.b. 22506 (L) ; Pikpik 200 m, 6.6. 22225 (L) ; Baho,
20 m, 6.6. 21830 (L) ; 4 km s.w. Bernhard Camp, Idenburg R., 800 m, Brass & Versteegh
1314.1 (L, L A E ) ; 6 km s.w. Bernhard Camp, Idenburg R., 1340 m, Brass & Vers-
teegh 12575 (L, LAE) ; Balim River, 1600 m, Brass & Versteegh 11176 (L, LAE) ,
11176A (L, L A E ) ; Manokwari, Prafi, 150 m, Bouwer S3 (L) ; Schram BW 468 (CANB,
L . L A E ) . T e r r i t o r y o f N e w G u i n e a : Near Wau, White 1455 (L, L A E ) .
Sepik Distr., Mapik Subdistr: Prince Alexander Range, s. side of Mt. Turu, Pullen
1585 (CANB, 2 sheets) ; ,,Sepik", Ledermann 6773 (SING). Madang Distr. n e a r Mawan
Village, Gogol Valley (ca. 25 km inland), 60 m, Hoogland 4903 (A, CANB 2 sheets,
L, L A E , U S ) . Morobe D i s t r : Umi R., Markham Valley, 480 m, Brass 32547 ( N Y ) ;
Sattelberg, 1000 m., Clemens 935 (L). T e r r i t o r y o f P a p u a : Sogeri,
Heather 2835 ( L A E ) ; Lower Fly R., Gaima, Brass 8359 (BO, L, L A E ) . Northern
Distr., Tufi Subdistr: along Kopwei R. bet. Aku and Kuruaku, 5—10 m, Hoogland
4388 (A, BRI, CANB 4 sheets, L 2 sheets, LAE, U S ) ; near Ridubidubina Camp,
400 m, Hoogland 4519 (A, BRI, CANB, L 2 sheets, LAE, U S ) . Fergusson I., Angamoia,
200 m, Brass 27251 (A, CANB, L, L A E ) . Koitahi, 1500' ft., Cwrr 12833 (CANB 2
sheets, L, SING 3 sheets, N Y ) . Bisiatavu, 1500 ft., Brass 576 (BRI). Kaiser-Wilhelms-
land, s . loc. Schlechter 16812 (SING). N e w B r i t a i n : Siwai, Waterhouse
9 4 (NY). S o l o m o n I s l a n d s : Bougainville I., Kugu-maru, Bum, 150 m ,



90 R B I N W A R D T I A [VOL. 7

Kajewski 1987 (BRI, L). Ysabel I., Tiratona, 600 m, Brass 3223 (BRI, L ) . Malaita I.,
near Auki, Walker & White 69 (BRI 2 sheets, CNB, L).

Specimens from planted t r e e s : M a l a y P e n i n s u l a : Selangor: Wild
Hill Reserve, Cabin's coll., CF 857 (SING); Kepong Plantation, Hamid 16918 (SING).
S i n g a p o r e : Lornie Road, Sinclair i n 1951 (L). P h i l i p p i n e s : Mindanao:
Cinchona Plantation, Koatoam, Malaybalay, Bukidnon, For. employee 2590 (L, SING).
Luzon: Laguna, Mt. Makiling, 120 m, Canicosa 22874 (L). I n d o n e s i a . S u m a t r a :
Kaban Bjahe, 1420 m, Batten Pooll in 1940 (SING). J a v a : s. loc. Koorders
15643 ( L ) ; Tandjong Priok, 1 m, Backer 32837 ( L ) ; Tjibinong near Bogor, Van
Heeteren in 19S8 (L, SING) ; Mt. Gedeh, Tjibodas, Boerlage ( L ) ; 1S00 m, van Oosts-
troom 13990 ( L ) ; Hort. Bogor in 1903, s. coll. I.B. 60 (NY) ; Kostermans 11152 ( L ) ;
Durttnd 2000 (L). B o r n e o : Sabah: Sandakan: 250 ft., Agama 6715 (L, S I N G ) ;
100 ft., Keith 4964 (SING) ; near Headquarters Forest Office, 100 ft., Hepburn in
1960 (L). W e s t I n d i e s : Cuba, Cienfulgas, Atchison 190 (US). S o u t h
A m e r i c a : Trinidad, P o r t of Spain, Broadway 7687 ( N Y ) ; Broadway 10547 (NY).
H a w a i i a n I s l a n d s : Oahu and Hawaii, seen growing but not collected.
I n d i a : The type locality of Adenanthera falcataria L. is given as " I n d i a . "

This plant was widely known as Albizia moluccana Miq. until Backer
in 1908 applied to it the name Albizia falcata (L.) Backer, based on Ade-
nanthera falcata L. (1754). This was originally based on Clypearia rubra
Rumph. but the plate of Clypearia alba Rumph. was also cited without ment-
ion of the latter name. Linnaeus in 1762 proposed Adenanthera falcataria,
citing Clypearia alba and its plate (111) as its basis, thus leaving Clypearia.
rubra as clearly the type of Adenanthera falcata. So there seems no choice
but to use the epithet falcataria rather than falcata for the plant typified
by Clypearia alba, which is the common Albizia under discussion.

Clypearia rubra, with spiral pods and different leaflets, is clearly
a Pithecellobium (or Abarema).

Kostermans, in his Monograph of the Old World Pithecellobium and
its allies (Bull. No. 20 of the Organisation for Scientific Research in Indone-
sia, Dec. 1954) includes Adenanthera circinalis DC, which is based solely
on Clypearia rubra Rumph., in the synonymy of Abarema clypearia (Jack)
Kost., thereby, by inference including Clypearia rubra in that species. Rump-
hius' plate has leaflets much more like those of A. angulata Benth. than of
P. clypearia. It is certainly an Abarema rather than an Albizia.


	Rein.Vol.7,Part 1,pp.1-90_Page_41
	Rein.Vol.7,Part 1,pp.1-90_Page_42
	Rein.Vol.7,Part 1,pp.1-90_Page_43
	Rein.Vol.7,Part 1,pp.1-90_Page_44
	Rein.Vol.7,Part 1,pp.1-90_Page_45
	Rein.Vol.7,Part 1,pp.1-90_Page_46
	Rein.Vol.7,Part 1,pp.1-90_Page_47
	Rein.Vol.7,Part 1,pp.1-90_Page_48
	Rein.Vol.7,Part 1,pp.1-90_Page_49
	Rein.Vol.7,Part 1,pp.1-90_Page_50
	Rein.Vol.7,Part 1,pp.1-90_Page_51