



■ïj\'-.\',	v w-quot;.

. i -
■ \' V

v ^

\'■V3

-J

TER VERKRIJGING VAN DEN GRAAD VAN
DOCTOR IN DE WIS- EN NATUURKUNDE
AAN DE RIJKS-UNIVERSITEIT TE UTRECHT.
OP GEZAG VAN DEN RECTOR-MAGNIFICUS
Dr. B. J. H. OVINK, HOOGLEERAAR IN DE
FACULTEIT DER LETTEREN EN WIJS-
BEGEERTE. VOLGENS BESLUIT VAN DEN
SENAAT DER UNIVERSITEIT TEGEN DE
BEDENKINGEN VAN DE FACULTEIT DER
WIS- EN NATUURKUNDE TE VERDEDIGEN.
OP MAANDAG 31 OCTOBER 1927.
DES NAMIDDAGS TE VIER UUR

Gaarne maak ik gebruik van de goede gewoonte, die
aan een promovendus de gelegenheid biedt in officieelen
vorm de momenten uit zijn studietijd naar voren te brengen.

Dat ik tenslotte de planten-systematiek en -geographie
als mijn hoofdvak heb gekozen, vindt voor een belangrijk
deel zijn oorzaak in het enthousiasme, waarmede. Gij Uwe
studenten tracht te bezielen, door hen zelfstandig aan
onderdeden der flora van onze Koloniën te laten werken.
Gij maakt hun dit mogelijk, door een kostbare verzameling
materiaal, het herbarium en een steeds meer volledig ge-
worden bibhotheek als werkmilieu aan te bieden.

Steeds zijt Gij echter bereid hen aan de bestudeering
van onze eigen flora te zetten, een onderwerp, dat altijd
Uwe belangstelling heeft. Toen ik dan ook eerst bij U
kwam met een onderwerp over de genetische planten-
geographie van Nederland was U hiervoor dadelijk te
vinden. Achteraf bleek, dat de volledige opzet zooals ik
mij dien voorstelde, verschillende jaren arbeid zou kosten.
Met de grootste bereidwilligheid heeft U mij toegestaan
het overzicht der Nieuw-Guineesche Bignoniaceae, hetwelk
ik voor mijn doctoraal-examen samenstelde, uit te breiden
tot een bewerking dezer familie voor het geheele Maleische
gebied. Ik ben U voor den steun en belangstelling, die Gij
mij hierbij betoonde zeer erkentelijk.

Gedurende het jaar, dat ik het voorrecht had Uw assistent
te mogen zijn, heb ik getracht van Uwe bewonderens-
waardig uitgebreide phytographische kennis te proiiteeren.
Dit jaar zal voor mij van beteekenis zijn in mijn toekom-
stigen werkkring.

Hooggeleerde Went. In dit korte bestek kan ik onmo-
gelijk alles samenvatten, wat Gij voor mij gedurende en
voor mijn studententijd hebt beteekend. Reeds in mijn H.B.S.-
tijd moedigde Gij mijn belangstelling voor de botanie aan
en had ik het geluk met U in persoonlijke aanraking te
komen. Dat ik tenslotte de planten-physiologie niet als
mijn hoofdvak heb gekozen, ligt zeer zeker niet aan de
wijze waarop Gij de Utrechtsche biologen leiding geeft
op de moeilijke wegen der physiologie. Mijn wetenschap-
pelijke vorming heb ik allereerst aan U te danken.

Meer en meer heb ik een bijzonder groote waardeering
gekregen voor de méér dan vaderlijke instincten, die Gij
voor de biologen bezit en waarvan ik in zoo ruime mate

De hartelijke vriendschap die Gij en Mevrouw Went
mij in Uw gastvrij huis hebben betoond neem ik als een
waardevol iets mee naar de tropen.

Hooggeleerde Nierstrasz. Gij hebt mij op dc U eigen
kritische wijze de sobere resultaten der biologische weten-
schap aangetoond. Gij zijt in dat opzicht een gevaar voor
mijn enthousiasme geweest, maar ik ben U daar zeer

Hooggeleerde Jordan. De wijze waarop Gij weten-
schap en dillettantisme weet te ontmengen uit de troebelen
der dieren-physiologie zijn ongetwijfeld van belang geweest
voor mijn wetenschappelijke opvattingen.

Hooggeleerde Rut ten. Uw heldere en aantrekkelijke
colleges heb ik tot mijn spijt maar ten deele kunnen volgen.

Hooggeleerde Westerdijk; De boeiende wijze waarop
Gij de phytopathologie voordraagt en de belangstelling
voor Uwe leerlingen heb ik steeds zeer op prijs gesteld.

Uwe vriendelijke belangstelling, hooggeleerde De Graaff,
en de leerzame gesprekken, die ik met U mocht voeren in
de laatste jaren, zullen mij steeds in aangename herinnering
blijven.

Deuren, die ik met U Zeergeleerde Entz, door mocht
brengen met de bestudeering der Protisten hebben door
Uw groote kennis en bereidwilligheid haar waarde voor
mij verkregen.

Voor de hulp, die Gij, Zeergeleerde Smith en Gij,
waarde Any Joles, Uit tien en Zorn mij bij het
samenstellen van dit proefschrift hebben verleend, ben ik
U zeer erkentelijk.

De duidelijke illustraties, waarde A. de Bouter, geven
het resultaat weer van de groote zorg door U aan mijn
schetsen besteed.

De finantieele hulp, die Gij Excellentie, Minister van
Onderwijs, Kunsten en Wetenschappen, mij gedurende
mijn studietijd hebt willen verleenen, heeft het mogelijk
gemaakt, dat ik mij aan de studie der biologie heb
kunnen wijden.

Hooggeachte Stakman. Ik zal niet trachten U als mijn
vroegeren leeraar en vriend hier aan te spreken, daar dat
in eenige woorden toch niet mogelijk is. Dit proefschrift
zij mede Uwe voldoening. De groote vriendschap, die ik
van U in Uw huis ondervond, zal voor mij van groote
beteekenis blijven.

Gij, mijn aanstaande vrouw, hebt met onuitputtelijk
geduld dit proefschrift helpen voltooien. Uw medeleven
in mijn werk zal voor ons beiden later van groote waarde
blijken te zijn.

My publication on the Bignoniaceae of New Guinea
has been the prime cause to study this family for the
whole Malayan region. So this study forms a parallel to
the publications of the families worked out by the botanists
at Buitenzorg, which are an attempt to constitute a modern
review of the Malayan flora.

That a review of this family was necessary may appear
from a short historical survey.

History. Since the publication of „De flora van Neder-
landsch Indiëquot; by Miquelquot;), only 4 of the 5 Javanese
species have been discussed byKoorders and Va 1 eton ;
Koorders^) gives only a short enumeration of the 5 Java-
nese species, together with the cultivated ones of Java.

Boerlage^) discussed the genera and gave an account
of the species known then, but did not give a critical dis-
cussion of the species.

For the extra-Javanese species there are only the elder
works of B1 u m e and M i q u e 1 and for Deplanchea
the descriptions of Scheffer^).

As the genera and species were known only incompletely,
the geographical distribution could not be indicated with
any certainty and about the affinities was not known much
either.

Geographical distribution. The remarkable distribution
of the genera and species in the Malayan region have
made me study especially the affinities of these with those
of the surrounding countries and with allied genera. As it
appeared to me that the Bignoniaceae of the Malayan
region sensu stricto, without Malacca, the Philippines and
New Guinea, are so closely connected with those of the
surrounding countries I have tried to treat a more or less
phyto-geographical unit, which extends from Malacca to
New Caledonia. With some genera it became necessary to
enlarge the region for the understanding of the geogra-
phical distribution of the genera or species, and so the
reader will find some Chinese, Burmese and Indian species
in some genera. The Umits for a „Flora Malesianaquot; will
be very difficult. Years ago H. Zollinger^) has given
some critical remarks about the area used by Miquel.
Since Miquel, affinities of the Malayan flora are more
closely studied by several Dutch botanists and by Ridley,
Merrill, etc. It has appeared that Malacca, Sumatra,
Banka, Billiton, Riouw and Borneo show many phyto-
geographical affinities, whereas Java and the isles eastern
of it do not belong so close to this unit. The Philippines,
though undoubtedly allied with the Borneo-Sumatra-Ma-

\') Ann. Mus. Bot. Ludg. Bat. 1 (1864) 197. 3 (1867) 249.
3) Nat. Tijdschr. Ned. Ind. 31 (1870) 332.

Over hct begrip der omvang eener Flora Malesiana. Nat. Tijdschr.
Ned. Ind. Ser. III. 3 (1856) 293-322.

lacca-group, and on the other hand with the Moluccas,
show many endemics and form in several respects a dis-
tinct phyto-geographical area.

The Moluccas show already signs of Australian influen-
ces. and New Guinea forms in many respects an area
containing a mixed Malayo-Australian flora. On the other
hand this enormous island contains a mass of endemics,
wich do not occur neither in Malaya nor in Australia and
forms also a unit. New Caledonia, of which R. Schlech-
ter^) made a plant-geographical study, shows close affi-
nities with Australia, just as one should expect, but con-
tains a large endemic development like nearly all isolated,
old isles.

After this discussion one will agree with me that it is
difficult to limit the Malayan region, though a political
unit, in plant-geographical respect. For different families
it will not be the same.

For the 5/^noniaceae-genera I have used a very large
area; whereas, however, within this area many genera
are limited, it forms in my opinion for my aim a genetic-
plant-geographical unit. Thus it is explained that for some
genera e.g. Deplanchea, Oroxylum, Tecornanthe. Pan-
dorea, Radcrmachera, Nyctocalos, the treatise has become
a floristic-monographical revision, though perhaps incom-
plete, due to lack of materials.

An other cause that I have used such a large area, is
because of the fact that whereas many genera are limited
to this area, it would be of interest to make more general
conclusions in connection with the geological theory of
A. Wegener-) and the genetic-plant-geographical ones

Pflanzen-geographischc Gliederung der Insel Neu Caledonia. Inaug.
Diss. Reprint. Engl. Bot. Jahrb. 36 (1905) 1-41.

of J. C. Willis^) and A. Engler. 2) In relation with
the remarkable endemism of a lot of species, an interes-
ting result could perhaps be obtained.

Scheffer\'\') makes already some remarks about two
Deplancheas, D. bancana {Scheff.) vSts. and D. tetra-
phylla {R. Br.) F. v. Mueli: „Het is een zeer opmerkelijk
feit. dat, zoo niet dezelfde, dan toch een zeer verwante
species van een genus, dat onder zijn verwante scherp
is gekarakteriseerd en daarvan door den habitus zeer af-
wijkt, voorkomen in twee, in vegetatie zoo zeer verschil-
lende landen.

Wel zijn de vleugels van de platte zaden zeer voor-
deelig voor de verspreiding, doch in geene der tusschen-
liggende streken is de plant waargenomen; ofschoon
sommige daarvan nagenoeg geheel bekend zijn.quot;

Seemann^) when discussing the native country of Teco-
maria capensis Seem., also says of the remarkable question:
quot;At first sight it would appear that the question respecting
the native country could easily be settled by assuming the
species to be endemic in both Africa and America, were
it not opposed to the fact that all Bignoniaceae, notwith-
standing their winged seeds, have a limited geographical
distribution, and that no species, as far as I know, has
been claimed as a citizen of both hemispheres.quot;

Bureau pointing to the very interesting Bignonia-
ceae of New Caledonia, also remarks the typical distribution
of the species. He says: ..Les Délostoma sont américains,
les Catalpa sont les uns de l\'Amérique. les autres de la

1)nbsp;Age and Area. A study in geographical distribution and origin of
species. 1922.

5)nbsp;Het geslacht Diplanthera Banks et Solandcr. Nat. Tijdschr. Ned.
Ind. 31 (1870) 336.

Chine; il est donc remarkable de trouver, dans les trois
espèces de Bignoniacées, de la Nouvelle Calédonie, trois
types d\'organisation différents : l\'un special à la Nouvelle
Hollande, la terre la plus voisine, qui produise des Big^
noniacées, mais les deux autres rappelant des flores très
éloignées : le premier identique avec un type indien, le
dernier se rattachant à des formes plus spécialement
américains. Quelque peu nombreuses que soient les Big-
noniacées de la Nouvelle Calédonie, elles ne sont pas
sans intérêt au point de vue de la géographie botaniquequot;.

The origin of the Bignoniaceae of New Caledonia must
indeed be explained in connection with those of Australia
and New Guinea, though it possesses an endemic develop-
ment, no species being found elsewhere.

In my opinion the Bignoniaceae, with their disjunct areas
and their endemism, notwithstanding the —apparently— easy
manner of spreading by the wind, are especially fit to
give some points of view about their origin and distribution
in general, based upon geological theories.

Something will be said also about adaptation in some
genera, e. g. Tecomanthe and Dolichandrone and of parallel
forms in Tecomanthe.

The geographical distribution of most of the genera is
to be found on my maps. They are founded on both
herbarium-materials I saw and literature. The limits, of most
of the areas are in general rather well established, but it
Will appear probably that those of the Moluccas, Borneo,
Celebes and New Guinea are not satisfactorily known
nowadays. The numbers of the areas accord with those
of the species of the text. Of a single species, e. g. Haplo-
phragma macroloba (Miq.) vSts., the finding-places are
markedj^h a

M a. Pandorea austro-calcdonica Bur,: b. Dolichandronc spathacca
• {•) K. Sch.; c. Dcplanchca spcciosa Vicill.

Materials studied. The materials of New Guinea and
of the herbarium at Utrecht I had at my disposal by the
kindness of its Director Prof. Dr. A. Pulle; I was very
much obliged by the materials I received from the respective
Directors of the herbariums at Leyden, Buitenzorg, Dahlem
and Kew. Prof. H. Lecomte granted me his hospitality
in the Paris Museum to study materials and rare literature.
Further I am greatly indebted to Prof. Dr. W. M. D o c t e r s
van Leeuwen (Buitenzorg). who lent to me the Big-
noniaceae, he gathered in New Guinea last year: to Prof.
E. D. Merrill (Berkeley), who was so kind as to give
me some information about PhiUppine plants.

I am greatly indebted to all of them; without their aid
the results would not be so complete as I hope they are now.

All materials were dried specimens, except a few New
Guinean ones gathered by Docters van Leeuwen. So,
figures, descriptions, etc. are all made after dried materials,
in connection with the remarks of the labels.

It is a remarkable fact that the collected specimens
never show capsules, with only few exceptions Flowers,
too, are often lacking. This was especially striking with
the materials I saw from the Forest Research Institute
(Boschbouw-Proefstation) at Buitenzorg: e. g. those of
Radermachera gathered in Borneo and Sumatra.

It is a pity that without flowers, the bulk of the Big-
noniaceae cannot be identified. The leaves of Bignoniaceae-
species of one genus, those of different genera and even
those of different divisions are hardly distinguishable. So
determinations after sterile materials are better to be kept
in petto. Moreover it occurs that some Bipnoniaceae-genera
in sterile state have often been mistaken with other families
and so on the contrary. E. g. Oroxylum with Araliaceae

As to the species, it is the same question, e. g. in Deplan-
chea, Radermachera, Tecomanthe, etc.

Further there is the difficulty how to distinguish parallel
forms. E. g, I found this in Tecomanthe; T, nitida vSts.
and T. ar/afci vSts, show a more than striking resemblance,
so that one would even suppose them to belong to the
same species. Further examination showed, however, that
there are differences in the calyx and the corolla. The
geographical distribution also pointed to the acceptance of
2 different species, T. nitida vSts. being found in the Mt.
Wichmann (Central N. Guinea), T. arfaki vSts. being col-
lected on the Arfak-Mts., the complex on the north-western
peninsula. Both at high altitudes of about 3000 m. Now
these mountains are only connected via the Mt. Warikeri
of about 3^700 m. high. So T. arfaki vSts. and T. nitida
vSts., both alpine plants, are isolated from each other, but
show a remarkable resemblance as to their habits. Now
the ovary has shown that the species are of different
origin; T. arfaki vSts. has certainly affinities with § Den-
drophilae. whereas T. nitida vSts, shows the most close
alliances with T, saxosa Diels.

Sub-spccies, varieties, forms, races. I have followed
most of the systematists in regard to the most rational
conception of these different forms.

Sub\'species form in my opinion together a Linneont:
however, they show many differences, though not important
enough to distinguish them as distinct species. Most times
there are links between the sub-species. They have a varia-
bility-curve of their own.

Oxera Labill. was first arranged as a doubtful Bignoniaccac by
Billardi^re (1824) and so Faradaija F. v. Mueli (Fragm. 5 (1865)21.

sidered as sub-species, but often these units are too small
for this. It is a well-known fact that plants of one finding-
place are very much alike, more than those of different
finding-places; the resemblance of specimens of one species
is mostly reversely proportional with the distance between
the finding-places. The closest resemblance is found among
specimens of one collection. This is among other things
due to the fact that the individuals of one finding-place are
grown up under the same conditions, probably belonging
to one or few „reine Linienquot;. Moreover specimens of one
collection are preserved in the same manner.

I accept varieties, when specimens differ in one heriditary
character, morphological or physiological.

I accept a forma for specimens, certainly differing in
one or more characters, which are not heriditary and
appear to belong to the natural variability-curve of the
species. They are due to external influences.

Though all this is rather well-defined it is very difficult
to distinguish them in dried materials; one ought to cultivate
many specimens of different areas. On the other hand the
routine of a florist gives help to discover the forms,
varieties, etc. also with dried materials. Besides the culti-
vation of rare species from isolated regions is rather
impossible.

Anatomical researches. I did not make any anatomical
investigation, however remunerative it would appear for the
Bignoniaceae, amongst which many lians are known. First
I had no materials of several species and only a complete
account of the anatomical structure of all species would gain
scientific value. On the other hand anatomical differences
were not necessary for the recognizance of the species,
the morphological differences being sufficient.

Descriptions, nomenclature, figures. In the descriptions
I have tried to give a most complete account of the
variability of the species and I have mentioned small, but

New or altered diagnoses are in Latin, according to
the rules adopted by the international congresses of Vienna
(1905) and Brussels (1910). It was necessary therefore to
alter some of the names I gave in quot;Nova Guineaquot;.

If I knew the type of the genus or the type specimen
I have cited it behind the description. Figures and descrip-
tions are made if possible after these types. For very
variable species it is impossible to give a satisfactory diag-
nose and in this case I availed myself of all specimens.

When I have accepted characters, diagnoses and specimens
I have not seen, the author is cited in brackets behind.
Of all (and only of) the specimens I examined, the her-
barium where they are preserved is mentioned behind it
in brackets.

The abbreviations of the herbariums are to be found
at the end of this introduction.

Dates of collection are not mentioned for convenience\'
sake: often a date is missing on the labels. Mostly it is
known, when a^ certain collector gathered in a certain
area.

As to the geographical distribution I have divided the area
in sub-areas, which are in general enumerated from east to
west and from north to south : British India—Burma—Indo-
China—China—Malay Peninsula—Nicobars and Andamans
—Sumatra—Riouw Archipelago—Banka—Billiton—Java,
Bali, Lombok, Timor — Borneo — Celebes — Philippines —
Moluccas—New Guinea—Australia—New Caledonia. This
differs a little for different species, according to their
peculiarities.

Of every specimen I first mentioned the finding-place
with contingent particularities; all this and the name of
the collector are separated by a colon. The collector is
cited with the collection-number. If this was missing I

As to the nomenclature I accepted the names given by
B. D. Jackson^). The geographical names of the Dutch
East-Indies are to be found in the dictionary of Ch. F.
H. Dumont-). Since the latter work has been published,
the government of the Dutch East Indies introduced other
rules for spelling However rational the new names may
be and though at Buitenzorg one has to accept them, I
have not made use of this new list. Firstly, the Hst contains
but the principal names for the present and further the
new names are not to be found in the large atlasses as
e.g. Stieler\'s Hand-Atlas, etc.

The names of the species often occur in the text without
author; the latter is to be found at the top of the diagnose.

I often cited native names from literature, but I also
used those of the labels. How far they are right I cannot
judge. The collector and finding-place are always cited,
so that they may be corrected later on.

Herbarium-names (nomina nuda) are accepted as synonyms
or as new species, firstly in appreciation of the authors,
on the other side for completeness\' sake.

Keys. I have made 2 keys. The first represents a rather
natural scientific one, based upon important systematic
characters. The second key is for the use of a studentin
Malayan plants. It contains also the often cultivated
species of the Archipelago. It is based upon sterile cha-
racters if possible, as in most of the specimens capsules
and often also flowers are missing.

3)nbsp;Lijst van de voornaamste Aardrijkskundige namen in den Ned. Ind.
Archipel. Landsdrukkerij Weltevreden (Java) 1923.

Species, which are only cultivated in the Buitenzorg
botanical gardens are not mentioned.

with 2 fertile thecae . . 2. Oroxylum (p. 816)
/3. Calyx small 0.2-0.3 cm. 1., stamens 4 fertile,
didynamous, one cell developed, calcarate
at the base .... 3. Millingtonia (p. 825)
B. Septum perpendicular to the valves

a.nbsp;Septum terete, sometimes notched, spongy,
a. Ovulus in one row on each placenta, germ of

the winged seeds trigonous, sunken in not-
ches of the septum 15. Stereospermum (p. 946)
/?. Ovules in many rows on each placenta,
germ of the winged seeds flat, septum
slightly bubbled. . 16. Radermachera (p. 953)

a. Septum quadrangular, false septum present,
other septum sometimes reduced.
I. Calyx spathaceous.

1.nbsp;Corolla-tube short, concealed in the
calyx, not white, hmb bilabiate, flowe-
ring at day-time. . . Markhamia(p. 929)

2.nbsp;Corolla-tube long, funnel-shaped, much
exceeding the calyx, white, fragrant,
flowers nocturnal, limb almost acti-
nomorphic.

X. Septum reduced, false septum large,
leaves scattered, opposite or whorled,
coriaceous. 13. Dolichandrone

§ Coriaceae (p. 931)
X X. Septum nearly as broad as the false
septum, leaves always opposite,
mostly membranaceous. 13. Doli-
chandrone § Membranaceae (p. 936)

Heterophragma(p. 999)
P. Capsule 2-locular, false septum absent.
I. Calyx spathaceous.

1.nbsp;Only 1 row of ovules on each placenta.
X. Lian, calyx without glands, cymes

cauliflorous, corolla distinctly 2-la-
biate, leaves opposite, pinnate,
thecae glabrous . Perichlaena(p. 840)
X X. Shrubs or herbs.

□. Erect shrubs, calyx mostly glan-
dular, tubular, campanulate, 5-
toothed, leaves opposite, simple
or pinnate, thecae hairy

11. Stenolobium(p. 904)
Cn. Herbs, calyx splits open to the
base, with 5 lobes, leaves scattered

X. Trees, shrubs or herbs.
□. Capsule oblong, leaves large,
simple, opposite or verticillate,
trees . • . 12. Deplanchea (p. 906)
Cn. Capsule linear or oblong-linear.
A. Thecae calcarate, herbs annual
or perennial, leaves simple or
1-pinnate . . Incar villea (p. 835)

Calyx tubular-campanulate
coriaceous, capsule very long,
glabrous little trees, thyrses
cauliflorous . Kigelian the (p. 1020)

Calyx campanulate, not coria-
ceous, mostly large trees,
with rather dense pubescent
or tomentose inflorescences,
f. Disk cushion-shaped, sep-
tum thin, capsule very long,
spirally twisted, inflore-
scences cauliflorous

ff. Disk annular, septum rather
thick, spongy, capsule
mostly curved, not spirally
twisted, inflorescences not
cauliflorous, terminal or
lateral

*. Calyx regularly 5-toothed,
campanulate, when in flower
inflate, disk cupular, 8 series of
ovules per cell Podran.ea(p. 839)

Calyx campanulate, 5-toothed,
not inflate when flowering,
disk flat dish-shaped, many
rows of ovules per cell

Camp sis (p. 837)
AA. Calyx coriaceous or membra-
nous, without glands, climbing
with the ultimate twigs.
Corolla actinomorphic, disk
cupular, leaves 3-foliate or
digitate.

9. Neosepicaea (p. 899)
Corolla more or less zygo-
morphic, disk annular, seldom
sub-cupular, leaves 3-foliolate
or 1-pinnate.

f. Calyx small up to 0.7 cm.,
truncate, mostly indistinctly
5-toothed, mostly thyrses
(seldom racemes: P.austra\'
lis s. sp. linearis), corolla
small up to 2.5 cm. (seldom
longer: § Leptophyllae and
§ Grandiflores), mostly
bearded at the throat and
on one side inside the
corolla, not annularly stu-

7. Pandorea (p. 833)
ff. Calyx large, 5-lobed (sel-
dom truncate: T. auran-
tiaca), 1.5-4 cm. 1. (seldom
smaller: T. montana, T.au-
rantiaca), corolla large 6—9
cm. (seldom smaller: r.mon-
tana), neither bearded at
the throat nor on one side
inside the tube, but dis-
tinctly (except T. ternaten-
sis) stupose near the inser-
tions of the stamens

THE PHILIPPINES, NEW-GUINEA, AUSTRALIA
AND NEW-CALEDONIA, WILD AS WELL AS CULTIVATED.

1.nbsp;Leaves 1-jugate, mostly with tendrils ...nbsp;2
Leaves 3-foliolate............ 3

2.nbsp;Tendril 3-forked, leaflets ovate to ovate-lan-
ceolate, flowers red-yellow, stamens inserted
at the middle of the tube, cultivated plant

4.nbsp;Pyrostegia (p. 829)
Tendril simple or none, leaflets obovate-
oblong, flowers purple-violet, stamens inserted

5.nbsp;Corolla valvate. red, leaves 3-foliolate, disk
cupular, Australian climber

7.nbsp;Leaves 3-foliolate, corolla mostly large, flowers
in racemes, S. E. Asiatic and Malayan climber

8.nbsp;Calyx large, 1. 5—3 cm. 1., 5-lobed, corolla
large, 5—8 cm. 1., axillar racemes, seldom
thyrses (see also Pandorea leptophylla. Pan-
dorea Curtisii and P. jasminoides pp. , ),
Malayan and New-Guinean climbers

8. Tecomanthe (p. 864)
Calyx small, truncate, ca. 0.3 cm. 1., corolla
small ca. 2-cm. 1., thyrses, seldom racemes
(see also Tecomanthe aiirantiaca and Teco-
manthe montana pp. , ), Malayan, New-
Guinean and Australian climbers

9.nbsp;Corolla with a transverse anterior plait, leaves
simple or 3-foliolate, flowers 2.1abiate, culti-
vated small trees, fruit a calabash

11.nbsp;Flowers cauliflorous, small shrubbish tree, ca.
3 m. high, capsule linear-cylindrical, candle-

like, cultivated tree . . 18. Parmen tiera (p. 1008)
Flowers in terminal or axillar inflorescences 12

14.nbsp;Thecae pubescent, tree 5 m. high, leaflets
serrate, flowers yellow, calyx glandular, cul-
tivated tree......11. Stenolobium(p. 904)

15.nbsp;Calyx very large and broad, coriaceous, sta-
mens 5, capsule flat, broad very large 1 m.

1., Indo-Malayan tree . . . 2. Oroxylum (p. 816)
Calyx rather small membranaceous, stamens
as a rule 4, didynamous, capsules always
smaller, terete............................16

Septum entire, at most slightly bubbled, seeds
flat, germ not sunken into the septum, ovules in
00 rows on each placenta, Chinese. Indo-
Malayan trees ... 16. Radermachera (p. 953)

In Journ. Bot. Néerl. 1 (1801) 366; Bureau. Monogr.
(1864) 52; Miquel in Ann. Mus. Bot. Lugd. Bat. 1 (1864)
201; ibid. 3 (1867) 249; Seemann Journ. Bot. 8 (1870) 147;
Benth. 6 Hook. Gen. PI. 2 (1876) 1040; K. Schumann
in Engler-Prantl Pfl. Fam. IV. 3b (1895) 219; Boerlage
HaiTdl. Fl. Ned. Ind. 2 (1899) 588.

This genus was founded by Teysmann and Binnen-
dijk on a flowering Javanese lian discovered by the former
on the southcoast of Java near Wijnkoops Bay.

As there have come more materials and species certainly
belonging to the. genus, the diagnosis must be altered at
some points, which I stated here.

Descr. emend. Frutices volubiles vel scandentes, glabri;
folia opposita 3-foliolata, foliolis integerrimis, petiolulatis,
ovatis vel lanceolatis; flores in racemis in apicibus ramulorum
brevium foliigerum dispositi; calyx tubuloso-campanulatus
brevis, truncatus, 5-dentatus; corolla mediocris vel magna,
tubo longe cylindraceo fauce campanulato-dilatato, hypo-
craterimorpha vel elongato-infundibuformis regulariter 5-
lobata lobis rotundatis, stamina 4 fertilia didynamia quinto
rudimentario vel 5 fertilia subaequalia vel inaequalia binis
anterioribus quidpiam longioribus, corollae tubo inserta atque
mclusa; antherae biloculares, tortiles, lamellatae connectivo
excedente; discus hypogynus carnosus annuliformis; ovarium
oblongum compressiusculum quadrangulare, biloculare; ovula
oo in utroque loculo dissepimenti marginibus 2-seriata inserta.

Fig, 1, Geographical distribution of Nyctocalos T, et B.
1. N. Thomson! Hook, f. 2. N. shanica Gregor et Smith. 3. N. brun-
felsiaeflorus T. et B, 4. N. cuspidatum (Bl.) Miq.

stylus elongatus filiformis, stigmate bilamellato ; capsula
lanceolato-siliquaeformis, plana, septifraga, valvis coriaceis
glabris interdum obtuse carinatis, septo subcoriaceo ; semina
plana alata tenuiter membranacea praedita.

(Tribus Bignonieae. Spec. Asiaticae
austro-orientales, Malayenses et Philippinenses.

Type spec.: N. brunfelsiaeflorus Tcysmann
et Binnendijk).
As the first (1861) species described, N. brunfelsiaeflorus
T. et B. and N. cuspidatum (Bl) Miq. (1867) possess five

fertile stamens and N. Thomsoni Hook. f. (1865) is didy-
namous with a fifth staminodium, it seems to be necessary
to expand the generic diagnosis. N. Thomsoni Hook f.
is nearly allied to N. cuspidatum [Bl] Miq. but differs at
first sight in the stamens and is further absolutely separated
in geographical distribution.

How Schumann could suppose that the punctual original
description of Teysmann and Miquel might be wrong.
I cannot understand. He thinks that it may appear that
N. brunfelsiaeflorus T. et B. and N. cuspidatum (Bl) Miq.
have four fertile stamens and a great staminodium. I proved
this on the species, but found of course his supposition wrong.

However it is remarkable that here in a Bignoniaceae-
genus, wherein the species are closely allied, an origin of
didynamy is to be seen.

I investigated the flowers of the here mentioned species.
The stamens of N. brunfelsiaeflorus and N. cuspidatum
are indeed equal in length: with a slight indication of 2
longer anterior and 3 shorter posterior ones. The Burmese
species N. sharjica I know but from the description; it
has 2 shorter (anterior) stamens and 3 longer (posterior)
ones, whereas N. Thomsoni is didynamous.

Taxonomical the latter represents a derived species with
an intermediate state to the original scheme in which 5
equal stamens must be supposed. I do not mean that N.
shanica is an intermediate form. At least N. Thomsoni
shows in the other points and in habitus a close alliance
with TV. cuspidatum. This 1 have demonstrated in the
following scheme.

If we want to explain the geographical distribution, we
are obliged to accept, that:

1.nbsp;N. cuspidatum is the eldest species and at the same
time the ancestor of the 3 others, or:

2.nbsp;N. brunfelsiaeflorus represents the ancient type and
has given appearance to the others, or:

I neglect the possibility that Nyctocalos could have a
polyphyletic origin, as the species show such close alliance

N. brunfelsiaeflorus—N. shanica—N. Thomsoni
I--^ N. cuspidatum.....................................

In this case it should be very remarkable, that N. cus-
pidatum and N. Thomsoni, which are closely allied, must
be understood as parallelly originated forms.

At any rate we are obliged to suppose that the ancient
Nyctocalos was widely distributed in S. E. Asia and the
Malayan Archipelago.

For the rest we can suppose that the areas yet known
are incomplete. The regions in S. E. Asia, Java and the
Philippines however are for a good deal worked out to-
day. It may be said that the flowers of Nyctocalos are
nocturnal. Among the materials I studied the bulk however
was in a flowering state.

In my opinion I may conclude that the species of
Nyctocalos do not represent progressive endemism, but
on the contrary are relic-endemics on account of their
taxonomy and geographical distribution;

Eu-Nyctocalos Seem, including the species with 5 fertile
stamens: N. bmnfelsiaefloms and N. cuspidatum.

Hausmannia {F. v. M.) Seem, including N. Thomsoni
and Hausmannia jucunda F. v. M. both with 4 fertile
stamens and a rudiment. The latter Austrahan lian being
first described by F. v. Mueller. (Fragm. Phytogr. Austr.
4 (1864) 148).

Seemann considers Hausmannia identical with Nycto-
calos. However he was truly wrong in his supposition.
Hausmannia jucunda remains still the only species of a
distinct monotypic genus Hausmannia F. v. M., differing
principally from Nyctocalos in the long exserted stamens
and the valvate corolla. Yet Hausmannia shows a remar-
kable conformity with Nyctocalos, being both lians, in
the 3-foliolate leaves and the habitus.

1.nbsp;Stamens 4, didynamous with a fifth staminodium, co-
rolla very large ca. 18 cm., white . \\. N. Thomsoni.
Stamens 5, fertile...............2

3.nbsp;Corolla short 5-7.5 cm., pale sulphureous, calyx 1-1.5 cm.,
teeth triangular acute .... 3. N. bmnfelsiaefloms.
Corolla large 15-20 cm. white, calyx 1.5-2.5 cm., teeth
calcariformous, linear......4. N. cuspidatum.

1. Nyctocalos Thomsoni Hook f. Bot. Mag. t. 5678;
C. B. Clarke in Hooker Fl. Br. Ind. 4 (1885) 376; =
N. assamica Hook. f. (nom. nud.) K. Schumann in
Engler-Prantl. IV. 3b (1895) 231.

High glabrate climber, leaves opposite 3-foliolate, entire;
folioles ovate to oblong-lanceolatc caudate acuminate ca.

15-7.5 cm. Flowers large ca. 17.5 cm. 1., white in long
ca. 15 cm. peduncled terminal cymes; calyx minute, sub-
campanulate. equally 5-dentated; corolla elongate infundi-
buliformous with a terete tube slightly arcuate; limb
5-lobed, lobes sub-equal, the 2 inferior ones slightly
greater, broad imbricate; stamens 4 with a fifth rudiment,
at the throat of the corolla inserted, didynamous, included,
filaments filiform-subulate, thecae divaricate pendulous at
an apiculate connective; disk cushion-like; opary bilocular;
style elongate filiform; stigma bilamellate; ovules numerous
at the margin of the placenta in 2 rows inserted; capsule
lanceolate, plane, septifragal; valves in the middle cari-
nate. leathery, almost woody; septum ultimately free, simple,
flat; seeds 0.8-1.2 cm. diam., or including the wing
2.5 cm. diam.

Geogr. distcib. iAssam — Mikir Hills: Simons; Gow-
hatty Hills: Jenkins (Hooker I.e.) — Assam — Herb.
Bernhardi: H. amp; S. (D) — Java (cultivated) — Iter War-
burgianum n. 1676, cult, in Hort. Bog. sub n. 1676. (D);
cult, in Hort. Bog. sub. n. X. F. 150 a. from the Bot.
Gard. Calcutta (from Assam) (B).

This species is closely allied with N. cuspidatum and
differs in the stamens and the longer peduncled cymes.
Moreover N. Thomsoni has much shorter 0.5-1 cm.
petioled leaflets and these are long acuminate and
opaque; the capsule is shorter ca. 15 cm., and broader,
ca. 5 cm.

With N. assamica Hook f. (nom. nud.) mentioned by
Schumann, the latter means the Assam species. However
he points out: „hält die Mitte in der Länge der Blumen-
krone zwischen N. brunfelsiaeflorus and N. cuspidatumquot;.
a Statement that is undoubtedly wrong.

C. B. Clarke (in Hook. I.e.) draws from the didynamy
the remark: quot;The Assam species constitutes a marked
sub-genusquot;. As I have stated the four species form a

well-defined genus in which the species are closely allied
and that shows a tendency to didynamy, a fact which
appears in several Bignoniaceae-Qemia.

2,nbsp;Nyctocalos shanica Me. Gregor et W. W. Smith
in Rec. Bot. Surv. Ind. 4 (1911) 280.

A high glabrous climber, bark grayish much covered
with warts; leaves opposite 3-foliolate, leaflets ca. 12 cm.
1. broad elliptic, acuminate, entire, with 6-8 pairs of side-
nerves; petiole 4-5 cm. 1., petiolules 1-4 cm. 1. Flowers
ca. 1 cm. pedicelled in sub-corymbose pauci- (4-7) florous
inflorescences. Calyx short 0.7-0.8 X 0.4-0.5 cm., campa-
nulate, truncate equally ca. 0.1 cm. 1. filiform dentate.
Corolla white with a 5-6 cm. 1. cylindrical tube towards
the throat campanulately ampliate; hmb equally 5-lobed,
lobes 1-1.5 cm. 1. ovate, obtuse, non-apiculate; stamens
5 fertile, 2 shorter, in the middle of the corolla-tube in-
serted and just reaching the throat. Anthers glabrous with
oblong sub-parallel cells. Ovary on a very short gyno-
phore, style filifojm with a bilamellate stigma; capsule and
seeds unknown.

Geogr. distr. Burma. — East of Keng Tung. Southern
Shan States. Upper Burma at an elevation of 330 m.:
Mc. Gregor. n. 672.

This species is more closely connected with those of
the Malayan Arch., especially with N. brunfelsiaeflorus,
than with N. Thomsoni. The only representative of the
genus within the Indian boundary.

3.nbsp;Nyctocalos brunfelsiaeflorus Tcysm. ct Binnend.
-Miquel in Journ. Bot. Neerl. 1 (1861) 367; Miquel.
Choix des PI. rar. ou nouv. 1863. t. VII; Miquel. Ann.
Mus. Bot. Lugd. Bat. 1 (1864) 201 excl. syn. Blumei et
stirpe celebica; ibid. 3 (1867) 248. t. 8A; Koorders Exc.
Fl. Java. 3 (1912) 183.

Descr. emend. Frutex volubilis, ramis ramulisque tere-
tibus, glabris. griseis; petiolo 5-8 cm. longo, semiterete,
superne obsolete angulato vel teretiusculo. basi et in
geniculis incrassato: folia ternata; foliola varia ovali-
oblonga. falcato-cuspidata. basi inaequali-rotundata, arti-
culata, supra obscurius viridia. 9-12 cm. longa. 4-6 cm.
lata. 1-2 cm. petiolulatum. petiolulis supra aanaliculatis:
racemus laxus. lateralis terminalisve. pedicellis bracteolatis.
teretibus. laevibus; calyx 0.5 cm. longus. campanulatus.
truncatus extus adnato-5-dentatus. viridus. margine purpu-
rascente, intus laevissimus; coro//a hypocraterimorpha alba,
dein sulphurascens. regulariter quinque-lobata. lobis obtu-
sissimis. rotundatis. basi angustatis. subauriculatis, subtus
multiglandulosis, tubo longissimo, cylindraceo. superne am-
pliato, laeve. violascente; stamina 5 subaequalia, tubi parti
tertiae superior! villosae affixa, inclusa, fertilia, filamentis
teretibus basi incrassatis; antherae profunde bipartitae
apice connectivo linguaeformi excedente affixae: pollen
ellipsoideum; pistillum staminibus longius, glabrum, disco
annulari crenulato integrove cinctum; ovarium oblongum
4-angulatum, glabrum, superne in stylum attenuatum.
biloculare, dissepimentis utrinque in marginibus ovulis
horizontalibus anatropis obsessis; fructus oblongus ca.
12 cm. longus, 4 cm. latus, utrinque rotundato-cuneatus,
apice vix vel modeste acuminatus, valvis coriaceis,
extus (in sicc.) fusco-virescens medio longitudinaliter sul-
catus, intus luteus medio lintatus, sulcis indistinctis grosse
reticulatis.

Geogr. distr. — Java — Java occidentalis: van Hasselt
(L sub n. 898, 100... 102); Cult, in Hort. Bog. sub n.
X. F. 138 a. (B, L sub n. 920, 240... 43 and 920, 240... 44).
Cult. Hort. Bog. (ex Wijnkoops Bay) sub n. 2248 HB
(B); Palaboeanratoe in Res. Preanger: Kds. 34438/? (spec.

ster.) (L. B)alt. 50 m.: Java: Dr. Ploem (B): Wijnkoops
Bay ad litora (Teysmann).

Though Mi quel had already described the capsule,
his description is wrong because his specimen was the
Celebic form N. cuspidatum, which I tested on his original
specimens.

Miquel first thought the Javanese climber identical
with the sterile specimen of B1 u m e from Celebes, described
by Blume as N. cuspidatum. Indeed it is remarkable that
the foliage of the Ni/cfoca/os-species is very difficult
or impossible to distinguish, nor in form neither in ner-
vation.

The colour of the corolla in Miquel\'s table (Choix)
is wrong, not being pink but sulphureous to white.

Probably all the herbarium materials have descended
from the only place hitherto known pr. Wijnkoops
Bay in the south of the Res. Preanger near the coast
up to 50 m. alt. Teysmann has already introduced
it into the Hort. Bogor. where it is cultivated with
success. The fr^rant flowers open at 8 o\'clock in the
evening and drop already in the early morning. Notwith-
standing that, it never produces capsules there, as far as
I know.

N. brunfelsiaeflorus in only found in the cited region
in Java and there strongly endemic. With respect to the
geographical distribution of the other species it stands
totally apart.

Nyctocalos cuspidatum (Bl) Miq. — = Tccoma
cuspidata Bl. Rumphia 4 (1848) 35; Miquel Ann. Mus.
Bot. Lugd. Bat. 3 (1867) 249, t. 8 B; - = Nyctocalos
macrosiphon Teysm. et Binnend. (nom. nud.) in Cat. Hort.
Bogor. (1856) 155; — = Nyctocalos brunfelsiœflorus Teysm.
et Binnend. (quoad spec, celeb.) Miquel in Ann. Mus.
Lugd. Bat. 1 (1864) 201; Merrill Philip. Journ. Sci. 1
(1906) Supp. 237; C. B. Robinson Philip. Journ. Sc. 6

(1911) Bot. 211; Merrill. Enum. Philip. Fl. PI. 3 (1923)
443 — = Beaumontia trifoliata (In herb. Paris).

High climber with terete lenticellate branches; leaves
3-foliolate, opposite, petioles ca. 9 cm. 1. sub-terete-angulate
with a thickened base; folioles elliptic-ovate, sometimes
oblique at the base, modest to long acuminate at the apex,
the terminal leaflet more or less obovate, lucid above,
pale beneath with 6—8 pairs of sidenerves; petiolules
1—2 cm. Flowers in short pauci-(10-i\'12) florous racemes
at the end of short fohate pendulous sidebranches; calyx
campanulate ca. 0.6 cm. 1. truncate with 5 erect-patent
equal teeth inserted immediately below the margin. Corolla
in bud pale green later on creamy, hypocraterimorphous
with a long ca. 15—17 cm. cylindric tube ampliate to the
throat; limb equally 5-lobed, lobes rotundate obtuse re-
flexed; stamens 5 fertile, equal, thickened to the base,
glabrous, inserted on 2/3 of the length of the tube and
just reaching the mouth of the corolla; anthers sagittate,
divaricate with a short toothed connective. Ovary sub-
quadrangular, bilocular with ovules in two rows pro loculo;
disk annuliformous crenulate; stigma bilamellate. Capsules
shortly stipitate, broadly lanceolate to oblong much like
those of N. brunfelsieefloras, plane, compressed parallel to
the septum, varying in length, when ripe 16—24 x 4—4.8
cm.; valves plane coriaceous median costulate; septum
plane membranous-sub-coriaceous, in sicc. nigrescent on
each side with two rows of seeds inserted at the margin;
seeds sub-orbicular 3—4 cm. broad (including the wings
0.6—1 cm. br.) thin membranously winged.

Geogr. distr. Celebes — Moluccas: Zippel (L sub n.
898, 199... 103. Type specimen of Blume); Menado:
Teysmann and H. de Vriese (L sub n. 907, 135...
804, n. 898, 199... 104, n. 898, 199... 105); Cult, in

Hort. Bog. ex. Menado sub n. H. B. 5729 (B, P. U);
Celebes: Teysmann n. 14105 (B).

Philippines — Luzon — pro v. Sorsogonpr. Irosin : Elmer
n. 14426 (L. U. B); prov. Cavite et Batangas: Cur ran
n. 7668 (D); Calawan, Manila: M. Callery n. 64 (P);
Manila: M. Barthe (P) —Pa/au^an —Bermejos n.39771
(= F. B. 4718) (D) - Polillo (Merrill) - Biliran (Merrill)
-- Basilan (Merrill). In primary forests at low and medium
altitudes (Merrill).

Habitus gracilis; foliolis oblongis vel oblongo-lanceo-
latis, basi cuneato-rotundatis 9—11 x 4—4. 5 cm.; calyce
0.6—0.7 cm. 1.; pedunculus ca. 8 cm., tubus corollae
17-19 cm.

Nyctocalos cuspidatum {Bl) Miq. is connected with N.
Thomsoni from which it differs in the already mentioned
characters.

Merrill notes a form from Luzon (prov. Ceram : V i d a 1
n. 3395) which differs from the usual specimens by the
longer calyx, the more acute not cuspidate leaves and the
shorter petioles. This he thinks a mere form however. The
cuspidate apex is indeed not the main point characterizing
the species.

The geographical distribution is remarkable. Whereas
in the Philippines it is very common in nearly the whole
Archipelago, it is found in the Malayan region only in
Celebes near Menado, where probably also Zip pel col-
lected the type-specimen. Whether it occurs in the Talaud
Archipelago I do not know, but I did not receive specimens
of Lam\'s collection which botanist explored there some
time ago.

Ventetat. Dec. Gen. Nov. (1808) 8 ; — = Calosanthes
indica Bl. Bijdr. (1825) 760; Endlicher. Gen. Plant.
(1836-40) 4119; D. C. Prod. 9 (1845) 177; Bureau.
Monogr. (1864) 45. t. 9; Benth. amp; Hook. Gen. Pl. 2
(1876) 1040; Bâillon. Hist. Pl. 10 (1891)39; K.Schumann
in Engler-Prantl. Pfl. Fam. IV. 3b (1895) 225; Boerlage.
Handl. Fl. Ned.-Ind. 2 (1899) 589.

A glabrous tree. Leaves opposite, large. 2—3-pinnate;
leaflets ovate, entire. Raceme terminal, long. Calyx large,
leathery, campanulate, truncate or obscurely toothed. Co-
rolla large, campanulate-ventricose. white or purplish;
lobes 5. subequal. round, crisped, toothed, stamens 5;
anthers glabrous. 2-celled; cells parallel, oblong. Capstde
large, hnear, compressed parallel to the septum, septici-
dally 2-valved. Seeds thinly discoid; wing hyaline, broad.

= Bignonia indica Linn, var x. Spec. Plant. 2 (1753)
625; Ventetat Dec. Gen. Nov. (1808)8; Lamarck Encyd.
Meth. 1 (1783) 423; Willd. Spec. Plant. 3306; - = Pa-
lega-Pajanelli Rheede Hort. Malab. 1 (1686) 177, t. 43;
Linné. Fl. Zeyl. 236 ; — = Calosanthes indica Blume.
Bijdr. (1825) 760; Roxburgh Fl. Ind. Or. 3 (1850) 110;
^ = Bignonia quadripinnata Blanco Fl. Filip. (1837)499;
Moritzi Syst. Verzeichn. (1845-46) 46; D. G. Prod. 9
(1845) 177; Wight Icones. 4 (1850) t. 1337-8; Blanco
Fl. Filip. ed. 2 (1845) 349; — = Spathodea indica Pers.
Ench. 2.173; — = Bignonia pentandra Lour. Fl. Cochin.
Ch. 2. 460; Miquel Fl. Ned. Ind. 2 (1856) 752; Blanco
Fl. Filip. cd. 3 (1878) 283, t. 219; Brandis Ind. Timb.
(1906) 496; Cooke. Fl. Bombay. 3 (1905) 327; Royle 111.
295; Prain Beng. PI. 787; Villar Nov. App. (1880) 151;
Vidal Synopsis. Atlas (1883) 35. t. 73; Vidal Rev. Plant.

Vase. Filip. (1886) 202; C. B. Clarke in Hooker Fl. Br.
Ind. 4 (1885) 378; Kds. and Val. Bijdr. Booms. Java. 1
(1894) 66: K. Schumann in Engl. Prantl. Pfl. Fam. IV.
3b (1895) 225: Gammie. Rec. Bot. Surv. Ind. 1 (1895)
65. 83; Gage. ibid. 1 (1901)352; Forbes-Hemsley. Journ.
Linn. Soc. (Bot.) 26 (1902); Wood. ibid. 2 (1902) 14. 25;
Gage. ibid. 3 (1904) 86; Prain. ibid. 3(1905)255; Brandis
For. Fl. 347; Dalz amp; Gibbs. Bomb. Fl. 161; Beddome
For. Man. 148; Wallich. Cat. n. 6514; Kurz. For. Fl.
254; Kanjilal. For. Fl. 254; Gamble. Man. Ind. Timb.
(1902) 510; Foxworthy. Philip. Journ. Sc. 6 (1909) 558;
Duthie. Rec. Bot. Surv. Ind. 2(1911)170; Merrill. Philip.
Journ. Sc. 1. Suppl. I (1906) 124; Burkill. Rec. Bot. Surv.
Ind. 4 (1910) 123; Kds. Exc. Fl. Java. 3(1912) 183; Kds.
and Val. Atlas Baumart. Java 2 (1914) t. 358; Fischer
Rec. Bot. Surv. Ind. 9 (1921) 132; Merrill Enum. Philip.
Fl. PI. 2 (1923) 548; Burkill Rec. Bot. Surv. Ind. 10
(1925) 331 ; Heyne Nuttige Plant. Ned. Ind. 4 (1917) 165.

Smal tree 6—13 m. with large leaves and {lowers ca.
7 cm. with a leathery broad campanulate calyx and con-
spicious large siliquiform flat capsules ca. 60—120 x
7—10 cm. The main stem is generally somewhat crooked
and below devided in several other stems, these strongly
branched. The remark of Roxburgh (1. c. 110) quot;oneof
the tallest trees of the Coromandel coastquot; is, also after
C. B. Clarke (in Hooker Fl. Br. Ind. 4, 378) certainly
wrong. However Endert gives for the collecting number:
\'Borneo: Endert n. 2325, a height of the tree up to 25 m.
Perhaps this may be a mistake; unanimously the other
collectors mark for the average height 10—12 m.

The branches have a smooth ca. o. 6—1 cm. thick
bark of a light-gray or grayish colour with few lenticels.
The leaves are very large, full-grown ca. 60—90 cm. 1-,
3-pinnate. The petiole is glabrous, terete, sulcate and at

the base and near the insertions of the side-axes articulate.
Therefore they show a striking resemblance with several
Araliaceae. Indeed Miquel (Ann. Mus. Lugd. Bat. 1
(1863) 27) described a new species, namely ArthrophyHum
zeylanicum n. sp.? on a sterile specimen collected by
Potielle in Ceylon. The same mistake he made with
an other sterile specimen, which he named A. reticulatum
Bl These specimens I found in the Leyden Herbarium;
Hallier f. had already recognized them.

In sterile state Oroxylum is always to be recognized
by the reticulate structure at the underside of the leaves,
which is lacking in Arthrophyllum.

So I was able to distinguish a specimen in the Paris
herbarium from Australia (East coast: M. Baume (P)).
which seemed to be a mistake already from a geographical
point of view, as the area of Oroxylum is established for
a good deal now. Indeed the sterile specimen is not
Oroxylum, or any other Bignoniacea, but perhaps also an
Araliacea, though not an Arthrophyllum.

The folioles are petioled. smooth and shining above, and
mostly sparely sliort tomentose not shining or totally gla-
brous beneath.

Along the midrib near the insertions of the primary
nerves there are minute scattered glands. This was most
clearly expressed in the cited specimen of Ceylon; that
of Manila (Luzon: El m e r n.15636) possessed not as many
of such glands and the specimen of Java (Bantam: Kds.
n. 223 /?) showed only at the base of the folioles some
single glands.

The folioles are ovate to ovate-oblong, oblique cordate
or sub-cordate at the base, whilst the top is apiculate or
short acuminate. The folioles are on the average 14.5 X 9 cm.;
I saw several specimens with minute orbiculate folioles.
So for instance a Philippine specimen (Merrill n. 125)
measured 4.5-7 X 3. 5-5. cm. and so did another from the

Kangean Archipelago (e.g. Backer n. 26828), which was
6.5-8 X 4-4. 8 cm. and from Borneo (E n d e r t n. 2325) where
the folioles were broadly ovate 7-9 cm. 1.. C. A. B a c k e r
also collected such a specimen near the k. Lima pr. Batavia,
but remarks that it was a young tree;soBeumée(Beumée
n. 531) notes the same of a specimen from Pekalongan
(Java). Moreover the structure of the leaves of these
young specimens is not sub-coriaceous but papyraceous.
So 1 think this fact is owing to the age of the tree.
Whether the habitat has something to do with it, I do not
know, but it is not probable.

Miquel (Fl. Ned. Ind. 752) states a form ^ with sparely
acute-dentate folioles. Of this form I saw but few examples.
It is certainly an inherent quality of very young trees, this
also occurring with Dolichandrone spathacea (L ƒ.) K. Sch.

The flowers are placed in large terminal racemes ca.
30 cm. 1., have a dirty-dark-violet colour and a disa-
greeable smell. The calyx is very large, broad campanulate
browny or violet-dirty-green coloured, thick leathery, per-
sistent, truncate or obscurely 5-dentate. The corolla is
outside dark-violet and inside dirty-yellow-gray. The
buds arc large and thick on stout pedicels. The large
pendulous capsules are very conspicious on the little tree
60-120 X 7-10 cm.; the seeds are large too, 3-4 x 8-9 cm.,
ca. 3.5-4 cm. thin membranous winged.

Geogr. distr. S. E. Asia—Throughout India from the
Himalaya to Ceylon and Malacca, common (Clarke in
Hooker Fl. Br. Ind.)-Wight n. 2331 {L.P); Br. India:
Jacquemont (P); Br. India: Herb. E. Lefèvre (P) —
Piindjab, in the Teray-West to the Cherab (Brandis) —
Malabar (Rheede); Malabar. Concan: Hook. f. ô Thom-
son (P) — Madras Presidency, Anaimatai Hills in the Coim-
batore distr. (Fischer) ; Cult, in Hort Pondichery : P e r r o 11 e t
n. 59 (P) — Chutia Nagpiir (Wood) — Ceylon, type
specimen of Arthrophyllum zeylanicum Miq.\'. Potielle

n. 116 (L): Ceylon Hook. f. ö Thomson (P) - Upper
Gangetic Plan (Duthie) — Assam, Sadija (Burkiii) —
Namlur Forest; G. Mann (L.) — Lakhimpur distr. (Gammie)
~ Nepal pr. Bichiakok. Khatmanda (Burkiii) — Calcutta
(Prain) - Bengal: Clarke (L. U); Bengal: Hook. f. Thomson
(P) ~ Siam (Ridley) — Burma (Gage) — Cochin China,
Hainan, Hongkong (Forbes-Hemsley) — Ava (Hooker) —
Andamans (Foxworthy) — Malay Peninsula, common
(Ridley).

V.n,: (Sanskr.) shyonaka (India); (Sanskr.) mundooka-
poerna (Roxb.); (Hind.) shyona (Roxb.); (Hind.) vanga
marum (Roxb.); (Hind.) pampena (Roxb.); gingen (Chutia
Nagpur); dak-dowa (Chutia Nagpur); szcaf (Chutia Nagpur
Wood); ullu (Upper Gangetic Plan: Duthie); pharkhat
(idem); pharri (idem); sanna (idem); tofila (Ceylon : Trimen);
bulai (Mal. Penins.: Ridley).

Malayan Archipelago. — Sumatra (Miquel). — Deli
(De Bussy). — Simaloer: Ac hm ad n. 121 (L. B). n. 1254 (B);
Atjeh P. Bras pr. Ooele-paya: Kds. 10653/3 (B). —
Siboelangit: Lörzing n. 5183(B).- W. Karoland pr.
Laoe Balang, Perbesi: Lörzing n. 11269 (B). — Palem-
bang. Afd. Maoera Doea: Greshoff n. 434 (B); Ins.
Sebesi (N. Sum.): Docters v. Leeuwen-Reynvaan
n. 5363 (B).

V.n.: mengleo (Simaloer: Achmad); /»atrenp (Atjeh : Kds.);
(Mad.) kapoeng-kapoeng (Palembang: Heyne, GreshofF).

Java — Java ? Moluccas ? (sub Arthrophyllum reticulatum
B1.(L); Java: Lesch enault (P.)—Res, Bantam.—Tjimara
Oedjong: Kds. 223/3 (L); Goenoeng Kantjana (L);Pasa-
roean: Backer n. 7274 (B); between Tjilèbès and G.
Kantjana: Backer n. 1172 (B); between Djepitoe and
Kemadang, G. Kidoel: Backer n. 2732 (B); Klappa
Noeggal: Backer n. 5869 (B); Tjiratjap, Djampang
Koelon: Backer n. 17336 (B); pr. Bantam: Zollinger
n. 1012 (Zollinger)—/^es. ßafay/a—Buitenzorg: Boerlage

(L): Laladon, Tji Apoes: Bakh. v. d. Brink. f. n. 2299
(B); Klappa Noenggal: Backer n. 23407 (B); Batavia:
Backer (B); Batavia: Vorderman (B); Batavia pr. k.
Lima: Backer (B)—/^e5.Preanger—C. Ploem (L); Tomo:
Kds. 221 ß (L); Palaboeanratoe: Kds. 220/S (L, P), Kds.
11710 (L, B), Kds. 33130 (L, B): Sea of Pendjaloe:
Kds. 47919 ß (B). Kds. 44308(B) — i?es. ßanyoemas-
Noesakambangan: Kds. 20042 ß (L, B), Kds. 24606/?(L),
Kds. 26857/3 (L); Bandjernegara Primgombo: Kds. 3864/?
(L. B), Kds. 229 ß (L, B), Kds. 34042 ß (L, B); Noesa-
kambangan: Amdjah n. 23 (B); Madjenang: Backer n.
18606 (B) — Res. Chenbon — Houten n. 191 (B) —
Res. Pekalongan — pr. Margasari: Burger n. 3359 (B);
pr. Weleri: Backer n. 16519 (B); Margasari: Beumée
n. 531 (B); Soebah: Kds. /3 (L, P, B), Kds. 228 ß{L,P, B),
Kds. 37058 ß (L), Kds. 14211 (B) - te. 5emarang-
pr. Boeloe: Oltmans(B); Kedoengdjati: Kds. 224 ß (L),
Kds. 26403 ß (L,B), Kds. 27221 ß (L), Kds. 225 /?(L);
Karangasem: K d s. 226 /? (L, B); Ngarengan: K d s. 35007 ß
(L), Kds. 35008 ß (L, B) - Res. Rembang - Sedan:
Kds. 36445 ß (L) - Res. Madioen - Pandan: Elbert
n. 473 (L); Madioen: Wisse n. 132. ^ Res. Besoeki —
Poeger Watangan: Kds. 230 ß (L), Kds. 12835)3 (L, B),
Kds. 30154/3 (L), Kds. 231 ß (L): Rogodjarapi: Kds. 680
ß (L). — East-Java: Forbes n. 1284 (L).

V.n.: (Soend.) pongporrang (Bantam, Preanger: Kds. and
Val.); (Soend.) pongporang (Preanger, Cheribon: Kds.,
Heyne); praon: ki tongtorang (Batavia); poempoeran (Kds.
and Val., Blume); ki lakaki (K d s.); (Mal.) boengli (Batavia);
(Jav.) woengli (Bantam, Pekalongan, Banjoemas); moengli
(Banjoemas, Pekalongan); detég (Besoeki: Kds.); lanang
(Besoeki: Kds.); kajoe lanang (Semarang, Tegal, Soerakarta,
Madioen, Kediri, Banjoemas, Pekalongan, Bagalen, Kedoe);
kadjen djaler (Besoeki: Kds. and Val.); padangan (Sema-
rang: Kds.); kêok (Semarang).

Madoera — Sampang: Backer n. 19535 (B) — Kan-
gean Archipelago — Ardjasa: Backer n. 26828 (L); Eteng
pr. Tambajangan: Backer n. 27451 (B); N. of Ardjasa:
Backèr n. 27211 (B); Doeko: Dommers n. 47? (B);\'
P. Paliat: Backer n. 29545(B); Saboenten: Backer n.
29724 (B); Kangean: Dommers n. 285(B); Mamboerit:
Backer n. 27329(B); Sepapan: Backer n. 28497 (B)-
Timor — Miquel in Fl. Ned. Ind. 2 (1856) 752; Timor:
Coll? (P.).

Borneo — Sanggouw: Hallier. f. n. B 920 (L, B); W.
Koetai: Endert n. 2325 (B) — Celebes — Madjene:
Noerkas n. 419 (L); Minahassa: Kds. 16096 ß (L); G.
Galesong pr. Malino, S.W. Gel: Bünnemeyer n. 10931
(B); Afd. Bonthain, 200 m. alt. n. bb. Cel. I 73.

V.n.; kajo karoe kadang (M. E. Borneo); (Mai.) kajoe
pedang (Menado: Heyne); boeli (Celebes).

Philippines — Northern Luzon (Cagayan) to Palawan
and Mindanao in most or all islands and provinces (Merrill
1923) — Island of Negros pr. Dumaguele: Elmer n.
6569 (L). — Boroboro: Com. Flor. For. Filip. n. 1626
(L). — Luzon pr. Irosin: Elmer n. 15636 (L); Luzon pr,
Manila (L); Luzon pr. Antipolo: Merrill. Spec. Blan-
coanae n. 125 (L, B); Isabella Prov., Luzon: R. J.
Alvarez n. F. B. 18576 (P).

V.n.: abang-abang (P. Bis); abong-abong {P. Bis); balay
uak (Sul.) barangan (Ilk.); fcun^/ui (Sul.); balilang-uak(Vaq.y.
balinag (C. Bis.); banlai (Sul.); bungai (C. Bis.); kamkam-
pilan (Ilk.); kampilan (Neg.); maidboid (Bik); pingka pingka
(Tag.) pingka-pinkahan (Tag.); taghilan (Tag.).

Oroxylum indicum (L) Vent. fl. citrinus J, J. S. mss.
Cult, in Hort. Bog. XV K (B. IX) II. (B).

at low and medium altitudes from 0 to 900 m. but con-
stantly occurring through the whole area. In the Malayan
Peninsula the species (Ridley I.e.) is quot;very abundant
and conspicious near rivers and in lowland swampy
countriesquot;. The species is certainly a striking one in the
forests and thickets, but in high-stemmed primeval forests
it lacks. However in the mountains south of the Himalaya
Oroxylum occurs at greater altitudes, as for instance near
Kumaon at 1300 m. above the level of the sea (D u t h i e 1. c.).

The climate on the other hand influences the occurrence,
e. g. it does not occur in the dry regions of N. W.
British India.

As for the Malayan Archipelago Backer gives many
remarks on frequency and habitat on his collecting labels;
so for instance in the Res. Bantam (Java) it occurs seldom
in secondary forests (shrubwildernesses), but never socially.
In the Res. Madioen (Java) Wisse notes the same and
in the Res. Banjoemas Oroxylum is rarely found according
to Koorders.

As to the Philippines Merrill (1923. I.e.) remarks
quot;occurring in thickets and secondary forests at low and
medium altitudesquot;. Thence it is remarkable that in the
Kangean Archipelago, perhaps part the east frontier of
the area. Backer gives some remarks on the frequency
there. Oroxylum seems to be very rare in this region
which is excellently worked out by this botanist. So for
G. Eteng pr. Tambajangen he only found 1 specimen and
likewise in Saboenten and Mamboerit.

In general it seems that the total number of individuals
is consistent distributed in the whole area except in the
eastzone.

This may be clear if we assume that Oroxylum, original-
ly endemic in S. E. Asia, extended the area towards the
east and that this process still goes on. Then indeed it is

probable that on the Kangean Archipelago we have got
to do with pioneer-specimens. From Bali, Lombok, Soem-
bawa, Soemba, the Solo Arch., the Alor Islands, etc. I
miss all materials or other notes. Only in Timor it occurs.

The matter of spreading seems to go easily; Bur kill
(1925) states the appearance quot;in a particular type of new
jungle between Sadija and Saikhoa.quot; Prain (1905) says
about this near Calcutta: quot;occasionally planted and some-
times selfsown.quot; The wind appears to me the main factor
in the question of the matter of spreading. Whether this
goes on large distances I do not know, but thereby the
appearance in the Malayan Archipelago in so many islands
would be explained.

For the use of the wood, etc. Hey ne (1. c. 165-166)
gives some remarks on the use of Oroxylum in the Malayan
Archipelago. Scheffer notes that the roots are used
together with other remedies against impotentia virilis. In
Java the natives do not use the timber, this being too
small and not durable (Kds. and Val.). Koorders
(Tectona III 121)-notes that the timber is used in the
matches-industry. He (Minahassa) also remarks that in
North Celebes the weak part between the bark and the
wood is used for stanching blood.

Not only in the paleaotropics it is cultivated here and
there, but also in the neotropics, though perhaps very
\'^are; e. g. in Martinique (Introduced and cultivated in
the Jardin Colonial: Hahn n. 1297 (P)).

Suppl. (1781) 45, 291; A. L. de Jussieu Gen. PI. (1789):
C. Prod. 9 (1845) 182; Bureau Monogr. (1864) 45;
Benth. 6 Hook. Gen. PI. 2 (1876) 1040; K. Schumann in
Engler-Prantl Pfl. Fam. IV. 3b (1895) 226; Bâillon Hist.
P\'- 10 (1891) 39.

Monotypic genus; native country probably Burma, from
Ava to Tenasserim. Perhaps also wild in Centr. India and
on the Upper Godavery, according to Kurz and Clarke.
Planted extensively throughout India and the Malayan
Archipelago; in India sometimes sub-spontaneous.

Suppl. (1781) 291; — = Bignonia suberosa Roxb. Cor.
PI. 3. 11, t. 214; — = Millingtonia dubisa Span. Linnaea
14(1841) 326; Roxburgh. Fl. Ind. 3 (1832) 111; Wallich.
Cat. n. 6513; Beddome. Fl. Sylv. t. 249; Kurz. For. Fl.
2. 238; Brand. For. Fl. 347; Miquel. Fl. Ned. Ind. 2
(1856) 753; Bureau. Monogr.( 1864) 45, t. 8; Miquel. Ann.
Mus. Lugd. Bat. 1 (1864) 753; C. B. Clarke in Hook.
Fl. Br. Ind. 4 (1885) 377; Kds. and Val. Bijdr. Booms.
Java. 1 (1894) 65; K. Schumann in Engler-Prantl. Pfl.
Fam. IV. 3b (1895) 226, fig. 891; Wood. Rec. Bot. Surv.
Ind. 2 (1902) 125; Gage. Op. cit. 3 (1904) 86; Prain.
Op. cit. 3 (1905) 255; Saxton and Sedgwick. Op. cit. 6
(1918) 286; Kds. Exc. Fl. Java. 3 (1912) 183; Merrill.
Enum. Philip. Fl. PI. 3 (1923) 444; Foxworthy. Philip.
Journ. Sc. 4 (1909) 558; Watt. Diet. 5. 247.; F. Villar,
Novis. App. (1880) 150.

An erect tree, up to 25 m. high, bark corky. Leaves
opposite, pubescent when young, mature nearly glabrous;
to 1 m. 1., 2—3-pinnate; leaflets ovate-lanceolate acuminate,
sinuate or crenate 3. 5—5 cm. 1. Flowers white in many-
flowered, 3—7 cm. 1. peduncled terminal panicles ca. 25—40
cm. 1. on pendent branches; panicles minutely yellow
pubescent. Calyx very short ca. 0.2—0.3 cm. 1. campanu-
late, truncate with 5 obscure lobes. Corolla with slender
0.1—0.2 cm. br. tube ca. 7 cm. 1., campanulate-dilatate
towards the throat with 4 stamens and a rudiment; each
fertile stamen having one fertile cell which is calcarate at
the base, and a rudimental one also with a spur; stamens

shortly exsert, didynamous; corolla distincdy bilabiate
5-lobed, lobes ovate-acute short pubescent at the margin, in
bud sub-valvate. Capsule linear compressed parellel to the
septum, septicidally 2-valved ca, 30 X 1.7 cm.; seeds thinly
discoid with broad hyaline wings (includ. wings 1.4 X
3.5 cm.)

Geogr. distr. India — Cultivated throughout India, per-
haps wild in Burma (Clarke); India orient: Wight n. 2332
(L, P); idem Mt. Nilghiri: Thomson (L sub n. 898,
199... 197) - Siam - Teysmann n. 5959 (U); Cult,
in Hort. Calc.: Wallich n. 6513 c (P); Nellighery: M.
Perrottet n. 281 bis (P); pr. Pondichery: Perrottet
n- 285 (P); Bengal: L e s c h e n a u 11 n.? (P); Penins. Ind. Or.,
Mt. Nilghiri öKurg.(P); Pondichery: O. Debaux n. 503
(P); Coc/iin C/i/na. arbre ornamental: Réguier n. 241 (P).

Malayan Archipelago — Sumatra — (Miquel Ann. 1
(1864) 753) - Java - de Vriese and Teysmann (L
sub n. 898, 199... 92, n. 907, 135... 802) - Res. Sema-
J^np - J. Haak.,n. 568 (B,L) - Res. Djokjakarta -
Here and there along the road from Djepiloe towards
Kala: Backer n. 2807 (B) - Res. Rembang ^ pt. Toe-
ban: Ch. Coster (B) — Res. Soerabaya - (L. sub n.
898, 199... 100); Solo cult.: Hemken n. 16(B) -/?es.
Kediri — Forestry Tritik: Grut ter ink n. 3260 (B) —
Res. Besoeki - Rogodjampi cult.; Kds. 219 /? (B); Ban-
joewangi: Teysmann n. HB 1774 (B); Banjoewangi. pr.
Sobo: Zollinger n. 2910 (P); Java: Leschenault?(P).

n.: (Jav.) sekar poetic (Pekalongan: Kds.); amfioenan
(Kediri: Grutterink); kcmbang poetic (Java: Kds.); (Jav.)
tekar petah (Banjoewangi: Kds. and Val).

Madoera — pr. Sampang: Backer n. 19809 (B) —
kangean Archipelago — Bilibilis: Dommers n. 189 (B)
~ Bali pr. Beliling: Teysmann sub n. HB. 2764 (B);
idem(U sub n. 032833); Bali (L. sub. n. 898, 199... 99,

n. 898, 199... 94); Bali: Cunningham n. 321 (L);ex.
Herb. Paris (L sub n. 898, 199. . . 93) — Timor (U sub
n. 032830); pr. k. Soenimela: H. Therik n. bb. 9560;
N. M. Timor pr. k. Haoelassi n. bb. 7728 (L): Timor:
Riedlé(P) — Soembawa — pr. Taliwang: Zollinger
n. 2910 (P).

V. n.: ketanggar (Soemba); karpoti (Kangean: Dommers);
kanongoh (Bali: Teysmann); toka haoe (Timor).

Celebes — pr. k. Rappang Enrekang: Noerkas n. 332
(L.B); pr. k. Saloeang Wolangitang: Bontham n. bb. Cel.
I. 87.; Afd. Boeton, P. Moena: n. bb. 4998 (B) — Ceram
- n. HB 1993 (L); (U sub n. 032832).

In the Malayan Archipelago often cultivated for orna-
mental purposes, except perhaps in Sumatra, Borneo and
the Philippines (Merrill); ripe capsules are seldom found.

Mr. J. Haak sent in 1903 the cited specimen from
Semarang to Dr. Koorders with the remark that the
species was used as a substitute for opium. The dried
leaves were used in cigarettes; the same is done in Kediri.
Heyne however does not mention this.

Miquel (Ann. 1. c.) states some differences between the
Siamese. Sumatranan and Timorese specimens: „foliola in
siamensibus ovata vel deltoideo-ovata; in sumatranis ovato-
oblonga, jugis singulis trifoliatis in timoranis oblongolan-
ceolatis bi-tri-juga.quot;

With a cultivated species as Millingtonia hortensis it
is no wonder that such little aberrations appear, but these
do not represent essential characters on which even a foun-
dation for new varieties cannot be undertaken. For the
rest we do not know anything about the heredity of these
forms, nor whether they are the consequence of a natural
variability. It seems that generally Millingtonia hortensis
does not fructify frequently in the Malayan Archipelago.

Of this Bignonian genus only one species is cultivated
in the Malayan Archipelago; native of Brasil.

Pyrostegia ignca (Veil.) Presl. Bot. Bemerk. (1844)
93; Abh. Böhm. Ges. Wiss. 3 (1845) 523; Symb. Bot.
2. 28. t. 77; = Bignonia ignea Veil. Flor. Flum. 4.244.
t. 15; Miers in Proc. Roy. Hort. Soc. London 3. 188;
Bureau. Monogr. (1864) 42; Benth. amp; Hook. Gen. PI. 2
(1876) 1034; H. Bâillon. Hist. PI. 10 (1891) 31; K.Schu-
mann in Engler-Prantl. Pfl. Fam. IV. 3b (1895) 31.

Lian with costate twigs and opposite 1-jugate leaves
with a terminal 3-branched tendril, seldom with a terminal
leaflet; leaflets ovate to ovate-lanceolate, sub-obtuse to
acuminate towards the top. 5—7 cm. 1. Flowers red-yellow,
in dense terminal inflorescences ; calyx minute, broad cam-
panulate. truncate with 5 short lobes ca. 0.3 cm. I., corolla
in bud valvate, glabrous except the margin of the lobes,
tube narrow, cylindric, ca. 5 cm. 1., dilatate towards the
ca. 0.8 cm. br. throat; lobes lanceolate ca. 1.5 cm. 1.;
anthers exsert, ifiserted at the middle of the tube, didy-
namous. Capsule narrow, cylindric with coriaceous valves
and elleptic, membranous winged seeds.

Geogr. distr. Malayan Archipelago — Res. Preanger
Reg. — Tjisoeroepan pr. Garoet in garden of hotel:
Wig man Jr. (B); idem: Kds. 41004 ß (B); Bandoeng
cult.: Eyken (B); Soekaboemi cult.: Backer (B); Malabar
pr. Pengalengan: Backer n. 26210 (B) cult, in garden of
van Müllem (B); Lembang: v. Welsum n. 21. 30? (B);
Bandoeng in garden hotel Kota: Kds. 41140 ß (B).

Though this species is now and then cultivated in
Java it seems never to run wild.

C. Prod. 9 (1845) 183; Endlicher. Gen. Pl. (1836—40) n.
Sill\' Bureau. Monogr. (1864) 40; Benth. Ô Hook. Gen.
PI 2 (1876) 1032; H. Bâillon. Hist. PI. 10 (1891) 27;
K. Schumann in Engler-Prand. Pfl. Fam. IV. 3b (1895) 213.

Lians or shrubs, leaves 2—3-foliolate. mostly with ten-
drils. opposite, mostly with pseudo-stipules. Inflorescences
axillar or terminal, forming mostly composed paniculate
thyrses, seldom pauciflorous. Calyx truncate, membranous
or sub-coriaceous, sometimes toothed, glabrous, smooth,
seldom with small glands. Corolla funnel-shaped or cam-
panulate, mostly actinomorphic. Anthers 4, included, didy-
namous with a fifth rudiment. Disk variable. Ovary mostly
elongate, mostly scabrous, with 2 or more series of ovules
per cell. Capsule linear, mostly long, valves coriaceous.

(Plants of Santa-Marta, U. S. of Columbia, collected
by H. H. Smith. (1898-1901 n. 741).

Lian, branchlets terete, 2-foliolate, mostly with a large
(15 cm.) long tendril; pseudo-stipules 2—3 cm. diam., ca.
orbicular; petiole ca. 1. 5 cm. 1.; leaflets obovate oblong,
shining above, opaque beneath, 8—12 X 4—6 cm., largest
breadth at Vs ot the top. shortly obtuse-acuminate, cuneate
to the base; nerves 3—4 pairs, lower pair reaching Vs
of the length, prominent, reticulations lax, prominent.
Thyrses axillar and terminal, rather pauciflorous. Calyx
firmly membranous, cupular-campanulate, smooth, glabrous,
not toothed, 0.7 x 0.7 cm. Corolla rather large, purple-
violet, tubular-infundibuhformous, sub-equally 5-lobed,
lobes large; tube below narrow, 0.4 cm. diam., whitish
purplish-striate, enlarged above ca. 1.5 cm. diam. 5famcns
4, didynamous, fifth rudimentary. Ovary sub-scabrous,
with 2 rows of ovules in each cell.

Geogr, distr. U, S. of Columbia — Santa Marta, 500
m. alt.: H. H. Smith n. 741, type specimen (U).

Malayan Archipelago — Banka ^ pr. Muntok: Kor-
nassi (B) — Sumatra — Sibolangit, Arnhemia, ground
of a controller\'s house, 3 m. high, 100 m. alt.: Lorzing
n- 5501 (B, L) — Gouvt. Sumatra\'s Oostkust — Medan,
Deli, cultivated up to 850 M. alt. (Lorzing).

This plant is sometimes cultivated in the Malayan Ar-
chipelago. As the capsules are always missing, I am not
sure it is a true Arrabidaea, nor could I find the original
description of S p r a g u e The specimens I saw are just
the same as the type specimen I cited above.

Lorzing mentions it as a species, which is rather
often cultivated on Sumatra; it never produces capsules,
but is only propagated with suckers. This points to self-
sterility. Perhaps all specimens have originated from one
individual. The flowers stink a little.

Shrubs or sub-voluble plants with simple 1-pinnate
leaves, leaflets serrate. Inflorescences terminal dense. Flowers
orange or scarlet. Calyx regular, campanulate, 5-toothed.
Coro//a-tube narrowly funnel-shaped or almost cylindric,
curved; limb markedly bilabiate. Stamens 4, exserted;
anther-lobes connate for the upper third, divergent below.
Disk cupular. Ovules 4-seriate in each cell. Capsule oblong-
linear, much compressed parallel to the septum.

Dr. T. A. Spraguc was (Sept. 27, 1927) so kind as to inform
me about magnifica Spraguc; it is a new combination for Bignonia
magnifica Bull ex Card. Chron. July 19 (1879) 72. which Spraguedid
not publish yet.

Malayan Archipelago. It is a native of South Africa
(Schumann): but naturalised in the neo-tropics.

Tccomaria capensis (Thbg.) Fenzl. — Spach. Hist.
Veg. Phan. 9 (1840) 137; — = Tecoma capensis Lindl.
Bot. Reg. t. 1117; Bureau. Monogr. (1864) 47; Benth.
amp; Hook. Gen. PI. 2 (1876) 1044; K. Schumann in Engler-
Prantl. Pfl. Fam. IV 3b (1895) 229.

Climber without tendrils: leaves opposite imparipinnate,
3—4-jugate, 8—10 cm. 1.; leaflets ovate short acuminate
with few gross teeth especially near the top, 1—3 x 0.8-0.
2 cm.; terminal leaflet mostly the largest one. Racemes
or panicles terminal ca. 10 cm. 1.; ca/yx glabrous 0.4 cm. 1.
truncate shortly 5-dentated; corolla infundibuliformous, a
little curved, the lower part inside pubescent, imbricate,
ca. 4. 5 cm. 1. 5famens exsert, cells divaricate, didynamous
with a fifth rudiment inserted at the middle of the tube.
Capsule linear, flat, seeds narrowly winged.

Geogr. distr. — Malayan Archipelago — Java — Res.
Semarang — Salatiga: Backer n. 30322 (B) — Res.
Preanger — cult, in a garden pr. Soekaboemi: Backer
n. 22160, n. 23223 (B).

The native country of Tecomaria .capensis was discussed
by Seemann (1863), who gives several arguments for
the neotropical origin; indeed, the plant seems to be
naturalised in Brasil and was gathered by Miers and
others in the natural vegetation. Moreover, Seemann\')
quite rightly says that, if T. capensis was a native of
Africa, as well as of America, it would be the only Big-
nonian species occurring on two continents. The principal
argument of Seemann was that the other Tecomaria s
were found also in South America, viz. T. fulva Seem.
and T. rosaefolia Seem. This has turned out to be wrong.

K. Schumann (1895) mentions a second new African
species : T. Nyassae [Oliv) H. Bn. and therefore he thinks
it not probable that T. capensis would be a native of
the neo-tropics. Moreover Sprague \') has found that
the American species of Tecomaria belong to Stenolobium,
on account of the free anther-lobes and 2 rows of ovules
in each cell of the ovary of Stenolobium. Thus Teco-
tnaria has been established to be a true South African genus.

Gen. PI. (1836-40) 711, n. 41 Ha (Sect. Tecomeae);
Spach. Hist. Nat. Vég. Phanérog. 9 (1840) 136 ; Ventetat.
Jard. Malm. {Bignonia) t. 3; Maund. t. 8; Lindl. Bot.
Reg. t. 2002; Bot. Magaz. t. 865 {Bignonia) 4094; D. C.
Prod. 9 (1845) 225 (Sect. Tecomeae); Bureau Soc. Bot.
Fr- 9 (1862) 163; Seemann. Ann. Mag. Nat. Hist. ser.
ni. 10 (1862) 31; Journ.Bot. 1 (1863) 19; Bureau. Monogr.
(1864) 49 ; Bentham. Fl. Austr. 4 (1869) 537 ; Benth. Ô Hook.
Gen. PI. 2 (1876) 1045; Bâillon. Hist. PI. 10 (1891) 40;
K. Schumann in Engler-Prantl. Pfl. Fam. IV 36 (1895) 230 ;
Boerlage. Handl. Fl. Ned. Ind. 2 (1899) 590; Bailey.
Queensl. Fl. 4 (1901) 1133; Diels. Engl. Jahrb. 57 (1922)
^198; van Steenis. Nova Guinea 14 (1927) 294, 301.
Fig. 3 a, b, e; 4; 5 r; 16.
Descr. emend. Frntices parvi vel altiores, volubiles,
lignosi, glabri, habituTecomanthis. Folia opposita, 1-pinnata,
2-11-jugata, rariter in apice ramulorum 3-foliolata, foliolis
parvis vel magnis. Infîorescentiae terminales, parte infima
interdum foliatae. rariter cauliforae, cymis in racemis dispo-
sitis, thyrsum efFormantibus, pedunculi infima basi semper
squamis minimis distichis vel verticillatis connatis praediti.
Flores parvi 1—3. 5 cm. in P. leptophylla {Bl.) Boerl. et
Curtisii Ridl. 5—7. 5 cm. Calyx semper parvus, trun-

catus vel 5-dentatus, campanulatus vel cupulatus. Corolla
parte inferiore anguste cylindracea, intus glabra vel stuposa,
parte superiore infundibuliformis sub-campanulata vel rariter
tubulosa, intus parte altera anteriore stuposa, altera glabra,
inaequaliter 5-lobata, sub-2-labiata. Stamina 4, didynamia,
quinto rudimentario, filamentis glabris vel basi stuposis.
Ovarium oblongum, 2-loculare, placentis binis pro loculo,
ovulis multi-seriatis praeditis. Capsula oblonga, plerumque
apice acuminata, laevis; valvis aequalibus, firmiter coriaceis,
concavis, septo piano, semina oo applanata, tenuitermem-
branaceo-alata.

(Tribus Tecomeae. Type P. australis
{R. Br.) Spach. Species 10. Malay Peninsula
to New Caledonia).

When giving a revision of this genus it seemed to be
needful to give a short account of the history, together
^•th that of the allied genera, which form in my opinion,
though certainly of distinct generic rank, a large group or
^\'ppequot;, having the same origin in geographical and
Phylogenetical respect. Here it is not the place and more-\'
°ver I do not feel competent to give a critical revision
of these genera, which I have not especially studied for
^he greater part, because they do not inhabit the countries
spoken about in this work. But to understand the origin
and distribution of several Malayan Tecomeae, it is abso-
lutely necessary to notice their affinities, which will appear
to be of fundamental moment later on.

A- de Jussieu (Gen. 1789. 139) the creater of the
Qenus Tecoma took elements together in this genus, which
on appeared to give raise to several distinct genera.

EndliCher (Gen. 1836—40. 4114) distinguished 3
sections amongst which the true Pandorea in section a
^ampsis Lour, he reckons synonymous with Incarvillea
Juss.. Stenolobium D. Don he arranges as a synonym

with Bignonia Juss. and Tecomaria forms section c.
of Tecoma. The section b. was composed of American
Tecoma s in Eu-Tecoma.

De Candolle (Prod. 9. 1845) accepts 2 sections, the
first of which included sect, b c ofEndlicher and a.o.
Tecomella Don. and Campsis Lour.. Stenolobium he
placed also in Bignonia. Section II contained Pandorea.

Thus it was rational that Blume (1848) supposed the
Malayan Tecoma s : P. leptophylla, T. dendrophila and
T. amboinensis to belong to Tecoma (Campsis) and so
did Miquel (1856). However, the latter (Ann. Mus. Lugd.
Bat. I (1864) 197) supposes their affinities with Pandorea,
but cannot decide, as the fruit is unknown to him.

Bureau (Monogr. 1861) distinguishes on the contrary
in his group II (1. c. 47) distinct genera, e. g. Tecomaria
Fenzl, Tecoma Juss. (part.), Campsis Lour., Campsidium
Seem, et Reiss., Tecomella D. Don., which he points next
to Stenolobium D. Don the latter being unknown to him,
Amphicoma Lindl and Pandorea (Endl) Seem.

Bentham and Hooker (Gen. PI. 1876) are of the
opposite opinion and again took together several of the
above mentioned genera to a large genus Tecoma owing
to their interpretation in general about the limits of a genus.
Within Tecoma they distinguish 4 sections, some according
to mentioned distinct genera, but others still composed of
heterogeneous elements. It is curious that they took
T. dendrophila and 2 African Pandorea\'s together in
section 3. Eu-Tecoma {Campsis Lour.), but left the
Australian and New Caledonian Pandorea\'s separated in
section 4. Pandorea Endl.

H. Bâillon (Hist. PI. 10.1891) agreeing with B u re a u,
again restores distinct genera and so does K. Schumann
(Pfl. Fam. 1895).

Boer lag e (Handl. Fl. Ned Ind. 1899) gives a review
of the history of Pandorea and supposes the 4 Malayan

Tecomas: Tecoma leptophylla BL, T. ceramensis T. etB.,
T. dendrophila Bl. and. T. amboinensis Bl. preliminary
to belong to Pandorea, however he points to their afflinities
with Tecomanthe H. Bn., an imperfect know genus then.

There are in my opinion the following relations between
the genera allied with Pandorea.

These genera contain all lians, chmbing without tendrils,
with opposite 1-pinnate (seldom digitate) leaves, showing a
Tecoma-flower, that is to say a calyx truncate 5-toothed,
campanulate 5-lobed or rather tubular; a corolla narrow
tubular below, enlarged above with 4 didynamous stamens
3nd a fifth rudiment, included or exsert, an oblong cap-
sule (seldom linear) with 2 concave coriaceous valve.s with
several rows of seeds on each placenta.

Pandorea {Endl.) Spach. Spec. 10. Malacca to New
Caledonia. Flowers mosdy small, calyx truncate, corolla
^^^ one side hairy inside, inequally 5-lobed, always thyrsal
inflorescences, anthers-included.

Flowers mostly large, corolla not on one side hairy
Within, inequally 5-lobed. Connected with transitional forms
^\'th Pandorea, having both racemes and thyrses, anthers

Calyx short, truncate, corolla small, regularly 5-lobed,
Inside on one side hairy; leaves digitate; anthers included,
\'osely allied with Pandorea, not so immediately with

Campsis Lour. Spec. ca. 5, perhaps more. Extending from
\'quot;do-China (Burma?) to China, Japan and in North America
quot;om Illinois to Florida.

mostly allied with Tecornanthe, though having an other
habit of the foliage. Fruit similar to that of Pandorea and
Tecornanthe, disk cushion-shaped and not annular as in
Tecornanthe.

As I have noted (p. 878) in Tecornanthe there are also
found roots on the stem but these are of no use to them
for climbing.

I do not know the Burmese Tecoma bipinnata Coll. et
Hemsl. (Journ. Linn. Soc. 28 (1890) 912), nor Tecoma
Cavalereia Leveille (Fl. Kouy-Tcheou. 1914/5.50-China:
Kweichan), nor Tecoma Mairei Leveille (Cat. PI. Yunnan
(1915) 20 —China: Yunnan). Perhaps these all belong to
Campsis Lour.nbsp;.

C. grandiflora [Thbg.) K. Sch. (= C. adrepens Lour.
Fl. Coch. Chin. 2 (1790) 377; — = Tecoma sinensis Spach.
Hist. Veg. Phan. 9. 135 = Tecoma grandi flora Loisel. Herb.
Amat. V. t. 286). It occurs in Japan, Korea, China and
Indo-China.T. Mori(Enum. PI. Corea (1922) 319) calls it a
native of Japan. H e m s 1 e y (Journ. Linn. Soc. 26 (1889-1902)
235) supposes it to be indigeneous throughout China, Indo-
China and Japan, where it is also cultivated.

C. radicans (L) Seem., occurs in the U. S. of America
from Southern New Yersey and Penssylvania to Florida
and Texas and north to Illinois and Iowa, escaped from
cultivation further north, according to B r i 11 o n and Brown
(111. Fl. North. Unit. States 3. (1913) 237).

Whether Tecoma fllicifolia Mc/io/s. (Diet. Card. 4 (1887)
13) of the Fiji Islands is a true Tecoma or a Campsis I
do not know. I think it is the first.

glands, cûrolla with rather short lobes, anthers included,
filaments attached in the middle, capsule linear, with sturdy
coriaceous valves. Certainly allied with Tecomanthe and
Pandorea, but possessing a linear capsule and somewhat
different anthers.

Allied with group I having a similar capsule, especially
showing affinities with Pandorea, but rather isolated. The
cupular disk and the foliage show affinities with Neosepicaea.
It differs however, having exsert stamens and a remarkable
valvate corolla, very curious characters indeed.

Anthers mostly exsert. Ovules 4-seriate per cell. Flowers
in terminal racemes or thyrses. Capsule linear, compressed.
According to Sprague (in Dyer. Fl. Cap. 4. 11(1904)448)
the American species.belong to Stenolobium D. Don., having
only 2 series of ovules in each cell.

The group is rather isolated, but has affinities with
Gampsidium and Campsis, though differing in the 4-seriate
ovules and the linear capsule.

Podranea Sprague spec. 2. Tropical Africa and the Cape.
ricasoliana Sprague (= Tecoma Maikenii Hort. ex
Watson = Pandorea ricasoliana Baill.) in Dyer Fl.
Cap 4 II 1904) 449. P. Brijcei Sprague in Dyer. Fl Trop.
Africa. 4 11(1906)515. Calyx campanulate, regular, 5-toothed.
inflate when in flower, disk cupular; ovary oblong, capsule
linear with 8 series of seeds per cell.

Certainly Podranea shows affinities with Pandorea
in which genus Bâillon first arranged P. ricasoliana
•^Prague.

As in Tecomaria the fruit however is hnear and the
plants are lians or undershrubs.

P. Richardii H. Bn. is a lian, with cauliflorous cymes.
The calyx is tubular, the corolla distinctly 2-labiate with
a curvate tube. Disk cone-shaped, ovules in 2 rows per cell.

The Madagascarian lian is rather isolated from all the
preceeding groups on account of its reduced number of
ovules.

Shrubs with simple or pinnate leaves; calyx tubular-
campanulate, often glandular. Thecae hairy, connective
enlarged. Capsule linear with coriaceous concave valves.
Ovules 2-seriate per cell.

Perhaps the American genus shows feeble affinities with
some of the other groups, but it is not certain this is of
true phylogenetical importance.

St. molle Seemann, St. stans (L) D. Don, St. fiilvum
Sprague. St. alatum Sprague.

In a general review of the affinities of the groups it
appears that the first 6 groups, though totally differing
in geographical distribution, show such characteristic com-
mon characters, that in my opinion there must be a nar-
row relation between them as to their origin. This I hope
to make clear in chapter II.

Urban \') made an interesting study on the importance
of the characters of tendrils and pollen-grains for the
distinction of the genera of the Bignoniaceae. As the

1) Uber Ranken und Pollen der Bignoniaccac. Ber. Deutsch. Bot.
Ges. 34 (1916) 728—758.

I have not studied the pollen either. For the distinction
of the genera it will be useful. For phylogenetic relations
it seems to me of hardly any value. Tecomaria FenzL,
Campsis Lour., Campsidium Seem, et Reiss., Pandorea
(Endl.) Spach. and Podranea Sprague have the same kind
of pollen, indeed, viz. quot;dreifurchigen granulierten Pollenquot;.
^arabignonia Bur. and Dolichandra Cham, also possess
this pollen. Both however, have 2-foliolate leaves with
tendrils; moreover, Parabignonia has only 2 rows of
ovules per cell. And Rhigozum Burch., Catophractes Don.
and Digomphia Benth., too, have the same pollen; all
three of them having only 2 rows of ovules per cell and
totally different leaves. On the other hand Deplanchea
Yieill. has quot;3-furchigen Pollenquot; and Delostoma D. Don
furchenlosen Pollenquot;. No doubt however these genera
are closely allied.

KEY TO THE SPECIES.
Leaves narrow 7—11-jugate, leaflets gross-serrate,
small 0.75—1 cm. 1., flowers large 6-7.5 cm.

Leaves mostly 2—4-jugate, seldom up to 7-jugate,
(but then flowers small ca. 2 cm. 1 and leaflets
linear), leaflets mostly larger (if small not peculiar
gross-serrate and then flowers ca. 1.5 cm. 1), flowers
small or large, not surpassing 5 cm. in length ... 2

Flowers always smaller up to 2.5 cm. 1 § Parviflorac 4
Flowers pinkish-yellow, ca. 5 cm. 1, inside orange-
yellow, lobes white, inflorescences racemes, leaflets
5. elliptic-caudate, 5 X 2.5 cm., 0.25 cm. 1 petioled

streaked in the throat, leaflets S—7 (-9), ovate to
lanceolate, obtuse acuminate, 2.5—5 x —2 cm.,
nearly sessile.........3. P. jasminoides.

4.nbsp;Leaves exceeding 60 cm., opposite or in whorls of
3 and 4, leaflets very large 12.5 x 5 cm., pale pro-
minently reticulate at the underside, thyrses 7.5—
45 cm., flowers cream-coloured, the lobes and throat
shaded with pink, small 1.2 cm. 1. 4. P. Bailey ana.
Leaves never reaching 60 cm., but always less than
25 cm., always opposite, leaflets if having a promi-
nent reticulate structure smaller than 12.5 cm., at
most 9 cm. 1. thyrses mostly smaller, corolla yellow,
throat marked with purple...........

5.nbsp;Leaves sub-coriaceous, with 8 distinctly prominent
nerves and reticulations, corolla inside only stupose
at the throat and not on one side of the tube, co-
rolla 2 cm. narrow tubular, tube 0.5 cm. wide, disk

Leaves papyraceous, nerves never distinctly prominent,
mostly less than 8, reticulations lax and indistinctly
prominent, tube inside stupose on one side, 2—3 cm.
1. much broader, seldom tubular, mostly campanulate-
infundibu liformous, disk annular.........

6.nbsp;Leaves 4—7-jugate, leaflets oblong-ovate never linear
or obtuse and then sub-orbicular, but long acuminate
(acumen ca. 1.5 cm.), 3—4 X 1.2—1.8 cm., nerves

Leaves 2-3-, seldom 4-jugate. if 6—7-jugate leaflets
either linear, or obtuse and sub-orbicular, nerves
5—7 or less..................

7.nbsp;Leaflets ovate or sub-orbicular, rounded and always
obtuse at the top, 2—3 (4—7) pairs, never linear,
rather small 3—4 X 2—2.8 cm., sub-coriaceous when
mature, nerves 3—6, thyrse rather small, 6—8 cm.
1, 15—20-flowered. Calyx yellow, corolla outside

sulfureous on one side and yellow-purplish on the
posterior one, inside the anterior part yellow, stupose
and glandular, the posterior one deep-purple coloured

Leaflets 2—4 pairs, in case more, then leaflets linear,
never obtuse and rounded at the top, seldom ovate,
never sub-orbicular, mostly rather large and oblong
to lanceolate, papyraceous, nerves mostly 6—8, if
less leaflets lanceolate, corolla coloured in another

8. Leaflets large,- lanceolate, ovate-oblong and acumi-
nate, the greatest breadth in the middle, rather long
1 — 1.2 cm. petiolulate or sub-sessile, entire, 8—11 x
3.6—4.5 cm.; thyrses small, lateral, 4—14 cm. 1,
corolla small rather narrow 1.5 X 0.5 cm., pale
yellow, purple-brown streaked on the inside of the

Leaflets smaller, most times all sessile, smaller or if
rather large (6—10 cm. 1) and with 7—9 pairs of
nerves, then the greatest breadth in the middle,
thyrses mostly larger, corolla broader campanulate-
infundibuliformous, mostly larger than 1.5 cm.. . . 9
Thyrses puberulous or shortly pubescent,. branchlets
densely lenticellate, leaves robust 10—20 cm. 1,
oblong, rounded or sub-cuneate at the base. 6—10 X

Thyrses entirely glabrous, branchlets elenticellate or
with few ienticels, leaves not robust. 5—10 cm. 1
seldom longer (and then leaflets mostly linear), mostly
cuneate at the base. 2—7 X 2—2.5 cm.

10. P. australis,
1- Pandorea leptophylla (Bl.) Bocrl.
~ = Tecoma {Cyathocoma) leptophylla Bl. Rumphia.
(1848) 35; Blume Mus. Bot. Lugd. Bat. 1 (1849-50)
27; Boerlage. Handl. Fl. Ned. Ind. 2 (1899)600; Miquel.

FI. Ned. Ind. 2 (1856) 758; Miquel in Ann. Mus. Bot.
Lugd. Bat. 1 (1864) 197; Warburg. Pfl. Papuas. Engl.
Jahrb. (1890) 419; K. Schumann Fl. Deutsch. Schutzgeb.
Südsee. (1900) 540; - = Pandorea ? leptophylla (Bl.) Diels.
Engl. Jahrb. 57 (1922) 499; van Steenis Nova Guinea
14 (1927) 301, t. 33.

Many-branched glabrous shrub, sometimes climbing
with roots (Blume); ultimate twigs voluble, sub-herbaceous,
striate or sulcate, very slender 0.1 cm. thick, internodes
1—8 cm. 1. Leaves patent, opposite 8—15 cm. 1., lance-
olate, when young tetragonous, later on sub-terete, remote
or shghtly crowded, 1—pinnate, 7—11-jugate. Petioles minu-
tely alate, sub-terete, base somewhat thickened, 2—3 cm. 1.
connected with a sub-puberulous line; rhachis slightly
alate, contracted near each pair of leaflets. Lateral leaflets
small 0.75—1,5x1,5-3 cm., nearly sessile, peculiarly tra-
peziformous, membranous, acute or obtuse-acuminate,
oblique, margin scabrous, anterior margin entire or 1—4
gross-obtusely toothed, posterior (1)—3^6 gross-obtusely
toothed; terminal leaflet narrow lanceolate, entire or gross-
obtusely toothed, ca. 0.5 cm. 1. petiolulate. Nerves 2—3
pairs, slightly prominent, reticulations indistinct. Lamina
beneath light-green with few scattered small glands as in
Radermachera. Racemes axillar and terminal (Blum e). Calyx
short, cupular-truncate as usually in Pandorea, without teeth
or with 5 equal minute teeth, blush-red, 0.5 x 0.7 cm.,
glabrous, smooth, coriaceous. Corolla rather large, 6—7.5 cm.
1. (incl. lob.) blush-red, inside fuscous-striate, glabrous; tube
rather narrow beneath ca. 0.5 cm. diam., slightly curved
or nearly straight, stupose inside near the insertions of
the stamens ca. 0.5 cm. above the base, infundibuliformous
enlarged to the throat, ca. 2—2.5 cm. diam. at the throat.
5-lobed; lobes inequal. ovate, acute, margin sub-tomentose.
Stamens 4, didynamous, ca. as long as the tube with a

fifth rudiment; filaments fihform, stupose at the immediate
base, glabrous. Style slighdy longer than the filaments;
stigma bilamellate. Disk annular, fleshy. Ovary cylindrical,
sub-compressed, bilocular. Ovules oo in several rows on
each placenta. Capsule unknown.

Geogr. distr. New Guinea — 5. W. New Guinea, coast-
forest. flowering specimen, near Triton Bay?: Zippel
(L sub n. 898,200.. .112, U sub n. 010795).

This very interesting shrubbish lian of the litoral forests
of New Guinea seems to be very rare. Only Zippel
gathered a flowering specimen nearly a century ago perhaps
near Triton Bay. Warburg (1. c. 419) says that it, judging
from the sterile twigs (he knows), the species also occurs
near Finschhafen and Sig ar. I found in the indeterminate
sterile New Guinean materials of Pulle at Utrecht a
specimen with scattered leaves, certainly no Bignoniacea,
the foliage of which was very like that of P. leptophylla,
so that it may be Warburg was mistaken. No German
collector seems so have found any flowering specimen later
on in N. E. New Guinea.

Taxonomically it is very interesting, as it shows affini-
ties with several allied genera.

The habit of the foliage is much like that of Campsi-
(e.g. C. chilense Seem, et Reiss.) also in its flowers,
though the calyx of Campsidium is not as truncate as that
of P.leptophylla.

Though I did not see any thick branches it seems to
ch\'mb now and then with roots, a character pointing to
^ampsis.

Por the rest possessing racemes (axillar as well as terminal,
®fter Blume) and having a rather large corolla it has the
•nearest affinities with Tecomanthe, especially with T.auran-
^\'^ca, but differs totally by its foliage.

between Pandorea and Tecomanthe one may just as well
call it a Tecomanthe, but it remains a link, showing affini-
ties with both genera, though the nearest to Pandorea in
my opinion.

—nbsp;= Tecoma Curtisii Ridl. Journ. As. Soc. S. Br. 49.26;
Fl. Malay Penins. 2 (1923) 553, fig. 125; van Steenis.
Nova Guinea 14 (1927) 302.

Glabrous slender twiner. Leaves opposite, lO—12.5 cm. 1.,
including the 2.5\'—5 cm. 1. petioles. Leaflets 5, elliptic-
caudate, base rounded, 5 X 2.5 cm.; lateral petiolules
0.25 cm. 1., terminal one 0.6 cm. 1., sidenerves 6—7.
Racemes axillar and terminal. Flowers numerous, pedicels
ca. 0.3 cm. 1. Calyx cup-shaped, small 0.6 cm. 1., greenish
purple, 5-toothed. Corolla 5 cm. 1., base cylindric,
funnelshaped above, 2.5 cm. diam., outside pinkish yellow,
inside orange yellow, lobes sub-orbicular, white. Fruit
unknown.

Geogr. distr. Malay Peninsula Rare and local
(Ridley) — Penang, Batu Feringhii| and near the Chitty
Temple (Curtis).

This species which I did not examine for lack of mate-
rials is a most interesting one with respect to the distri-
bution of the genus. The very small calyx points undoub-
tedly to Pandorea; it is the only species known from the
continent of Asia.

It is well distinguished by its 2-jugate leaves, its large
5 cm. 1 pinkish yellow corolla-tube with white lobes.

—nbsp;= Tecoma jasminoides Lindley. Bot. Reg. t. 2002;
K. Schumann in Engler-Prantl. Pfl. Fam. IV. 3b (1895)
230; All. Cunningham in London Hort. Brit. 582; Don.
Card. Diet. 4. 225; D. C. Prod. 9 (1845) 225; Bot. Magaz.

t. 1004; F. v. Mueller. Sec. Syst. Cens. Austr. PI. Part. I
(1889) 166; Bentham. Fl. Austr. 4 (1869) 537; van Steenis
Nova Guinea 14 (1927) 302; Bailey. Compreh. Cat. Queensl.
PI. (1909) 364; Bailey. Queensl. Fl. 4 (1901) 1134.

Tall glabrous woody climber, ultimate branchlets terete,
elenticellate. Leaves opposite, 1-pinnate, 2—3 (—4)-jugate.
Leaflets nearly sessile, ovate to lanceolate, obtuse acuminate,
shining, entire, slightly concave at the base, 2.5—5 x 1—2
cm. or in ovate leaflets 3—5 X 1.5—2.5 cm., not showing
the remarkable variations of P. australis: 3—5 pairs of
nerves, indistinct, reticulations none or minute, microsco-
pically punctate. Thyrse terminal, sub-corymbose, compact,
ca. 3 cm. 1., ca. 3 cm. 1. peduncled. Flowers large, showy.
Calyx glabrous, glandular, ca. 0.6 x 0.45—0.5 cm., trun-
cate or 5-toothed, teeth equal broad-triangular, acute.
Corolla infundibuliformous-campanulate, delicate milkwhite
rose-red streaked in the throat, ca. 4 cm. 1., short pubes-
cent outside. Tube ca. 0.8—1 cm. br., inside near the
insertions of the stamens ca. 0.5 cm. above the base
stupose. Limb flat, expanded; lobes rounded somewhat
waved and crenate. very broad, ca. half as long as the
tube, inside pubescent; throat scarcely bearded or inside
marked with 2 decurrent lines of short hairs. Stamens 4,
didynamous with a fifth rudiment, included; filaments gla-
brous. Stigma bilamellate. Disk annular-cupuliformous.
Ovary 2-celled, each cell with 2 indistinctly separated
placenta\'s; ovules in many series on each placenta. Capsule
similar, but longer than that of P. australis, seeds rather
broader, almost obcordate, the wings either entirely sur-
rounding them or chiefly .on the 2 sides.

Geogr. distr. Australia — Queensland — Brisbane river,
Moreton Bay (A. Cunningham, F. v. Mueller); Burdekin
river (F. v. Mueller); Ipswick (Nernst) — New South Wales
— Richmond river (Henderson); Clarence river (Beckler ?)

Malayan Archipelago — Cultivated in Hort. Bog. sub
X. F. 133a. (B); idem sub n. 6012 (B); K. Bidara Tjima
pr. Meester-Cornelis: Edeling (B); Gamboeng, north of
Pengalengan. Bandoeng: Kds. 42081 p (B); Telagapatengan,
south-west of Bandoeng, cultivated in garden, 1650 m.
alt.: Backer n. 12533 (B).

The Australian climber differs distincdy from P. australis
and from other Pandorea s in its large, white corolla, the
throat being rose-red streaked, the rather compacted in-
florescence and the glandular corolla.

It is often cultivated in the tropics (so e. g. in British
India, Chutia Nagpur, according to Wo o d in Rec. Bot.
Surv. Ind. 2 (1902) 125) and in the temperate regions: in
the latter it is treated as a greenhouse-plant.

The leaflets are often infested with a blight-fungus,
Melasmia tecomatis C. et M. (Bailey).

4. Pandorea Bailcyana (Maid, ct Bak.) vSts. nom.
nov. — Tecoma Baileyana J. H. Maiden et R. T. Baker
Proc. Linn. Soc. New. South Wales. Sec. ser. 10 (592),
Pl. 51; Bailey Queensl. Fl. 4 (1901) 1134; F. M. Bailey
Compr. Catal. Queensl. PI. (1909) 364.

Tall woody glabrous climber. Leaves opposite 1-pinnate,
3~4-jugate, sometimes exceeding 60 cm., mosdy 15—25
cm. 1.. Branchlets ca. 0.4 cm. diam., ribby-alate; bases of
^he petioles thickened leafscars connected with each other.
Petioles 3—5.5—7.5 cm. 1., sub-terete, rhachis thickened
quot;ear the insertions of the thickened bases of the petiolules;
^^aflets oblong, distinctly acuminate at the top, margin
entire, chartaceous, oblique and rounded or slightly tapering
at the base, nearly sessile; 10—11 sidenerves, just as the
reticulations distinctly prominent at the underside; shining
dark-green above, dull pale-green beneath. Inflorescences
thyrsoid rather multiflorous, axillar, 7,5—45 cm. 1., mostly

15—20 cm. 1.; peduncle terete, 0.25 cm. diam., side-axis
2—3 cm. 1., 7-flowered on the average; calyx articulated
with the pedicel by a hypanthium, truncate, irregularly
lobed or more or less distinctly 5-lobed, 0.3 x 0.4 cm.,
smooth, glabrous, purplish-brown. Corolla tubular, incurved,
1.2—1.3 cm. 1., short tomentose outside, cream-coloured,
lobes and throat shaded with pink; lobes almost equal,
sub-orbicular, densely tomentose, tube inside somewhat
bubbled, shghtly hairy. Stamens 4, didynamous, included,
glabrous, cells ovate-oblong, style glabrous, exceeding the
stamens, stigmatic lobes sub-orbicular, crenate. Ovary sub-
orbicular or ovate. Capsule unknown.

Geogr. distr. Queensland — Near the Tweed River:
W. Bauerlen (Bailey) — New South Wales — pr.
Mullumbumby: W. Bauerlen (Kew).

Without doubt this species belongs to Pandorea {Endl.)
Spach. I did not see any 3- or 4-verticillate leaves as
Bailey mentions. Abnormous flowers occur sometimes,
showing 5 stamens (3 longer, 2 shorter) and 10 equal
corolla-lobes. Just as Tecomanthe Hillii [F. v. Muell.)
vSts. the species seems to be rare.

It is nearest allied with P. stenantha Diels, on account
of the structure of the leaves and the tubular small corolla,
and not allied with Australian Pandorea s.

Large glabrous lian, stem ca. 2.5 cm. thick, slightly
rough, sulcate, nearly elenticellate; bark light-gray. Leaves
pinnate, 3-jugate, sub-coriaceous. 10—15 cm- 1.; petiole
2.5—3 cm. 1. with several dark, not prominent ovate-
oblong glands 0.2—0.3 cm. 1., papillose-puberulous; leaf-

lets sessile or up to 1 cm. 1. petiolulate, oblong-acuminate,
shortly serrate at the top, oblique at the base, all ca. of
the same length, 6—-8. 5 X 3.5—4 cm., lateral nerves ca.
8, curved, below as well as the reticulations prominent,
connected with each other near the margin. Thyrses
axillar and terminal, 10—25 cm. 1. peduncled, many-flo-
wered, especially dense at the top, minute puberulous.
Peduncle slender ca. 0.1 cm. thick, naked below, except
single opposite bracts at the immediate base, papillose-
puberulous. Bracts minute-triangular, ultimate stalks 3-flo-
rous. Pedicels thin, ca. 0.5 cm. 1. Ca/yx glabrous, smooth,
articulate with the pedicel by a short hypanthium, campa-
nulate-truncate, 0.4—0.5 x 0.3—0.35 cM., teeth small ine-
qual, apiculate, broad triangular. Corolla with a yellow,
narrow tube 2 x 0.5 cm., outside minute papillose pubes-
cent, inside glabrous, 5-lobed, lobes short 0.3—0.4 cm.
broad, ovate or sub-orbicular, inside pubescent at the
throat, sub-inequal, 3 anterior ones emarginate, white.
Stamens 4, included, didynamous, anthers divaricate,
filaments glabrous, with scattered minute glands. Ovary
oblong, glabrous, irregularly sulcate when dry. Disk cupu-
lar-annuliformous. Ovules oo on 2 placenta\'s per cell in
many series. Capsule onknown.

Geogr. distr. Neiv Guinea — N. E. New Guinea,
Sepik-district, on inclinations near the April river in dense
®oist primeval forest 2—400 m. alt.: Ledermann n.
9809, 2 specimens (D).

This species is characterized by its small narrow
flowers, its somewhat cupular disk, and the glands on the
petioles.

■ ^t is closely allied with P. Baileyana {Maid, et Bak.)
quot;Sfs. from E. Australia.

— Tecoma ceramensis Teysm. et Binnend. Nat.
Tijdschr. Ned. Ind. 25 (1863) 412; H. Bâillon. Hist. PI.
10 (1891) 40; Miquel. Ann. Mus. Bot. Ludg. Bat. 1 (1864)
197, t. 5; — = Pandorea ceramica [T. et B.) Baill.
K. Schumann in Engler-Prantl. Pfl. Fam. IV 3b (1895) 230;
Boerl. Handl. Fl. Ned. Ind. 2 (1899) 600; van Steenis
Nova Guinea 14 (1927) 302.

Stem voluble, branches terete, ultimate branchlets 0.3
cm. diam., sub-quadrangular, bark brownish, striate when
dry, with few small lenticels. Leaves opposite, 1-pinnate,
(4.) 5^7-jugate, 15—25 cm. 1.; petioles 3—4 cm. 1.,
thickened and terete at the base, sub-sulcate above, eglan-
dular or indistinctly glandular towards the base; rhachis
striate, thin 0.1 cm. diam. minutely alate towards the apex.
Leaflets papyraceous, sub-sessile, dark-green and shining
above, paler beneath, margin entire or gross-dentated,
obhque at the base, oblong-ovate and long acute-acuminate
(acumen ca. 2—2. 5 cm.) or even caudate, the upper pairs
even lanceolate-acute-acuminate, mostly 3—4 X 1.2—1.8
cm., rounded; or sub-cuneate at the base, lateral ones
sessile or subsessile, terminal 1 cm. 1. petiolulate; nerves
5—7, slightly sunken in above when dry, indistinctly pro-
minent beneath, reticulations lax, indistinctly prominent,
underside distinctly glandular, glands pezizaeformous. small,
orbicular. Thyrses axillar and terminal, ca. 15 cm. 1..
puberulous, rather multi-flowered; peduncle terete, at the
base with some small opposite or verticillate cupular-con-
nate bracts; primary axis being cymes. 2—7-flowered;
pedicels slender 0.8 cm. 1. with 2 minute opposite brac-
teoles at the middle. Cahjx fulvous, cyathiformous or
cupular-campanulate. truncate, mostly indistinctly irregu-
larly toothed, articulating with the pedicel with a short
hypanthium, slightly puberulous, 0.25 X 0.35 cm. Corolla

tubular-campanulate, rather broad, glabrous, outside yellow,
inside brown-purple striate and spotted, tube 2 X 1 cm.;
lobes pale yellow, at both sides shortly pubescent, sub-
orbicular, 0.9 cm. diam. ; tube inside on one side stupose.
Stamens 4, didynamous, inserted ca. 0.2 cm. above the
base of the corolla-tube; filaments thickened at the base
and glandular there, further glabrous; anthers oblong,
cells divaricate. Disk annular. Ovary oblong, glandular
as well as the base of the style, 2-locular, with 2 placenta\'s
in each cell and oo ovules in several series on each placenta.
Capsule oblong, rostrate at the apex, sub-compressed,
quot;i-angular, 8 X 2.5 cm. ; valves coriaceous, septum thick,
shining, glabrous. Seeds oo, transverse alate, obcordate,
brownish.

Geogr. distr. Malayan Archipelago — Ceram — in
forests: Teysmann n. HB. 5050 (L, B, U) - Ambon:
Porsten (L sub n. 898, 200... 89, n. 898, 200...
88. n. 898, 200... 85, U sub n. 010796); in forests near
the coast, G. Nona or Noesa Niwi: ƒ. Bo es veld (B).

Java - Cultivated in Hort. Bog. sub n. X. F. 118 (B.);
Cult, in Hort. Bogor. ex. Ceram sub n. 6613 (B); Cult.
^^ Hort. Bog. ex Klein Kei, pr. Toeal sub n. 32 (B).

The species occurs at low altitudes in Ceram, Ambon,
and the Kei Archipelago. It is closely allied with P. Poin-
^illantha. See for the differences the notes about that
species on p. 859.

The specimens cultivated in the Hort. Bogor. ex Kei
slands differ from the type, showing broader and larger
\'eaflets (5-7x2.5—3 cm.). Though the preserved specimens
^•^e sterile, I suppose them to belong certainly to P. cera-
^ensis: the leaves are 5-6-iugate.

Voluble, glabrous lian, ultimate twigs thin, terete, striate;
bark yellowish when dry. Leaves opposite, 1-pinnate. 2-3
(-4-7) jugate, 10-13 cm. 1., upper ones under the inflores-
cence sometimes 3-foliolate. Petioles 1.5-4 cm. 1., terete,
thickened at the base, eglandular, connected with a pro-
minent line; petiole and rhachis sulcate above; internodes
1.5 cm. 1.. Lateral leaflets sub-sessile, terminal 0.5-1 cm. 1.
articulate-petioled, sometimes connate with one of the
pair immediately beneath; 3-4 x 2-2.8 cm. 1., concave,
mostly broad elliptic or even sub-orbicular, seldom sub-
oblong, rounded and oblique at the base, rounded and
obtuse at the top; margin entire or with few gross teeth
(2-4) at the top, revolute when dry; sub-coriaceous when
mature, shining above, opaque beneath, narrowly running
down along the petiole, microscopically punctate and with
scattered sunken glands at the underside; nerves 3-6,
curved, connected with each other near the margin, pro-
minent, reticulations lax, indistinct. Thyrse 6-8 cm. 1. inclu-
ding a 2-4 cm. 1. peduncle, few (15-20) flowered, cymes
opposite, 1-3 flowered, pedicels very slender, ca. 0.5 cm. 1..
bracts minute, oblong. Calyx yellow, small, ca. 0.3 cm. 1.,
truncate, cupular-campanulate, ciliate, indistinctly 5- toothed,
shordy mucronate, articulating, with the pedicel with a short
hypanthium, glabrous, smooth, membranous. Corolla outside
yellow on one side and yellow-purplish on the other (poste-
rior), campanulate-infundibuliformous; tube 1.3 X 0.7 cm.,
glabrous outside, inside on the anterior part yellpw, glandular
and stupose, the other side being purple and glabrous;
lobes 5 inequal, yellow, at both sides dense tomentose,
sub-orbicular, 0.4 cm. diam.. Stamens 4, didynamous, fifth
rudimentary, filaments arcuate, slightly thickened at the
base with few glands, inserted 0.2 cm. above the base of

the tube. Anthers ovate, cells divergent. Ovary sub-orbicular,
compressed, each cell with 2 distinctly separated placenta\'s :
ovules 00 in 6-7 series on each placenta; ovary glandular
as well as the basal part of the style, yellowish. Stigma
bilamellate, lobes orbicular, ciliate. Capsule unknown.

Geogr. distr. New Caledonia — N.W. New Caledonia
~ Mt. Balade: Vieillard n. 1002, n. 538 (P); Mt. Panié,
thyrses only lateral, leaves 3-4-jugate: R. H. Compton
n. 1804 (P); Mt. Canola, serpentine mountains 700 m. alt.,
leaves even 7-jugate : M.T. Le card (P); forest of Guaro:
de Pompéry (P).

This species is well characterized with its obtuse, small,
almost ovate or even sub-orbicular leaflets, the varying
number of pairs 3-7 and the colour and structure of its
small corolla.

It seems to me to be allied with P. ceramensis, Poincil-
lantha, stenantha and australis, but to differ distinctly
from them.

It occurs at medium altitudes in the forestrial north-
western part of the isle.

Large glabrous Ham stem terete 2 cm. diam. ; bark
quot;^ough, splitting, argentate when dry. Branches sub-quadran-
Qular. brownish, firmly striate, densely lenticellate, Ienticels
small nearly 0.05 cm. diam.. Leaves 1-pinnate, 3-(ormore?)
jugate, ca. 27 cm. 1.; petiole striate when dry, glabrous,
with many small glands at the lower half ; rhachis articulate,
sub-terete, sulcate above, internodes 4-5 cm. 1.. Inferior
leaflets 1-1.2 cm. 1. petiolulate. upper ones nearly sessile,
^11 entire, papyraceous, 8-11 X 3.6-4.5 cm., ovate-oblong
With the greatest breadth below the middle, apiculate or

acuminate, sub-truncate and oblique at the base, somewhat
concave, shining above; lateral veins ca. 9 pairs, sub-
straight, connected with each other at the margin, a little
prominent, reticulations lax, indistinct. Thyrses lateral,
puberulous, on the old wood as well as on the young
branches, narrow, rather lax but multi-flowered, 4-14 cm. 1.,
sometimes foliate at the base, short ca. 1 cm. 1. peduncled;
stalks slender, the peduncle at the immediate base with
2 small opposite broad cupular-connate bracts; the opposite
primary axis are 3-flowered cymes; pedicels 0.7-1.1 cm.
1. with short scale-like bracteoles. Ca/r/A: sub-campanulate
minute, 0.2 cm. 1. on a short 0.1 cm. 1. hypanthium arti-
culate with the pedicel, 5-toothed; teeth equal, minute
pubescent, cihate, acute, mucronulate. Corolla small 1.5 x 0.5
cm., pale yellow purple-brown streaked on the inside of
the lobes; tube sub-campanulate glabrous outside, stupose
inside on one side; lobes unequal 0.3-0.4 cm. 1.. pubescent
at both sides. Stamens 4 didynamous, fifth rudimentary,
included. Fruit unknown.

Geogr. distr. New Guinea — 5. New Guinea pr. Kloof-
bivak, Lorentz river. 40 m. alt. in primeval forest - Pulle
n. 1223 (L, B.).

This species is well characterized with its small, pale
yellow, narrow flowers, the lobes being dark-purple
streaked inside, the rather large leaves, with apiculate
leaflets, the glandular petiole reminding that of P. stenantha
and the minute, 5-toothed calyx.

It is allied with that species, but differs in the smaller
calyx, the rather short peduncled inflorescences, which are
not contracted at the end, the smaller flowers, the lobes
being not white, but pale yellow, dark-purple streaked.
Moreover the leaflets of P. stenantha ate dentated towards
the top, are smaller and more elliptic-oblong, having the
greatest breadth in the middle; the structure of the leaflets

of P. stenantha is more firmly papyraceous and the
reticulations are narrower and more distinctly prominent.

The petioles of both species are glandular, but the
glands of P. stenantha are less in number and larger.

Most closely allied with P. Poincillantha. This species,
however, has a much broader corolla, a truncate-cupular
calyx with indistinct teeth, smaller leaves and long acumi-
nate leaflets with the greatest breadth mosdy at the middle.

9. Pandorea Poincillantha (Zipp.) vSts. — = Bignonia
Poincillantha Zipp. Nomen in Herb. Lugd. Bat.; van
Steenis Nova Guinea 14 (1927) 302; - = Tecoma ceramensis
T. et B. var. elliptica Miquel Ann. Mus. Lugd. Bat. 1
(1864) 198; SchefFer. Ann. Buitenz. 1 (1876) 40.

Glabrous climber ca. 4—6 m. (or more?) high; branch-
lets 0.5 cm. thick, terete, firmly striate, densely lenticellate,
lenticels minute, brownish. Leaves 1-pinnate. 10—20 cm.
L. 2—3 (-4).jugate, opposite, glabrous; petioles 2—4 cm.
L thickened at the immediate base, connected with a
prominent line, glabrous, sub-terete, the lower half glan-
dular. internodes 2—3 cm. 1.; leaflets firmly papyraceous,
glabrous, shining dark-green above, entirely or (seldom)
gross-dentated towards the top. lower pairs shortly 0. 2—0.5
cm. petiolulate. upper ones sessile, terminal 1—2.5 cm. •
1- petiolulate, all elliptic-oblong, oblique and rounded or
sub-cuneate at the base, acute or mostly acuminate, seldom
caudate at the top, 6—10 x 3—4 cm.; sidenerves 7—9
pairs, a little prominent, reticulations lax. indistinct, under-
side of the leaflets sometimes glandular. Thyrses lateral
^nd terminal 10-15 cm. 1.. 1—3 cm. 1. peduncled; pe-
duncle terete, 0.3 cm. thick, puberulous, at the base thickened
and with opposite or verticillate minute cupular-cuneate
bracts. Thyrse puberulous, especially multi-flowered at the
top, the side-axis are (3)-7-flowered cymes; stalks slender;
pedicels 0.5—0.8 cm. 1. puberulous, bracts minute, with 2

opposite bracteoles in the middle. Calyx cupular-campa-
nulate, 0.2 cm. 1., truncate, minutely equally 5-toothed,
margin ciliate, articulated with the pedicel with a short
hypanthium. Corolla yellow, not fragrant, 1.5—2 cm. 1.
sub-campanulate-infundibuliformous, inside on one side
stupose; tube 1 X 0.8 cm., glabrous; lobes 5, sub-equal,
sub-orbicular 0.5 cm. diam., short pubescent on both sides,
marked with brown-red inside. Disk cone-shaped. Stamens
4, didynamous, inserted near the immediate base of the
tube 0.1—0.15 cm. above it, fifth rudimentary; filaments
glabrous, white, thickened at the base and densely covered
with small glands there, nearly as long as the tube;
anthers divaricate brown-purple, on curved filaments, obtuse,
elliptic. Ovary glabrous, glandular at the top or through-
out, style glabrous, smooth or glandular at the base,
small ca. 0.4 cm. 1., 2-celled; each cell with oo ovules in
many rows on 2 placenta\'s. Capsule unknown.

Geogr. distr. New Guinea — N. Guinea: Zippel (L.
sub n. 898, 200... 92, n. 898, 200... 91, 898, 200...
90, n. 898, 200. . . 88 (partim), n. 898, 200... 86); North
New Guinea, Humboldt Bay, bank of the Mbaai river,
10 m. alt., 1 specimen: K. Gjellerup n. 971 (L, B.);
North New Guinea, Bivak Hollandia, river bank, 10 m.
alt.: K. Gjellerup n. 405 (B, U, L, D,Kew); banks of
the mouth of the Mamberamo river: R. F. Janowsky
n. 433 (L, B, D, Kew, U).

Geogr. distr. Neiv Guinea — Meervlakte, Motor-bivak,
bank of the Brown river, alt. 100 m., flowers yellow,
fragrant: W. M. Docters van Leeuwen n. 11152
(L, B).

This species is well characterized by its striate, slightly
ribbed dense lenticellate branchlets (lenticels minute), corolla
rounded at the base, the filaments which are thickened

and glandular at the base and inserted nearly at the base
of the tube and the rather broad corolla-tube, which is
inside stupose on one side.

It is allied with P. acutifolia, from which it differs in its
foliage, the large thyrses, the cymes of which are mostly
7-flowered, the truncate or obtuse-toothed calyx, the
filaments, which are thickened and glandular at their base
and the glandular ovary.

Moreover it shows affinities with P. stenantha.
It is most closely allied, however, with P. ceramensis,
even that one might suppose the 2 species being both of
the rank of a sub-species of one species P. ceramensis.
As no capsule of P. Poincillantha is known for the
present, it seems better to keep both as distinct species,
though closely allied. For easiness\'sake I give here the
principal points of difference.

Leaflets rather large when Leaflets rather small, when
mature, 6-10 x 3-4 cm.nbsp;mature3-4 X 1.2-1.8cm.,

10. Pandorea australis (R. Br.) Spach.
R. Brown. Prod. 471; ed. 2 (1827). 327; Spach. Hist.
Nat. Veg. Phan^rog. 9 (1840) 136; Sims, in Bot. Magaz.

(1795) n. 865; Maund. Botanist t. 8; — = Bignonia
Pandorana Andr. Bot. Rep. t. 81; — = Bignonia Pan-
dorae Vent. Malm. t. 33; — = Bignonia australis Ait,
Hort. Kew. ed. 2. IV. 34; Bailey Queensland Fl. 4 (1901)
1134; — = Tecoma australis R. Br. D. C. Prod. 9(1845)
225; — = Tecoma floribunda Cunn. in D. C. 9 (1845)
225; - =: T. diversifolia G, Don. 4. 225; D. C. Prod. 9
(1845) 225; Bailey Compreh. Cat. Queensl. PI. (1909)
364; Bailey in Queensl. Agric. Journ. Bot. Bull. 7 (1900)
349; F. V. Mueller Sec. Syst. Cens. Austr. PI. 1 (1889)
166; Bentham Fl. Austr. 4 (1869) 537; Diels Engl.
Bot. Jahrb. 57 (1922) 498; van Steenis Nova Guinea
14 (1927) 302.

Tall glabrous woody climber, branchlets twining. Leaves
10-15 cm. 1. very variable, opposite. 2-4-(6)-jugate. Leaf-
lets ovate-oblong, ovate-lanceolate or almost linear, entire
or here and there coarsely crenate, 2-8 cm. 1, exceedingly
variable, all small or all large, sometimes, especially on
barren shoots, all coarsely toothed and then occasionally
all very small and much more numerous, underside beneath
glandularly punctate, rhachis and petiole glabrous, with-
out glands, sidenerves 5-7, indistinctly prominent, reti-
culations indistinct. Thyrses glabrous, mostly loose and
multi-flowered, but also sometimes more contracted or
pauciflorous, leafy at the base, primary axis opposite.
Calyx smooth, 0.2-0.3 cm. 1. Corolla mostly yellowish-
white, tube 1.2-2.5 cm. 1., slightly curved and dilated
upwards, lobes broad not 1/3 as long as the tube, the
2 upper ones rather smaller with purple or red spots at
their bases, the throat bearded inside under the lower
lip, tube bearded on one side and with small glands. Tube
glabrous near the insertions of the stamens. Stamens 4,
didynamous, glandular at the base, inserted near the base
of the corolla. Disk annular. Ovary glandular, sub-com-

pressed, 2-celled, each cell with 2 placenta\'s, each placenta
■with oo ovules in several rows. Capsule 3-7.5 cm. 1. acute
or slightly acuminate at both ends; valves hard and very
concave. Seeds flat oblong or rounded, surrounded by a
broad wing.

This species is ultimately variable and Bailey (I.e. 1134)
distinguishes 3 forms, which he calls varieties. For the
present it seems to me the best to keep these forms,
though I mean they ought to be considered as sub-species,
because they vary in many rather important characters and
intermediate forms are present.

Diels (I.e. 498) who worked out the New Guinean
materials of P. australis quot;findet aber keine Möglichkeit,
diese Exemplare von der bisher aus Ost-Australien und
auch dort sehr vielförmigen Pflanze zu trennenquot;.

It will be necessary in the future to give a more dis-
tinct account of the sub-species, varieties, and forms,
especially by Australian botanists based upon fieldknow-
ledge and cultivation of P. australis.

s. sp. Pandorea Bailey — Queensl. Fl. 4 (1901) 1134;
Compreh. Cat. (1909) 364; — —Bignonia Pandorea Ventetat.
lard. Malm. t. 43; Bot. Mag. t. 865; Maiden and Campbell
Pl. PI. and Ferns. N. S. W. n. 11 ; — = Tecoma Latrobei

Leaflets on the flowering shoots 3, the lateral ones more or
less oblique at the base, smaller than the terminal one, which
is from 5—7. 5 cm. I., base broadly rounded, from whence
narrowed into a long acuminate sharp point; terminal petioles
\'ong, lateral ones short. Terminal panicle often wide sprea-
ding. Flowers 2.5 cm. or longer, expanding upwards to nearly
1-8 cm. wide, emitting a strong disagreable odour, capsule
dark-brown, the largest about 8 cm. 1., valves stiff.

Geogr. distr. New Guinea (I did not see the New
Guinean specimens cited below, but probably they belong
to this s. sp.) N. W. New Guinea — South river, 400 m.
alt.: Moszkowski n. 423;pr. Paraido: Moszkowski
n. 430 ^ N.E. New Guinea — Mt. Kani, 1000 m. alt.:
Schlechter n. 18271; forests of the Maboro: Schlech-
ter n. 19861; in forests in Mt. Gati: Schlechter n.
16998.

Queensland — Rockingham Bay: Dallachi (U, L, P) —
New South Wales — Port Jackson pr. Sydney (L sub n.
898, 200... 68); Blue Mountains pr. Sydney (P); pr.
Wingham: J. L. Boorman (L sub n. 910, 151... 539);
Paterson river: J. L. Boorman (L sub n. 908, 146...
2205); Australia Felix: F. v. Mueller (U).

I saw the following cultivated specimens: cultivated in
a temperate greenhouse Jard. PI. Paris, 1862 (P); Hort.
Noisette, 1876 (P).

s. sp. meonantha Bailey — Queensl. Fl. 4 (1901) 1134;
Compreh. Cat. (1909) 364; — = Tecoma meonantha G. Don
Syst. 4. 24; — = Bignonia meonantha Link. Enum. Hort.
Berol. II. 130; — Tecoma australis fi ? meonantha D.C.
Prod. 9 (1845) 225; Illustr. Bot Cook\'s Voyage. 71, t. 228.

Leaflets on the flowering shoots 3 or 5, usually oblong,
mucronate. Panicles usually very narrow. Flowers fragant,
seldom exceeding 1.25 cm. in length, narrow. Capsule the
largest about 6 cm., valves rather thin.

Geogr. distr. New Guinea — Alexishafen, flowers yellow,
inside blood-red striate, fragrant: P. Fr. Wiesenthal
n. 46.

Australia — Queensland — Moreton Bay: F. v. M u e 11 e r
(P, U, L, sub n. 900, 81... 135); Genoc river: A. Mar
6 Tailor (L sub n. 898, 200... 66); Australia Felix:
F. V. Mueller (U. P, L); Burdekin river: F. v. Muel-
ler (P); Australia, without locality (L sub n. 898, 200
. .. 72);exherb.Beaudoin(P); Banks and Solander(P) —

New South Wales — Port Jackson: L h o t s k y (P); pr.
Sydney: Arnony n. 44 (P); Port Jackson: R. Brown,
type specimen of Tecoma australis R. Br. Prod,? (P); New
South Wales: M. Busseuil (P): pr. Sydney: M. Ver-
reaux n. 53 (P); Port Jackson: Beaudoin n. 68 (P);
Port Denison (L sub n. 898,200... 65); pr. Fitzalan (L sub
n. 898, 200.. . 63); — Victoria — Dandenong Ranges pr.
Melbourne: F. v. Mueller (P, L sub n. 900, 256...
449, n. 900, 256... 448, n. 898, 200... 46); Victoria
river; F. v. M u e 11 e r (P); Victoria: D u n c a n (P); Victoria
(L sub n. 898, 200... 61). Tasmania — sub Pandorea?(P).

I saw the following cultivated specimens: Garden Antibes,
fruiting specimen: Thür es (P); Korfoe; Cassa Rosa:
C. Baenitz (P).

As the specimen of Antibes showed ripe fruits, the
species seems not to be self-sterile, this being probably
often the case with the Bignoniaceae.

s. sp. linearis Bailey — Queensl, Fl. 4 (1901) 1134, pi.
45; Compreh. Cat. (1909); 364 — = Tecoma Oxleiji
A. Cunn. in D. C. Prod. 9 (1845) 225.

Leaves remote, from 4—6 pairs and a terminal leaflet,
this one the largest, 2.5—5 cm. 1. and seldom exceeding
0.4 cm. broad. The lateral ones scarcely half the size of
the terminal one. Panicles on long peduncles, rather short,
■i—6 cm. 1., pauciflorous, about the size and form of s. sp.
meonantha. Capsule 2—2.2 X 0.4—0.6 cm.

Geogr. distr. Australia — Queensland — Herberton:
J- F. Bailey — New South Wales A. Cunningham
n. 386(U); pr. Coolabah; J. L. Bo or man (L sub n. 918,
3... 199); pr. Coolabah: Maiden and Boorman (P).

Pandorea Ricasoliana H. Bn. (Hist. PI. 10 (1891) 40) =
f\'odranea ricasoliana Sprague (in Dyer. Fl. Cap. 4 (1904)
-150) - S. E. Africa.

I do not know any Pandorea from Java and even do
not suppose that any Pandorea will be found in this island,
which is worked out for a great deal nowadays. Boer-
lage (Handl. Fl. Ned. Ind. 2 (1899) 591) does nôt sup-
pose the genus to occur in Java either.

Hist. PI. 10 (1891) 41; K. Schumann in Engler-Prantl.
Pfl. Fam. IV 3b (1895) 230; Boerlage. Handl. Fl. Ned.
Ind. 2 (1901) 590-1 ; L. Diels. Bign. Papuas. Engl. Jahrb.
57 (1922) 496; S. Moore. H. O. Forbes\'s New Guinea
Plants. Journ. Bot. 61 (1923) 38 ; van Steenis. Nova Guinea
14 (1927) 294.

The genus Tecomanthe is founded by Bâillon on
T. Bureaui Bâillon of which only an imperfect specimen
was gathered near Triton Bay by H o m b r o n. Only a
branchlet with a naked peduncle and some flowers are
present. So Bâillon (p. 11) says: quot;Le Técomanthe de
la Nouvelle Guinée, est une liane, a fleurs portées sur le
bois des branches volubiles, et qui parait voisine des
genres précédents, mais dont on ne connaît ni les feuilles,
ni le fruit.quot; Bâillon thus supposes the alliance with
Pandorea, Campsidium, Campsis, Tecomella and Tecomaria,
though he had seen neither the capsules nor the leaves.
And so did Blume. It is again a proof of the existence
of a certain quot;systematical intuitionquot; founded upon a quot;syste-
matical visionquot;, as I would say, that Blume (1848) already
placed the 3 Panc/orea\'s and the Tecomanthe s known then, in
the genus Tecoma (Campsis), though Blume also did
not know anything about the fruits. That he mentioned
the same about Tecoma (Campsis) cuspidata (=Nyctocalos
cuspidatum) is due to the fact, that sterile specimens of

Bentham amp; Hooker (Gen. PI. 1045), who adopted a
large genus Tecoma with several sections, which later on
have given raise to distinct genera, placed the Tecomanthe\'s
of Blume in the section 4. Pandorea.

The genus Tecoma in those days (1875) composed of
truly heterogeneous elements, was limited later on to
American climbers,quot; the species of the palaeotropics being
arranged in several new genera. This is amongst other
things the case with our Tecomanthe (see for this also
under Pandorea p. 836).

Bâillon marks the type as a doubtful Tecomea (1891)
the capsule being unknown. He says: quot;51.? Tecomanthe
H. Bn. Flores fere Campseos ; calyce membranaceo : lobis
5 inaequali-deltoideis ; majoribus 2. Corolla e basi angusta
sensim ampliata; limbi sub-2-labiati lobis acutiusculis.
Stamina didynama, inclusa ; loculis divergentibus. Stamino-
dium antheram sterilem gerens. Discus annularis. Germen
brevissime stipitatum; ovulis utrinque in septo oo seriatis;
styli lamellis stigmatiferis longe rhomboideis — Frutex
scandens; caule volubili (haud? cirrhoso) foliis. . . .?;
floribus amplis (ad 1 dm.) speciosis in ligno ortis. Genus
in ordine valde anomalum, ob specimen valde mancum,
»nale notum. (Nova Guinea. Spec. 1. T. Biireaui H. Bn.)quot;.

K. Schumann does the same, as he says (p. 232) that
he does not know other species than the single T. Bureaiii,
^hich he did not see. Later on Schumann recognized
Thecoma dendrophila Bl. as a Tecomanthe, In the key he
gives in the Pfl. Fam., he mentions as the only difference
between Pandorea and Tecomanthe the cauliflory of the
latter, against the flowers on foliate twigs of the former.
In the description he adds that in Tecomanthe the calyx

is unequally 5-lobed, the corolla-tube being campanulate
with a lower narrow part, the disk being annular; whilst
in Pandorea the calyx is small with 5 teeth, the corolla
being funnel-shaped or nearly campanulate and the disk
being cushion-like.

Boerlage (p. 590—1) discusses the systematical place
of Pandorea and Tecomanthe. He knows T. amboinensis,
T. leptophylla, P. ceramensis and T. Bureaui. The latter
species he did not see, otherwise he would have distin-
guished its immediate alliance with T. amboinensis. The
former three species belong in his opinion to Pandorea.
Thus Boerlage did not concept properly Tecomanthe
and Pandorea, but only Pandora, and for a great deal
he was quite right in his supposition, following Bentham
amp; Hooker (Gen. PI. 2. 1045), which authors placed T.
dendrophila Bl. also in the sub-genus Pandorea.

As I have pointed out in Nova Guinea (p. 294) the two
genera are not sharply separated by means of important
taxonomical differences, but are connected with 3 links.

In my opinion this fact is exactly the ultimate expression
of a true phylogenetic system, wherein the groups, small
as well as large are connected with each other according to
the evolution-hypothesis. Indeed the bulk of the Teco-
manthe s and Pandorea s are easily recognizable and there-
fore I accept 2 genera, with the remark that they both
belong to a large „Sippequot; with some other genera, in
earlier times united in Tecoma, thus having undoubtedly
a common origin, which will be worked out in chapter II.
It is of course necessary to give a new diagnosis, as there
are described a lot of new species in both genera. It is
a pity that only of T. nitida a capsule without seeds is
known, but it is surely sufficient, as it shows strikingly
the same structure as those of several Pandorea\'s.

rami lignosi, interdum radicibus juxta cicatrices foliorum
praediti, folia opposita, 1-pinnata, 1-5-jugata, linea dis-
tincta prominente inter sese connecta, rhachibus semi-
teretibus vel alatis; foliola opposita, plerumque foliolo
terminali excepto, breviter petiolulata, papyracea vel cori-
acea, glaberrima vel sub-glabra, Integra vel grosse dentata.
Inflorescentiae racemosae, rariter thyrsiformes {T. montana
Diels et T. sa;cosa Diels), solitariae vel geminatae, semper
in axillis foliorum vel e ligno ortae, plerumque breves usque
ad 5 cm. longae, rarius longiores; pedunculi infima basi
semper squamis minimis distichis 1-vel pluri-jugis, praeterea
nonnumquam fpliis rudimentariis praediti. Pedicelli brac-
teolis 2 bracteaque 1 praediti. Calyx campanulatus interdum
sub-urceolatus, magnus, 1.5—4 cm. longus, 5-lobatus, glaber,
eglandulosus, lobis plerumque margine tomentosis, rariter
parvus {T. montana Diels) vel sub-truncatus obscure
5-dentatus {T. aiiranf/aca Die/s). Coro//a magna, 5—10 cm.,
rariter minor (T. montana Diels), recta vel curvata, glabra
vel apicem versus puberula, plerumque rosea vel rubescens,
rariter aurantiaca; tubus parte inferiore angusta, intusstuposa,
rariter puberula (T. ternatensis vSts.), parte superiore infundi-
buliformi recta vel curvata, intus semper glabra, 5-lobatus,
lobis magnis, saepe puberulis, inaequalibus. Stamina 4,
didynamia, quinto rudimentario; filamentis glabris vel basi
stuposis. Discus annularis. Ovarium 2-loculare, placentis in
utroque loculo binis; ovulis oo, multiseriatis. Stylus filiformis
alaber, stigma bilamellatum. Capsula 2-locularis, valvis coria-
ceis concavis.Semina applanata, tenuiter membranaceo-alata.

The principal differences with Pandorea are the following:
Tecomanthe.nbsp;Pandorea.

Corolla-tube mostly in-
side at one side hairy,
mostly glabrous near the
insertions of the stamens.

The groups I have distinguished within the genus were
originally based upon habitus-characters and I did not
know much about the phylogenetical relations, though I
pointed out in New Guinea (1 c. 292) that perhaps the
montane species may have found their origin on successful
trips into the mountains (by means of mutation or other-
wise, nobody knows).

Now I have studied nearly all the species of Tecomanthe
and I have found that 1 can support the above said
statement for some species. Most sections have appeared
to be probably quite right in phylogenetical respect.

The main point is the structure of the ovary, (Fig. 5)
which shows two tendencies.

I take the normal structure of the whole Pandorea-
group (p. 837) as the eldest. It is found in Pandorea
(Fig. 5r), Neosepicaea (Fig. 5 u), Hausmannia (Fig. 5 s)

lar, racemes, seldom thyrses
[T. mont., T. aurant.) Often
cauhflorous, but also on fo-
liate branches.

Leaves 3-foliolate or 1-
pinnate.

Calyx large (except T.
mont.), campanulate, 5-
lobed (except T. aurant.),
never cupular.

Corolla large (except T.
mont.) tubular infundibuli-
formous or somewhat en-
larged at the throat.

Corolla-tube inside al-
ways glabrous, except near
the insertions of the sta-
mens always stupose.

etc.. and as Bureau already mentioned perhaps the primi-
tive type (eldest) of the Bi^nonraceae-gynaecium. It is
found in T. gloriosa (Fig. 5 k). T. Gjellerupii (Fig. 51),
T. dendrophila (Fig. 5 e). T. acutifolia (Fig. 5 f).

The first tendency is that the ovary remains sub-cylin-
dric, but that the placenta\'s are quot;goingquot; close to the endo-
carp and are inserted not only at the septum but also
at the endocarp. This is the case with several links to
the totally angular position of the placenta\'s.
We can make the following series of stadia:
r. aurantiaca (Fig. 5g) gt; T. cijdopensis (Fig. 5 h) -gt;
T.ellipfica (Fig. 5i). The reduction of the number of ovules
on the external side of the placenta is going on. The pla-
centa\'s are already asymétrie. After these follows T.
ternatensis (Fig. 5 o). where the placenta has an angular
insertion and then T. amboinensis (Fig. 5p) and T. venusta
S. Moore (Fig. 5q) showing an entirely angular position.
This seems really striking to me.

The second tendency consists in an applanation of the ovary.
We can use again the primitive Panc/orea-groupstadium
as the first type, e.g. T. Gjellerupii (Fig. 5 1), then follows
T. saxosa Diels (Fig. 5 m) and at last T. nitida (Fig. 5 n).

It is a problem what is the principal means, with which
We can distinguish ancestors. In the large groups there
3re indeed fundamental characters, which may be used.
However, in small groups, e.g. relative small and rather
Uniform genera it is very difficult how to understand the
phylogenetical relations. In this case e.g. we have the
habitus and morphological characters of foliage and flowers,
which I used for a division in groups, whereas later on
found the striking relations based on the structure of
^he ovary mentioned above; the first method being a
niore or less external one, the latter being more of inter-
nal character. Moreover, we have to consider the geo-
graphical relations (Fig. 6).

Let us take for instance the § Volubiles. T. volubilis
I have not seen, but of T. arfaki (Fig. 5d) and T. nitida
(Fig. 5n) I gave a description and a figure. It appears
that these species do not show such close affinities in the
structure of the ovary. As to its habit, if one did not
know the colour of the flowers, one would think them
at first sight to belong to the same species, so striking a
resemblance they show. The geographical distribution,
however, is totally different. T, arfaki is found at alpine
altitude (2500 M.) on the „Bird\'s headquot; of New Guinea,
a highland, isolated from the central alpine area by a
peninsula of low altitude, some tops attaining at most
1200 m. but ca. 5—800 m. on the average.

In my opinion the structure of the ovary points out
that both species have another origin and are one of the
most striking examples of parallel evolution 1 know within
a small group of species, (see Fig. 16. p. 1041).

Before I studied the structure of the ovaries I had
already (in ms.) pointed out T. saxosa, which is found
at medium altitudes (14—1500 m.), to be the immediate
ancestor of T. nitida. The first is found in the Sepik-
district, the latter in central New Guinea, so that from
geographical point of view no objections can be made.

Indeed my suppositon has turned out to be most pro-
bable. Thus geographical distribution, habitus and morpho-
logical characters (leaves, flowers, ovary), all point in the
same direction.

T. arfaki I consider as a parallel form in habitus but
originated, with a link that I do not know now, from
other lowland species such as T. dendrophila, T, Gjelle-
rupii, or another allied species.

Further T. ternatensis, T, amboinensis and T. venusta
have appeared to be closely allied with respect to their
ovaries. But it is striking that they are also closely allied
in habitus (2-jugate, rather large leaves and similar flowers).

I suppose there is a close relation between them in phylo-
genetical respect; perhaps they have originated from
elliptica- or acufi/o/ia-liking ancestors.

T. cyclopensis seems to represent a side-branch of the
main-branch T. acutifolia T, elliptica T. ternatensis. It
is a remarkable fact that the structure and length of the
leaves of T. cyclopensis show resemblance with those of
T. saxosa. Both species are found at medium to sub-
alpine altitudes, the first at 1800 m., the latter at 14—
1500 m.
We have first to conclude:

A.nbsp;That the way in which Tecomanthe has produced
montane and alpine species, shows, as to their habit:

B.nbsp;Species which are similar as to their habitus, may
have been originated from different ancestors.

C.nbsp;Some groups I recognized as to the habit, have ap-
peared to present phylogenetical value; others, however,
are composed of elements, which are not immediately
related in phylogenetical respect.

I have tried to give a figure of the development of the
Tecomanthe\'s in a scheme and their relation to Pandorea
and some other genera (Fig. 16 p. 1041).

When more fruits will be gathered, they certainly will
establish the above mentioned hypothesis, which is based
principally on the structure of the ovary. It is a pity that
now we know but 2 valves of T. nitida.

1.nbsp;Calyx truncate or shortly dentate at most 0.8 cm. 1. 2
Calyx not shortly dentate nor truncate, but larger
and distinctly 5-lobed, far above 0.8 cm. 1., at least

2.nbsp;Robust lian, calyx truncate, leaves 3-foliolate, leaflets
large, 10-15 cm. 1., corolla large 8.5-10 cm. 1., au-
rantiacous and firmly membranous . § Aurantiacae

Small lian, calyx shortly 5-lobed, leaves 3-4-jugate,
leaflets small, 3-6 cm. 1., corolla small 2-2.5 cm. 1.,
membranous, brown-yellowish .... § Montanac

4.nbsp;Calyx rather small ca. 1.6 cm. 1., lobes triangular,
leaflets broad elliptic abruptly acuminate 3. T. elliptica.
Calyx larger 1.6-2 cm. 1., lobes ovate-acute or lan-
ceolate, leaflets elliptic-oblong or lanceolate, not
abruptly acuminate..............5

Leaflets elliptic-oblong, acuminate, calyx lobes ovate-
acute, calyx broader campanulate. 5. T. dendrophila.

6.nbsp;Leaves small 4-7.5 cm. 1., 3-5-jugate, leaflets very
small at most 2 cm. 1., coriaceous 2-3-nerved, small
climbers of high altitudes ......§ Volubiles 7

Leaves and leaflets much larger, 2-3-(seldom 5-: T.
gloriosa) jugate, with) much more pairs of sidenerves,
high climbers of medium altitudes. . . § Saxosac 9

7.nbsp;Corolla rather small ca. 6 cm. 1.. very broad (ca.
3.5-4 cm. in sicc.) deep rosa, inside white longitudi-
nally striate, calyx rather large 2-2.3 cm. 1. firmly

broad in sicc.) calyx smaller ca. 1.8-2 cm. 1., papy-
raceous ...................8

8.nbsp;Leaflets very small. 0.8 cm. 1., somewhat crowded
on a rather long ca. 2.7 cm. petiole (entire length of
leaf: 5 cm.), corolla pink, outside pilose at the apex

Leaflets 1.3-2 cm. 1., leaflets not crowded on a rather
short 1-2 cm. 1. petiole (entire lenght of leaf: 5-7 cm.),
corolla rosea, yellow to the base, inside yellow with
red lines..............8. T. nitida.

9.nbsp;Leaves 5-pinnate, very large ca. 7 cm. 1. petiolulate,
petiole and rhachis pubescent, leaflets 9-10 cm. 1.,
puberulous-lepidote, corolla rose purple, creamy at

Leaves 2-3-pinnate, mostly smaller and shorter pe-
tioled, glabrous...............

11.nbsp;Small Han 8-10 m. high, leaves 2-3-jugate, 1.5-2 cm.
1. petioled, rhachis narrow winged, leaflets ovate-
oblong, dentated in the upper half, rather short
4-6 cm. 1., calyx white, membranous, corolla pale

Large climber, leaves 2-jugate, 4-7 cm. 1. petiolulate,
rhachis sub-terete, leaflets entire, 7.5-12 cm. 1., calyx
papyraceous, blush-red, corolla coccineus, 8-10 cm. 1.

12.nbsp;Corolla-tube near the insertions of the stamens nearly
glabrous, margin of the calyx glabrous, at the top
indistinctly ciliate, petioles and rhachis with small
glands. Leaves 3-jugate. Corolla 6 cm. 1., whitish,
later on pink. Very large lian, 20-30 m. high

Corolla-tube near the insertions of the stamens stu-
pose. margins of the calyx puberulous to densely

13.nbsp;Corolla 5-7.5 cm. 1. tube pale, marked with purple
lines, limb rosy purplish. Leaves 2-pinnate, leaflets

• 2-7.5 cm. 1., ovate-lanceolate, thin . . \\3. T. Hillii.
Corolla mosdy longer, tube rosa-purpureus, limb
mosdy creamy. Leaves 2-3-pinnate.......14

14.nbsp;Leaflets small, 4.5-5 cm. 1., lanceolate, sub-coriaceous,
corolla rosea, tube ca. 6 cm. 1. . 14. T, cyclopensis.
Leaflets longer 7-10 cm. 1., papyraceous or char-
taceous ...................15

15.nbsp;Leaflets lanceolate-oblong, gray when dry, acute or
shghtly acuminate, entire, 7-10 cm. 1., chartaceous,
nerves ca. 7 pairs strongly prominent beneath, sunken
in above, calyx dark rosea, lobes creamy, lanceolate-
acuminate, corolla-tube ca. 6.5 cm., stalks rather rigid

Leaflets-oblong ovate, somewhat abrupdy acuminate,
dark brown when dry, dentated to the top, 7-8 cm.
1., papyraceous, sidenerves 6-7 rather sofdy prominent,
not sunken in above, corolla outside roseo-purpureus,
ca. 6 cm. 1., stalks rather thin . . . 16. Tquot;. venusta.

1. Tecomanthe aurantiaca Diels. — Engl. Jahrb. 57
(1922) 497; van Steenis. Nova Guinea 14 (1927) 296.
Fig. 3d; 5g; 6 (1); 16.
Firm high lian. stem ca. 5 cm. thick; bark light-gray,
branchlets ca. 0.5 cm. thick, sub-terete, cortulate. Leaves
opposite, large, 3-foliolate; petioles stout, minutely winged,
sub-terete, the bases connected by an indistinct line,
6--7 cm. 1.; petiolules of the lateral leaflets short 0.6—1.5 cm.,
that of the terminal one -2.5-3.5 cm. Leaflets oblong,
acuminate at the top, lateral ones distinctly oblique,
shining above, coriaceous, 10—15 X 4.5—6 cm. with
8—10 pairs of veins distinctly prominent beneath. Thyrses
short, peduncle ca. 8 cm. 1., axillary, erect, on a foliate

branch (always?); primary stalks 3-florous, ca. 1.5 cm. 1.,
pedicels ca. 1 cm. 1.. bracts oblong ca. 0.15 cm. 1. Calyx
short, truncate, 0.5-0.8 X 0.8—0.9 cm. Corolla aurantia-
cous, curved, tubiform-infundibuliformous, minutely pube-
rulous outside 8.5—10 cm. 1., 2.2—2.7 cm. wide, inside
near the insertions of the stamens densely stupose; lobes
triangular -ovate, the. margin densely puberulous, inside
slightly puberulous, 1.2—2.5 cm. 1.; sternens 4, didynamous,
with a fifth rudimentary one. ca. 1 cm. above the base
of the corolla inserted, filaments glabrous; style glabrous.
Fruit unknown.

Geogr. distr. New Guinea — N. E. New Guinea —
Sepik-district, Etappen-mountain in 25 m. high rather
mossy forest with many epiphytes. 850 m. alt.: Ledermann
n. 9561 (D).

This species is rather interesting, possessing a compa-
ratively short, truncate calyx, the inflorescence being a
thyrse, whilst the other species having a raceme. From
T. dendrophila it differs at first sight, having much larger
leaflets, an orange corolla and a truncate small calyx.

In the flower it remembers much P. leptophylla, though
it differs in other respects, e.g. in the foliage, totally from
it. At any rate it is a Tecomanthe showing affinities with
Pandorea, just as T. montana. For the rest the inflores-
cence is a thyrse. a fact also pointing to Pandorea.

2. Tccomanthc montana Diels. — Engl. Jahrb. 57
(1922) 497; van Steenis Nova Guinea 14 (1927) 296.

Small lian, stem fingerthick, ca. terete, bark light gray.
Leaves 1-pinnate, 3-4-jugate. opposite, 10—13 cm. 1.;
petiole 2.5—3 cm. 1., sulcate above; rhachis especially to
the apex 0.1 cm. broad leafy winged, all stalks articulated,
thickened, leaflets papyraceous, nearly sessile, oblong, grosse
serrate in the upper part of the margin or with 1—2

teeth 3-6 X 1.5-2.5 cm.; nerves 5-6. curved, prominent
beneath Thyrses axillary on young shoots, genuine, ca.
4 cm. 1.. pauciflorous. pedicels ca. 1 cm. 1. Calyx mem-
branous. campanulate 0.7 x 0.6 cm. (inc lob.) with 5
equal teeth ca. 0.15 cm. 1.. triangular. CoroHa membranous,
obhque-campanulate. glabrous, except near the insertions
of le stamens stupose. 2-2.5 X 0.8--0 9 cm., brown
vellowish. brown spotted at the throat, lobes sub-elhptical.
rotundate. puberulous. dliate 0.5-0.6 X 0.4-0.5 cm., white.
Anthers 4. inserted just above the 0.6 cm. 1. narrow part,
of the tube, 1 cm. l. cells divaricate, elliptical, 0.15 cm.

Geogr. distr. New Guinea - N. E. New Guinea -
Sepik-district, Lordberg, 1000 m. alt. in mountain-forest:
Ledermann n. 9916 (D); Hunsteinspitze, alt. 1350 m.:
Ledermann n. 10925 (D); Hunstein Mountains, alt.
1350 m.: Led er mann n. 8518 (D); Hunsteinspitze,
1350 m. alt.: Ledermann n. 10962 (D).

This species seems to be endemic in the forests at
medium altitudes in the Sepik-district. It is a small lian
6—8 m. high or more, very well characterised by its
genuine thyrses, small flowers and 3-4-pinnate leaves.

T. montana forms with T. aurantiaca and P. leptophylla
the links between Tecomanthe and Pandorea. The rather
small flowers and the inflorescence, being a thyrse. point
to Pandorea.

Lian. ca. 5 m. high, branchlets ca. 0.5-0.7 cm. thick,
internodes ca. 9-11 cm. I., bark gray, striate (in sicc.)
with few lenticels; leaves 3-foliolate. 9-14 cm. 1.. petiole
2.5-3.5 cm. 1. the bases connected by a puberulous pro-
minent line, lateral leaflets 0.6-1.5 cm. terminal 1.3-2.8 cm.

Fig. 5. Tccomanthc H. Bn. a. Inflorcsccncc and roots on a naked
branch of T. clllptlca vSts. (Type speclmcnK b. Leaf of T. venusta S.
Moore. (Gjellerup n. 54). c. Leaf of T. saxosa Diels. (Type specimen),
d. Leaf of T. arfaki vSts. (Type specimen). Transverse sections of the
ovary, e. T. dendrophila (Bl.) K. Sch.. (Drs. van Leeuwen n. 11263).
f. T. acutifolia vSts. (Type specimen), g. T. auratianca Diels. (Type
specimen), h. T. cyclopensis vSts. (Type specimen). l.T. elliptica vSts.
(Type specimen), j. T. arfaki vSts. (Type specimen), k. T. gloriosa
S. Moore (Forbes n. 622). 1. T. Gjellerupii vSts. (Type specimen), m.
T. saxosa Diels. (Type specimen), n. T. nitida vSts. (Type specimen),
o. T. ternatensis vSts. (Type specimen), p. T. amboinensis (Bl.) vSts.
(Kornasi n. 1372). q. T. venusta S. Moore. (Gjellerup n. 54) Pandorea.
r. P. stenantha Diels. (Type specimen) Hausmannia s. H. jucunda F.
V. Muell. (L sub n. 898. 199.. 70). t. longitudinal section H. jucunda
P. v. Muell. (L. sub n. 898. 199.. 70) Neosepicaea. u. N. viticoidcs
Diels (Type specimen). Magnif. ^/j X, except e-u.

petiolulate, elliptic-obovate, rounded and oblique at the
base, broad rotundate at the apex, suddenly short acu-
minate contracted at the top, with few teeth, papyraceous,
acumen short 0.2-0.5 cm., leaflets concave; nerves 8-9,
curved at the top, reticulate, prominent beneath, sunken
in above (in sicc.). Racemes on the old wood as well as
on the younger shoots, short, pendent, few-flowered (3-5)
on the young shoots, larger and many-flowered on the old
wood (10-15), rather dense: peduncles 0.5-3 cm. 1. rather
thick, pedicels rather thin 0.8-1.2 cm., bracts setaceus.
Calyx campanulate, smaller than of T. dendrophila, tube
1.1-1.2 cm., lobes triangular, puberulous at the margin,
0.5 X 0.5 cm. acute. Corolla rosa, tubular-infundibulifor-
mous, curved, inside 1 cm. above the base near the in-
sertions of the stamens stupose, lobes sub-inequal, triangular,
1.5 X 1.3 cm., at the margin and the top puberulous.
Stamens 5, didynamous, with a fifth rudiment, 6.5 and
7 cm. 1., style 7-7.5 cm., 1.. Fruit unknown.

Geogr. distr. New Guinea ^ Mouth of the Mamberamo:
R. F. Janowski n. 427 (L, B, D, Kew): Pamoi near the
Mamberamo river, riverbank: Moszkowski n. 102 (D).

This species is nearest allied with T. dendrophila and
T. acutifolia. It is characterised by its peculiar leaflets, the
distinct concave leaflets and smaller calyx than that of T. den-
drophila, whilst the lateral leaflets are longer petiolulate. Of
the latter species Blume mentions that there are sometimes
roots on the stem. In the first cited specimen of T. elliptica
roots are also present, namely on a 0.5 cm. thick not-
foliate branchlet at both sides of each leafscar; they are
30 cm. 1. and not branched, as far as I could see on the
dried materials. (See under T. dendrophila p. 883). Certainly
the lian does not climb with this straight pendent roots,
as is the case with other allied genera.

peculiarly placed. In the axil of a leafscar there is a short-
shoot, a stalk, ca. 2-2.5 cm. 1. bearing 2 opposite 3-foholate
leaves, whilst the top is represented by a scar. Perhaps
the voluble apex of this shoot is broken off during collecting.
At first sight there seems to be present 6-foliolate leaves.

The specimen of Moszkowski, which Diels referred
to T. dendrophila appears to belong to this allied species.

Moszkowski n. 272 differs from the T. dendrophila
and points to T. elliptica as to the form of leaflets; however
it has another calyx and much shorter petiolulate leaflets,

Lian, stem sub-quadrangular; branchlets 0.3 cm. thick
glabrous, strigulose-striate slender, internodes 10-13 cm. 1.
Leaves 3-foliolate, opposite, papyraceous, 12-18 cm. 1.,
petiole 2.5-5 cm. 1., glabrous, sulcate above; lateral leaflets
0.5-0.7 cm., terminal 1.8-2.6 cm. 1. petiolulate, lanceolate,
acute or short sub-acuminate 7-9 X 2.3-3 cm., margin revolute
(in sicc.), entire or minutely 1.-toothed on each side of the
apex, sidenerves 6-7, prominent below, slightly prominent
above. Racemes short, axillary, on the foliate young shoots
(also the wood?), peduncle very short 0.4 cm. 1. pendent,
ca. 5-flowered. Calyx brown-purple 1.6-2 cm. 1. (incl. lob.)
tubular-campanulate, equally 5-lobed, lobes- narrow lan-
ceolate-acuminate 0.6-0.8 cm. 1., margin puberulous. Corolla
rosa, tubular-infundibuliformous, tube in the lower part near
the insertions of the stamens stupose. 8-9 x 2-3 cm. (excl.
lob.), lobes creamy, at the apex and the margin puberulous,
inequal, triangular, acute 1.2 X 1.5 cm.. Stamens filiform,
glabrous 6.5 and 7 cm. 1.; style ca. 8 cm. 1.. Fnzif unknown.

Geogr. distr. Dutch S. W. New Gurnea—Lorentzriver,
near Alkmaar-east: G. M. Versteeg n. 1538 (L, B);
V. Römer n. 951 (L).

This species is closely allied with T. elliptica and T.
dendrophila, but with lanceolate leaves, large flowers with
pale yellow lobes, a brown-purple calyx, sub-bilabiate
flowers, the calyx-tube being narrower, the lobes narrow
lanceolate-acuminate.

— = Tecoma {Campsis) Dendrophila BLRumphia A {184S)
35, t. 190; K. Schumann Fl. Deutsch Ost-Asiat. Schutzgeb.
Engl. Bot. Jahrb. 9 (1887)218; — = Dendrophila trifoliata
BL in tab. cit.; — = Bignonia rhodosantha Zippel. (Herb,
nomen): Blume. Mus. Bot. Lugd. Bat. 1 (1849-51) 25;
Miquel. Fl. Ned. Ind. 2 (1856) 757; Miquel. Ann. Mus.
Lugd. Bat. 1 (1864) 197; F. v. Mueller. Descr. Notes
Papuan PI. 9. 64; — = Campsis Dendrophila Seemann.
Journ. Bot. 5 (1867) 373; SchefFer Ann. Jard. Bot. Buitenz.
1 (1876) 40; K. Schumann FL Kais. Wilh. Land. (1889)
123; Warburg Pfl. Papuas. Engl. Bot. Jahrb. 13(1890)
418; — = ? Tecomanthe Bureaui Baill. Hist. PI. 10 (1891)
41 ; — = Pandorea dendrophila [Bl.) Boerl. Handl. Fl. Ned.
Ind. 2 (1899) 600; K.Schumann. Schutzgeb. Südsee. (1900)
539; L. Diels Bign. Papuas. Engl. Bot. Jahrb. 57 (1922)
496; van Steenis Nova Guinea 14 (1927) 297.

High climber, reaching the uppermost foliage of the
trees; stem . ca. 0.5—1 cm. thick, terete, often spirally
twisted, bark gray or brownish gray (in sicc.), glabrous,
rough with orbicular lenticels; branchlets terete, sub-
sulcate or striate, with much smaller lenticels; inter-
nodes 7—15 cm. 1. Leaves 3-foliolate, remote, opposite,
glabrous, 8—15 cm. 1., 2.5—3.5 cm. 1, petioled. Leaflets
membranaceous to firmly membranaceous, ovate to elliptic-
oblong, acute or shortly acuminate, entire or mostly
indistinct- to grosse-serrate to the top; petioles flat above,
minutely winged, the bases connected by a glabrous
prominent line; lateral leaflets nearly sessile. 0.2—0.3

cm. 1. petiolulate, terminal 1.5—2 cm. 1. petiolulate; leaflets
6—10 X 3-5 cm., sub-acute at the base, shghtly oblique;
sidenerves not very prominent beneath, in 6—7 pairs.
Racemes on the old wood (cauliflorous) as well as on the
younger foliate shoots, single, axillary, patent or bent back;
peduncles short, 1—2 cm. (or sometimes longer ca. 7 cm.),
3^10-flowered (on the old wood sometimes 10—13 fl.),
with some opposite minute lanceolate scales on the lower
part; pedicels short, ca. 0.75—1 cm., slender, pendent.
Calyx campanulate, with a small setaceus bract and 2
bracteoles, reddish-brown, ca. 2 cm. 1. (inch lob.), equally
5-lobed, glabrous or minutely puberulous, membranous
to firmly membranaceous, lobes lanceolate ca. 1 cm. 1.,
sometimes minutely mucronate or little acuminate, short
whitish sericeo-tomentose at the margin. Corolla infundi-
buliformous, outside phoeniceus, inside lactescent, glabrous
except 1 cm. above the base near the insertions of the

patent, lobes ovate somewhat obtuse. Stamens 4 included,
didynamous, with a fifth setaceus sterile one; filaments
lilac, filiformous, hairy at the base; anthers white, inserted
at the top, glabrous, pendent, sub-divergent at the base;
style somewhat longer than the stamens, glabrous; disk
fleshy, annuliformous; ovary elongate. Fruit unknown.

(Type specimens of Blume leg. Zippel
in the Leyden Herbarium).
Geogr. distr. New Guinea — Dutch New Guinea —
In forests near the coast: Zippel (L. sub n. 898, 200...
94, n. 898,200... 95, n. 898,200... 96. U sub n. 010794);
pr.Anday: Teysmann (B); Triton Bay: M. Hombron
n. 1814 (P), type specimen oi Tec. Burcaui Baill. \\ Lorentz
river, Geluks-hill near Alkmaar: v. Römer n. 467 (L);
Lorentz-river, on a hill near Nepenthes-hill: Versteeg
n. 1376 (LB); Mamberamo-district, east-inclination. Door-

(L, B, D); Taua, mountainforest: M. Moszkowski n.
272 (D); Naumoni, mountainforest, 75—300 m. alt.:
Moszkowski n. 386 (D); primeval forest pr. Albatros
Bivak, alt. 60 m., cauliflorous: W. M. Docters van
Leeuwen n. 11283 (B).

Kaiser Wilhelms Land — pr. Wengi in forest:
Schlechter n. 11610 (D); Neu-Mecklenburg: ? : n. ?
(D); Sepik-district, K. Augusta river: Schultze n. 178(D);
Kollua pr. Finschhafen, cauliflorous: Hollrung n. 166
(D); lower course of the Gogol river in high forest;
Lauter bach n. 1164 (D); K. Wilh. Land pr. Kelana,
high climber: Hell wig n. 53 (D); mouth of the Bumi
river near Finschhafen: Lauterbach n. 434 (D); Huon
Baypr. Samoahafen: Laut er bach n. 722 (D); K. Wilh.
Land, high primary forest, 150—500 m. alt. common and
spreaded: Laut er bach n. 2816 (D); near Kalueng (K.
Hollrung); Sattel Mountains (Hollrung); Konstantinhafen
(Hollrung); Bismarck mountains, high climber found in a
quot;Gallerie-waldquot;, 500 m. alt.: C. Lauterbach n. 2780
(D); Astralobe Gebirge: F. H. Brown n. 203 (D).

British New Guinea. — In woods on the coast of
New Naima (Seem.); Owen Stanley Range: Forbes (F.
v. Mueller).

T. dendrophila is an endemic lian of the forests at low
and medium altitudes up to ca. 500 m. in Papuasia. Though
many materials are at hand, it is a pity there are no
fruits collected as far as I know. It seems to be the com-
monest Tecomanthe, local as well as in general. So e.g.
Warburg notes: quot;die grossen rosafarbigen Blüten liegen
häufig massenhaft auf dem Waldboden dicht bei Finsch-
hafen.quot;

It is a high climber reaching the uppermost foliage in
high primeval forets. The beautiful flowers are found as
well on the old wood near the leafscars (axillary) as well

As to the remark of Scheffer: quot;folia passim bijugaquot;,
etc. I suppose this must be a mistake. In his material
collected by Teysmann I did not see this; it is a true
3-foliolate species. So I cannot agree with the determina-
tion of the specimen Weinland n. 174 and Wiesen-
thal n. 10, which is a true 5-foliolate species (See p. 898).

The variability is not as great as I supposed to bc
the case, when I had seen but few materials. The flowers
are uniformly described by the collectors, but the foliage
shows differences. Diels supposes a heterophylly, some
specimens having entire leaflets, others possessing a gross-
dentated margin especially near the top. It seems to me
that this is caused by light and shadow, influencing the
younger shoots, a case comparable to that in Lonicera
Periclymenum and Symphoricarpus, both showing the
same structure in this irregular teeth. To make out the
question one ought to have much larger specimens than
are collected hitherto.

As to T. Bureaui I saw the imperfect specimen on
which Bâillon founded the species and genus Tecomanthe.
Indeed it is, as Diels pointed out (1922) perhaps the
same as T. dendrophila. Only a branchlet with a naked
peduncle and some flowers are present. Whereas T. ellip-
tica and T. acutifolia are distinguished in the first place
by their foliage I am not absolutely sure in identifying
T. Bureaui with T. dendrophila, but it seems to be most
probable.

As to the quot;roots Blume noticed in his table, I am not
sure there are true roots as to their position, but on the
other hand I cannot believe Blume represented some
Hepaticae, which are indeed often found on the stem.
Of T. elliptica I saw true roots, but these are placed
truly on both sides of the leafscars, just as van der

Small ca. 2 m. high voluble climber, stem terete, spi-
rally twisted, with few lenticels, bark brownish, stem ca.
0.3—0.4 cm. thick. Leaves opposite, 1-pinnate, 4—5-jugate,
5—6.5 cm. 1., [petiole ca. 1 —1.5 cm. 1.. internodes of the
narrow winged rhachis 0.5—1 cm. 1. Leaflets glabrous,
coriaceous, shining, dark-green, elliptic. 1.3—1.8 X 0.7—
0.9 cm., shghtly oblique, cuneate at the base, nearly ses-
sile, terminal leaflet 0.3 cm. 1. petiolulate, acute at the
dentate top, top somewhat recurved, nerves obscurely
sunken in above, prominent beneath, 1—3 pairs. Racemes
axillar 1.5—2 cm. 1. peduncled or longer and then with
1 or 2 leafpairs, though reduced, pauci (4—5)-florous;
bracts ca. 0.8 cm. 1., pedicels 0.8—1.2 cm. 1., with 2 subu-
late bracteoles. Calyx firmly papyraceous 2—2.3 cm. 1.
(incl. lob.), campanulate, sub-inequal 5-lobed, lobes ovate,
obtuse, shortly mucronate, 1 x 0.65 cm., margin puberulous.
Coro//a campanulate-cyathiformous, tube 5.5—6 cm. 1. and
3.5—4 cm. broad at the throat, glabrous except inside near
the base stupose near the insertions of the stamens, deep-
rosa, inside white longitudinally striate; lobes 5, triangular-
orbicular 1.3—1.5 cm. diam., margin and apex of the lobes
pubescent. Stamens 4, didynamous, filaments 3.3 and
4 cm. 1., stupose at the base; style 5.5 cm. 1. Fruit unknown.

Geogr. distr. New Guinea — Arfak Mountains, be-
tween the Angi-Seas. alt. 2500m.: K. Gjellerup n. 1212
(L, B, D, Kew).

1) H. A. A. van der Lek. Over de wortelvorming van houtige stekken.
Diss. Utrecht 1926.

This small climber occurs as well as T. nitida on open
places in a heath-shrubbery-vegetation on a mountain-ridge
between the Angi-Seas. It has the same habit as T. nitida
and T. volubilis. but differs in several points. (See what is
said about this at the end of the descriptions of these
allied species).

Phytogeogr. and Fl. Arfak Mts. (1917) 179; Diels. Engl.
Jahrb. 57 (1922) 498; van Steenis. Nova Guinea. 14
(1927) 299.

Fig. 6. Geographical distribution of Tecomanthe H.Bn. 1. T. auran-
tiaca Diels. 2. T. montana Diels. 3. T. cUiptica vSts 4 T acutifoltó
vSts. 5. T. dendrophilla (Bl.) K. Sch. 6. T. arfaki vSts 7. T. volubi is
Gibbs. 8. T. nitida vSts. 9. T. gloriosa S. Moore. 10. T. saxosa D^cls.
11. T. amboinensis (Bl.) vSts. 12. T. ternatensis vSts. 3. T. Hillii
(F. v. M.) vSts. 14. T. cyclopensis vSts. 15. T. Gjellerupii vSts.

gray-strigulose, covered with prominent lenticels. Leaves
small 5 cm. 1. (incl. pet.) 2.7 cm. 1. petioled, sub-4-angled,
opposite, 1-pinnate, 4-jugate. rhachis contracted at the
insertions of the leaflets. Leaflets 0.8 cm. 1., the lower pair
shortly petiolate and often smaller, elliptic, obtuse or
acute, coriaceous, lateral veines 2—3, impressed above,
surface beneath glandularly-punctate, longitudinally striate
when dry and transversely so above. Racemes ca. 2 cm. 1,
axillar, short (1 cm. 1.) peduncled, peduncle with a pair
of subulate bracts and 2 or 3 pairs of reduced leaves or
folioles; pedicels ca. 1 cm. 1. with 2 opposite bracteoles
0.4 cm. 1. Calyx ca. 2 x 0,8 cm., sub-equally 5-lobed,
lobes ca. 1.3 X 0,6 cm., reflexed in flower. Corolla large
ca. 8.5 X 4 cm., broadly infundibuliformous, pink, outside
pilose at the apex of the tube, inside hairy at the base;
lobes 5 unequal deltoideus-acute, tomentose at the margin,
1.3—0.6 cm. Stamens 3 and 4 cm, 1. anthers 0.5 cm. 1.
with divergent cells, filaments inserted ca. 1 cm. above
the base of the tube, stupose at their base. Style ca. 6 cm. 1.;
stigmatic lamella oblong, glabrous. Ovary cylindric, glabrous,
bilocular, ovules in many rows, disk fleshy crenulate. Fruit
unknown.

Geogr. distr. New Guinea — Arfak Mountains, Koebre
ridge, twining in shrubberies, alt. 3000 m.: Gibbs.
n. 5603.

The species is nearest allied with T. nitida and T. arfaki.
From both it differs in the smaller leaflets and the longer
petioled leaves, the leaflets of which are somewhat crowded.
It has much larger flowers and absolutely otherwise (pink)
coloured flowers than T. arfaki and T. nitida. It is not
allied with T. leptophylla as Gibbs remarks, but this
species was known to Gibbs only by its description. The
pluri-jugate small, however gross-dentate leaves of the
latter remember in the description somewhat those of

It was not only found in the heath-shrubbery vegetation
of the Koebre ridge, but also in the forest below near
the lake. In the latter place it was plentiful and the fallen
corollas covered often conspicuously the ground, but on
the Koebre Mt. it was still in flower, both on the old
and young green wood. Certainly this variabihty is caused
by local climatic conditions.

This species shows again, that the cauliflory and the
character of terminal or axillar inflorescences does not
indicate an important taxonomical difference, these 4 cha-
racters being found on one and the same plant.

The habit of the small climber with its reduced leaf-
surface and the coriaceous structure of the leaves points
to a montane xeromorphous adaptation.

Small voluble climber, stem thin, sub-quadrangular,
bark with many Ienticels ca. 0.3 cm. thick, spirally twisted.
Leaves 1-pinnate, opposite. (3)-4(5)-jugate. S—7 cm. 1.,
internodes of the rhachis 1.5—2 cm. 1.. narrow-winged,
petiole 1.5—2 cm. 1. Folioles small, elliptic, cuneate at
the base, involute at the short acute top, shining, coria-
ceous, nearly sessile, terminal 0.2—0.4 cm. 1. petiolulate,
main nerve prominent, sidenerves 2—4 pairs, slightly
prominent, 1.5—2 X 0.65—1 cm. Racemes axillar. pauci
(3—5) florous, peduncle 1.7—2 cm. 1.. at the immediate
base with some minute opposite scale-Hke bracts and mosdy
a pair of reduced leaves; pedicels opposite ca. 1 — 1.4
cm. 1.; bracts lanceolate 0.5 cm. 1.; bracteoles 2 on each
pedicel, linear. Calyx coriaceous. 2.5—3 cm. 1., badious,
glabrous, sub-infundibuliformous or campanulate, tube
1.8—2 X 2 cm.; lobes triangular, margin puberulous,

1 X 0.8 — 0.9 cm. Corolla glabrous, infundibuliformous, basal
part of the tube narrow, enlarged above, ventricose, tube
8 X 3—3.5 cm., outside rosa, yellowish to the base, inside
yellow with red lines, tube to the base, especially near
the insertions of the stamens, yellowish-stupose ; lobes 5,
large, inequal, sub-orbicular 2.5 X 3 cm.\', at the margin
and towards the top yellow-pubescent. Stamens 4, creamy,
yellow-floccose at the immediate base of the 5.5 and 6
cm. 1. filaments. Ovary bilocular, with 2 placenta\'s in
each cell; ovules many-seriate. Capsule ca. 14 x 3 cm.,
valves concave, coriaceous, glabrous, shining inside. Seeds
unknown.

Geogr. distr. New Guinea — 5. W. Guinea in Mt.
Wichmann on open grounds, alt. 3100 m.:G. Versteeg
n. 2401 (L, B, D, Kew, U).

This species belonging to the section § Volubilis, is
found at high altitudes. It is allied with T. volubilis but
differs in the larger leaves, shorter petioles and the colour
of the flowers and also with T. arfaki, but possesses
rather dense lenticellate branchlets, much shorter flowers,
a differing calyx and nervature of the leaflets.

In habit the 3 species remember each other very much.
As well as the 2 others, T. nitida is supposed to be
originated from Tecomanthe s of lower altitudes and is
fully adapted to the open montane vegetation, the probable
ancestors being lians of primeval forests of Papuasia.

It is the single Tecomanthe of which a fruit, that is to
say without seeds or septum is known. It is the proof that
the genus belongs certainly to the Tecomeae and that it
is closely allied with Pandorea.

9. Tecomanthe gloriosa S. Moore — Journ. Bot. 61
(1923) 38; van Steenis. Nova Guinea 14 (1927) 299.
Fig. 5 k; 6 (9); 16.

Leaves opposite, 1-pinnate, 5-jugate, ca. 7 cm. 1. petioled;
petiole and rhachis pubescent, ca. 0.2 cm. diam., inferior
leaflets slightly smaller than the others 9—10 X 3—cm.,
the superior 11 — 12 X 4.2—4.5 cm., oblong ovate ca.

1nbsp;cm. 1. acuminate, top acute, base obtuse, entire, papy-
raceous, at both sides with hairy nerves, the lamina
puberulous-lepidote, ca. 9 pairs of nerves; petiolules ca.
0.5 cm. 1. pubescent. Racemes on the foliate branches
multi-flowered, peduncle with scale-like bracts below, ca.

2nbsp;cm. 1.; bracts setaceus, ca. 0.4 cm. 1.; pedicels slightly
swollen at the top, slightly puberulous 0.5-1.2 cni. 1.,
Calyx campanulate 2.3 X 1.1 cm., the 5 equal triangular
acuminate ca. 1.2 cm. 1. lobes included, at the margin
puberulous. Corolla tubular-infundibuliformous, rosa purple,
creamy at the top just as the lobes; tube ca. 7 cm. 1.
curved at the narrow basal part, suddenly dilated; lobes
puberulous, triangular, 1.2-1.5 cm. 1., inequal, the posterior
ones being shorter.

Geogr. distr. British New Guinea — Mt. Koikoko,
1000 m. alt.: H. O. F o r b e s n. 495, H. O. Forbes n. 622 (L).

Of this species I saw original materials except the leaves,
which must be 5-pinnate and very large as it is said. As
to the flowers, these remember those of T. dendrophila,
but with respect to the foliage it differs entirely from
that species.

10. Tccomanthc saxosa Diels — Engl. Jahrb. 57
(1922) 498; van Steenis. Nova Guinea 14 (1927) 297.

Lian, glabrous throughout, stem ca. 2.5 cm. thick,
8-10 m. high, voluble, bark grayish, corky. Leases opposite,
2-3-jugate, l-pinnate, 7-13 cm. 1., the terminal leaflet the
greatest, 1.5-2.5 cm. 1. petioled, narrow ca. 0.1 cm. br.
winged just as the rhachis, papyraceous, ovate-oblong
crenate-serrate in the upper part, acuminate, 4-6 X 1.5-2.

8 cm., nerves ca. 9, straight, prominent beneath, sHghtly
sunken in above, along the margin connected with each other.
Thyrses on the old wood, long (10-12 cm.) peduncled;
peduncle with many small scales at the base, further smooth
and glabrous, terete, genuine, 0.2 cm. thick. Calyx large.
1.5-2 X 1.5-1.7 cm. (excl. lob.), membranous, glabrous,
white, lobes 5, triangular 1.5 X 1.8-2.3 cm., cuspidate.
Corolla pale rosa, like that of T. dendrophila but smaller,
5-7 X 1.8-2.3 cm., tubular-infundibuliformous, lobes broad
triangular, acute, 0.8-1 x 1.2-1.4 cm. inside and at the
margin shortly pubescent, tube ca. 1 cm. above the base
dense long stupose near the insertions of the stamens.
Stamens 4, fertile with a fifth rudimentary one resp. 4 and
4.5 cm. 1., anthers divaricate, pendulous, oblong, ca. 0.5
cm. 1., filaments glabrous, style ca. 4.8 cm. 1., with 2
spathulate flat stigma\'s.

Geogr. distr. N. New Guinea — Sepik district, in a
shrubbery-forest, with few large trees, 14-1500 m. alt.:
Ledermann n. 12904 (D).

This lian does not grow in a high-stemmed primeval
forest, but in more bushy woods. In habit it reminds
much of the species belonging to the § Volubiles: alate rhachis
of the leaves, structure and nervature of the leaflets, large
flowers. As to the habit and habitat it forms in many
respects a link between the § Dendrophilae and the
§ Volubiles.

11. Tccomanthe amboinensis (Bl.) vSts. nom. nov. —
= Tecoma (Campsis) amboinensis Blume Rumphia
4(1848)35; van Steenis Nova Guinea 14(1927)298; — =
Campana rubra Rumphius. Herb. Amboin. 6(1755) Auctuar.
Cap. 52; Miquel. Fl. Ned. Ind. 2(1856)757;- = Bignonia
speciosa Reinw. (Herb, nomen); C. L. Blume. Mus. Bot.
Lugd. Bat. 1(1849-51)26; - = ? Pandorea spec. Hasskarl.
Neuer Schlüssel. Abh. Nat. f. Ges. Halle 9(1866) Heft 2.

143-389: — == Campsis amboinensis Seemann Journ. Bot.
5(1867)374; — = Pandorea amboinensis {Bl.) Boerlage
Handl. Fl. Ned.\'ind. 2(1899)600; — = Tecoma amboinensis
Reinwardt (Herb, nomen?) ex Boerlage; — = ? Pandorea
spec. E. D. Merrill. Interpret. Rumph. Herb. Philip. Bur.
Sci. (1917) 469.

Voluble lian, attaining the upper foliage of the trees,
branches 0.5-1 cm. thick, terete, bark gray, tuberculate
with orbicular ca. 0.075 cm. diam. lenticels, smooth, branch-
lets striate-sub-costate verrucose-punctate, glabrous. Leaves
1-pinnate, 5-foliolate, opposite, 15-25 cm. 1. glabrous or
the underside of the folioles minutely puberulous; petiole
4-7 cm. 1. subterete, canaliculate above, thickened towards
the base, the ultimate basal part and the narrow line
uniting the bases short pubescent, patent; internode ca.
4-5 cm. as the petiole; lateral leaflets short 0.5 cm. terminal
one 1-2.5 cm. 1. petiolulate, lateral leaflets smaller than the
terminal one; leaflets 7.5-12 x 3-6 cm., glabrous or the
underside slightly pubescent, elliptic, entire, smooth above,
sub-rotundate at the base, the top acuminate or even
cuspidate, papyraceous to thick membranous, sidenerves
8-9 at each side, curved, patent, prominent beneath. Racemes
erect, short 2.5-5 cm. 1. peduncled. single or 2 by 2. axillary,
8-12-flowered, patent, rhachis with small opposite scales;
bracts linear-acuminate, puberulous. deciduous, pediccls
0.8-1.2 cm. 1. with 2 minute bracteoles; ca/j/.vcampanulate.
sub-coriaceous, equally 5-lobed, glabrous, smooth, blush-red,
3-4 cm. 1. (incl. lob.), lobes ca. 2 cm. 1., lanceolate, aristate,
margin thickened to wards the top, scabrate; corolla
erect 8-10 cm. 1. (excl. lob.) broad infundibuliformous,
red-coccineus, paler so beneath, glabrous, narrow part of
the tube 2-3 cm., then enlarged, ca. I cm. above the base
densely stupose near the insertions of the stamens, lobes 5,
broad triangular-ovate, inequal, margin sub-pubescent.

Stamens 4, didynamous, glabrous, with a fifth setaceus
rudiment, included, anthers 4, divaricate, pendulous; ovary
cylindric, with a fleshy disk, style somewhat longer than
the filaments, glabrous ; stigma spathulate, 2-lobed, Fruit
unknown.

Geogr, distr. Malayan Archipelago—Ambon—\'Ambon:
Reinwardt n. 1413 (L. sub n. 898, 200... 57); idem
ex. Herb. Lugd. Bat. (U sub n. 010793); Ambon: leg.?
(L sub n. 898, 200... 55); Moluccas: Reinwardt (Lsub
n. 898, 200... 56): Ambon, valley of the Olifantsberg
(Rumphius); Ambon, on the G. Hariel: Boerlage n. 40(B).

Ceram-W. Ceram. 500—800 m. alt. pr. T. Lomoli
and Lohia Tala, along a stream: Rut ten n. 1571 (B);
M. Ceram, on the Mt. Isal: Kornasi n. 1372(B).

This beautiful lian seems to be truly common at medium
altitudes in the woods of Ambon; it is also found by
Rut ten in Ceram in the west and centre of the island.
Blume says about it; „una e stirpibus pulcherrimus, quibus
ornantur sylvae Amboinenses,quot; wich indeed must be the
case as to the very large scarlet flowers, the lian reaching
the uppermost foliage of the woods. C. L. DoleschalP)
however does not mention it.

Blume says that the leaves are rarely also 3-foliolate,
which I cannot affirm, his remark on this being surely
a mistake.

From T. dendrophila it differs at first sight in the
5-foholate leaves and the much larger calyx. The greatest
affinity is to be found with T. saxosa Diels. The latter

Korte karakterschets der flora van Amboina. Nat. Tijdschr. Ned.
Ind. 14 (1850) 139.

however, has smaller, gross-dentated more sub-coriaceous
leaflets, a smaller, more pale-red corolla and a thin mem-
branous calyx.

Frutex volubilis prealtus. glaber, ramosus, caule brachiali,
20-30 m. altus; rami lignosi 0.7 cm. crassi, teretes, cortice
rugoso, lenticellis paucis orbicularibus, ramulis ultimis tere-
tibus striatis. Folia opposita 1-pinnata, 3-jugata. 14-18 cm.
longa; petiolis 4 cm. longis. 0.2 cm. diam., semi-teretibus
supra sulcatis, striatis, basin versus glandulosis, rhachi
petiolo simili, interjugiis 3-4 cm. longis; foliola sub-sessilia
vel 0.5 cm. petiolulata. foliolo terminali 0.5-1 cm. longe
petiolulato, papyracea vel sub-coriacea. fragilia, oblonga,
basi inaequalia, 6-7 X 2.5-2.7 cm., integra vel apice dentis
paucis grossis praedita, apice sub-rotundata, acutata vel
breviter acuminata basi rotundata vel sub-cuneata; nervis
fere 4 angulo 45° divergentibus, vix prominentibus, reti-
culatione laxa indistincta. Racemi axillares e ligno orti,
solitarii vel geminata. 2-6 cm.; pedunculis bracteis multis
squamaeformibus aggregatis basi praeditis, teretibus, 0.2 cm.
diam.; pedicelli 1 cm. longi, bracteolis binis, bracteis lan-
ceolatis 0.3 cm. longis. Calyx in alabastro fere globosus,
breviter apiculatus, pallide viridis, postea campanulatus,
glaber, laevis, 1.5-2 X 1 cm. (lob. incl.) lobis ovato-trian-
gularibus, margine paullo sinuatis, apice ciliatis, 0.5 x 0.7 cm.
Corolla ca. 6 cm. longa (lob. incl.) tubo primo albido,
deinde roseo, 5.5 X 2.5-3 cm. parte inferiore 1 x 0.5 cm.,
parte superiore tubuloso^nfundibuliformi, glabro; lobis 5,
inaequalibus, brevibus, latae triangularibus, acutis, 0.5 X 1.5
cm., nonnisi marginibus puberulis. Stamina 4, didynamia,
quinto rudimentario, filamentis glabris, in apice partis
inferioris vix stuposae insertis, 4.5 et 3.5 cm. longis, tubum
haud superantibus. Discus annularis. Ovarium ovato-oblon-
gum, glabrum, sub-compressum, 2-loculare, placentis in

utroque loculo binis, valde distinctis. Ovulis oo, multi-
seriatis obtectus. Capsula ignota.

Geogr. distr. Malayan Archipelago — Ternate —
Branched Uan, 30 m, high, probably more than 100 m.
long, stem arm-thick, 600 m. alt. pr. Foramadiahi, flowering
specimen: Beguin n. 1201, type specimen (B) ; lian 20 m.
high, specimen in bud, pr. Foramadiahi, 1100 m. alt.:
Beguin n. 1414, co-type (B).

This very large climber of the Moluccas shows affi-
nities with T. amboinensis and several other species. It
is distinctly separated from all other species by its rather
small glabrous calyx, the glabrous 6 cm. 1. corolla, which
does not show the usual hairy ring near the insertions
of the stamens and the 3-jugate leaves.

From T. cyclopensis it differs distinctly in the foliage ; from
T. Gjellerupii in the glabrous pink corolla, the totally other
indistinct nervature of the leaflets and the shorter triangular
acute not ciliate calyx-lobes ; from T. saxosa in the foliage,
nervature, etc. and 3-jugate leaves; while T. venusta has
an entirely different foliage and a glabrous calyx and corolla.

13. Tccomanthe Hillii (F.v.M.) vSts. nom. nov. —
Tecoma Hillii F. v. M. Fragm. 10.101, 11.136; Bailey.
Queensl. Fl. 4 (1901) 1133 ; Bailey. Compreh. Cat. Queensl.
PI. (1909) 364.

Tall glabrous climber. Leaves 1-pinnate ; leaflets 5, ovate-
lanceolate, 2.5—7.5 cm. 1., thin. Racemes 6-flowered. Ca/yx
1.25—2.5 cm. 1., membranous. Corolla 5—7.5 cm. 1., limb
rosy-purplish, tube pale, marked with purplish lines, the
lobes 0.6—0.8 cm. 1., pubescent towards the margins.
Filaments filiform, anthers yellow, cells equal, 0.3—0.4
cm. 1., widely divergent. Staminodium 0.4—0.6 cm. 1. Style
filiform, glabrous, stigmatic lobes semi-lanceolate. Capsule
3.75 cm. 1. (Bailey I.e.).

There is but httle doubt this species represents the only
known Australian Tecomanthe. The few-flowered racemes,
the large calyx and corolla point to this genus. It is a
pity that I could not examine any materials.\' The species
seems to be very rare. Bailey mentions that the only plant
ever met with was found by Mr. Fred Turner in 1876,
when collecting near the Hervey Bay for Mr. W. Hill.

Lian. ca. 2 m. high, stem 0.6—0.7 cm. , thick, sub-qua-
drangular; bark grayish-yellow when dry, irregular, gla-
brous. Leaves 1-pinnate, 2-jugate, rigid; leaflets 5, lanceolate,
4,5^5 X 1.6—1.8 cm. nearly sessile, short cuneate and
slightly oblique at the base, short (0.5 cm.) acuminate at
the top with few (1—2) teeth at each side, shining dark-
green, firmly chartaceous, with 5—6 sub-curved pairs of
sidenerves, indistinctly prominent beneath, slightly sunken
in above, reticulations none; petiole 2.5—3 cm. 1., sub-
terete as well as the rhachis. Racemes on the old wood
and the branchlets, short 0.8—2 cm. 1., single or 2 by 2
in the axils, at the immediate base with scale-like short
opposite bracts, pedicels ca. 1.5 cm. 1. with setaceus 0.5
cm. 1. bracts. Calijx campanulate, 2.2 cm. 1. (incl. lob.),
glabrous, lobes sub-equal lanceolate-triangular, acuminate,
0.9—1.2 X 0.6—0.7 cm., puberulous at the margin. Corolla
rosea, tube ca. 6 cm. 1. broadly infundibuliformous, narro-
wed at the base, stupose near the insertions of the stamens,
sub-equally 5-lobcd, lobes short pubescent, obtuse triangular
1 — 1.3 X 1.6—2 cm. Stamens 4 and 5 cm. 1., filaments
glabrous, inserted 0.6 cm. above the base of the tube.
Style filiformous, glabrous, 6 cm. 1. Fruit unknown.

Geogr. distr. Dutch New Guinea — Mt. Cycloop, alt.
1800 m., 1 specimen found on mossy half-dead trees on
humus-soil: K. Gjellerup n. 544 (L, B).

This small lian of medium to high altitudes differs
distinctly from T. saxosa in the very short racemes, the
much smalle? firmer calyx, the corolla possessing a wide
mouth, the not narrow leafy alate petiole and rhachis and
the narrower leaflets. Moreover the calyx is much firmer
chartaceous and smaller.

It is allied with T. venusta and T. Gjellerupii as to the
flowers, but differs immediately in the foliage. The same
as to T. gloriosa.

Lian. ca. 3 m. high, stem ca. 0.4 cm. thick, sub-qua-
drangular, spirally twisted, internodes ca. 20 cm. 1.: bark
gray-brownish, striate, with many small Ienticels. Leaves
3-jugate, 17-20 cm. 1. opaque, darkgreen. opposite: petioles
3-4 cm. 1, slightly alate. sub-terete, sulcate above, the
bases connected with a strongly prominent hne. Leaflets
lanceolate-oblong, acute or shortly acuminate, entire light
grayish when dry, 7-10 X 3-4 cm.: nerves ca. 7-pairs
patent curved, prominent beneath, sunken in above, reti-
culations lax. prominent, petiolules of the lateral leaflets
0.1-0.5 cm., of the terminal ones 1.2-1.7 cm. Racemes
axillary, on foliate branchlets. peduncle puberulous short
ca. 2 cm. 1.. few (ca. 7) flowered, with many scale-like
short bracts at the immediate base: pedicels ca 1 cm 1
with setaceus 0.5 cm. 1. bracts. Ca/yx campanulate. equally
5-lobed. ca. 2.7 cm. 1. (incl. lob.), dark rosa. the lobes
cream-yellow ca. 1.5 cm. 1.. lanceolate-acuminate, margin
shortly pubescent. Corolla tubular-infundibuliformous rosa
sub-inequally 5-lobed, tube ca. 6.5 X 1.5-2 cm., glabrous,\'
except m the narrow lower part ca. 0.7 cm. above the

base stupose, lobes triangular, acute, 0.8—1.3 cm. Sfamens
included, didynamous, inserted ca. 1 cm. above the base of
the tube, 4.5 and 5 cm. 1., filaments glabrous; cells 0.5 cm. 1.,
divergent. Style filiformous, glabrous with 2 stigmatic
spathulate lobes. Ovary glabrous, bilocular, with many ovules
on 2 placentas in each cell in many rows. Fruit unknown.

Geogr, distr. Dutch New Guinea — North-coast near
Humboldt Bay, 300 m. alt. near Hollandia, spreaded:
K. Gjellerup n. 668 (L. B).

This slender lian is allied with T. venusta and less with
T. saxosa. As to the first it differs in the shorter corolla,
the larger calyx, the firmer papyraceous leaflets posessing
a totally different nervature. As to T. saxosa it has larger
entire leaflets, a much shorter few-flowered raceme and
different flowers, the calyx being darker rosa and more
firmly chartaceous.

16. Tecomanthe venusta S. Moore. Journ. Bot. 61
(1923) 38-39; van Steenis. Nova Guinea 14 (1927) 298.

Lian, branchlets 0.3-0.5 cm. thick, densely tubercular-
lenticellate. Leaves 1-pinnate, opposite, 5-foliolate, oblong-
ovate, 0.5-1 cm. 1., acuminate, entire, base obtuse, papy-
raceous, glabrous, 7-8 X 3.5-4 cm., sidenerves 7, reticu-
lations indistinct, 4 cm. 1. petioled, petiolules 0.5 cm. 1.
Racemes multi-flowered on naked branchlets, peduncle ca.
2 cm. 1. with scale-like bracts; pedicels ca. 1.2 cm. 1.
Cahjx ca. 1.5-1.7 X 0.8 cm. (lob. incl.), lobes 0.8 cm. 1.
Corolla inside creamy, outside rosa-pupureus 6 cm. 1.,
2 cm. broad near the limb.

Geogr. distr. British New Guinea — Mt. Gawada, 1700 m.
alt.: H. O. Forbes n. 857 — Upper course of theTami
river, ca. 85 m. alt.: K. Gjellerup n. 54 (L, B).

bilis; rami circ. 1.5 cm. crassi, spiraliter torti. Cortice
griseo sparse lenticellato. Folia ut in typo sed firmiora
apice integra. Racemi cauliflori singuli vel 2 superpositi in
axillus cicatricium foliorum robustiora. Calyx minor ca.
1.5 cm. longus\' pallide brunneus in sicco. Corolla minor
ca. 5 cm. longa angustior, apicem versus puberula.

Geogr. distr. New Guinea — Kaiser Wilhelms-Land .—
Alexishafen, climbing in trees, flowers like those of Digi-
talis purpurea (teste coll.): P. F. Wiesenthal n. 10(D);
on a hill west of Finschhafen: Weinland n. 174 (D).

The new sub-species is founded on a specimen Diels
(1922) arranged to T, dendrophila, but is certainly another
species as to its 2-jugate leaves; it belongs without doubt
to T venusta, but differing in several characters I believe
it represents a marked sub-species. Perhaps it is due to
the remark of Scheffer (see p. 883) that T. dendrophila
is not a true 3-foliolate species, but it is. I never met
with 3- and 5-foholate leaves on the same specimen.

Though I have not seen the original specimen of F o r b e s,
the species is well characterised and though differing in
altitude, the remarks of Gjellerup are just conformable to
Mr. Moore\'s description. There are 5 leaflets, which are
sometimes gross-dentated at the top. Moreover the lower
part of the corolla-tube is very narrow 1.5 cm. 1. and
stupose inside near the insertions of the stamens. The
species is allied with T. dendrophila.

The specimen of Gjellerup shows narrow winged
lines along the midribs of the calyx-lobes, the latter
being triangular acuminate, with short pubescent margin;
the stigmatic lobes are spathulate shortly mucronate.
Gjellerup notes it as common near the Upper Tami.

From T. saxosa it differs in the thinner, different leaves,
the much smaller, less loose calyx of firmer structure and
the larger corolla.

Glabrous lian. Leaves opposite, digitate. The inflores-
cence is a corymbiformous panicle, the ultimate stalks being
cymes. Inflorescence terminal, the lower side-axis in the
axils of the leaves. Calyx short, cupular-campanulate, sub-
coriaceous, 4—5-dentated, sub-bilabiate, glabrous, smooth.
Corolla with a narrow lower tubular part, infundibuli-
formous. inside the throat dense pilose. Stamens 4. didyna-
mous insert, with divaricate cells; disk cupular, covering
the ovary. Style filiform, with a bilamellate stigma. Ovary

bilocular, septum with 2 placenta\'s in each cell, each
placenta with oo ovules in several rows. Fruit unknown.

As Diels remarks the genus is allied with Pandorea,
with which I fully agree. It differs immediately in the
digitate leaves, the irregularly toothed calyx, the actino-
morphous corolla and the cupular disk.

I suppose it has been originated from Pandorea-like
ancestors with actinomorphic corolla\'s. The digitate leaves
may be derived from 1-pinnate leaves, wherein the rhachis
is reduced. Indeed several Pandorea\'s show unequal leaflets
on one leaf, perhaps a tendency of importance here. In
the American Tecoma s digitate leaves are common.

Though the fruit is unknown, I suppose the capsule will
turn out to be like that of Pandorea and Tecomanthe. It
is certainly a Tecomeae.nbsp;*

Large glabrous climber, stem arm-thick, bark gray, on
the young ca. terete, smooth branchlets elenticellate. Leaves
digitate, (3)-4-5-foliolate. 5—7 cm. 1. petioled. Leaflets
papyraceous, metallic shining when fresh, microscopically
punctate, entire, broad oblanceolate. acuminate, nearly
sessile, inequal. median 10—11 X 4—5 cm., laterals smaller.
Inflorescence paniculate, foliate at the base, 10—15 cm. I.,
primary axis ca. 4 cm. 1., ultimate stalks cymes. Calyx
0,4—0.5 X 0.35—0.4 cm. Corolla outside pubescent, brown,
inside yellow with dark brown-purple lines, lower narrow
part of the tube 3—4 cm. 1., at the top near the insertions
of the filaments annularly stupose; upper part of the tube
infundibuliformous, on one side longitudinally stupose

0.6 X 8 cm., lobes ovate-acute, ciliate, inside stupose, out-
side pubescent. Ovary bilocular, with 2 placenta\'s on the
septum in each cell, each placenta with oo ovules in several
rows. Fruit unknown.

Geogr. distr. New Guinea — Sepik-district, April river,
Standlager,\' ca. 100 m. alt., in well passable, 20—25 m.
high primeval forest : Ledermann n. 8665 (D) ; pr. Sogere ;
H. O. Forbes n. 437 (S. Moore).

This large lian seems to belong to the Pandorea-
Tecomanthe-„Sippequot;, I mentioned already (p. 837).

The colour of the corolla seems to differ a little. S.
Moore describes it as pale orange spotted with crimson.
As I did not see Forbes\' specimen, I do not know
whether this specimen represents a variety, or whether it
is due to the description. It is also possible that the colour
is varying indeed.

= Campsis F. v. M. Coll. Fragm. Phytogr. Austr. 4
(1864) 148; Bentham. Fl. Austr. 4 (1869) 539; = Nycto-
calos (sub-genus Hausmannia) jucundum Seeniann. Journ.
Bot. 8 (1870) 149; Benth. 6 Hook. Gen. PI. 2 (1876) 104;
F. V. Mueller. Sec. Syst. Cens. Austr. Pl. (1889) 167;
H. Bâillon. Hist. Pl. 10 (1891) 21. note 2; K. Schumann
in Engler Prantl. Pfl. Fam. IV. 3b (1895) 223; Bailey
Compreh. Catal. Queensl. Pl. (1909) 368, fig.345; Bailey
Bot. Bull. Contr. Queensl. Fl. 13 (1896) 1136.

It is due to lack of materials that Seemann, Bentham
and Hooker, Bâillon and K. Schumann did not
know whether the monotypic genus belonged to the
Bignonieae or to the Tecomeae. They all placed it in the
first tribus. Bailey described the fruit (Bot. Bull. 13(1896)
11) as opening loculicidally and thus it is stated that it
belonges to the Tecomeae. Its habit and flowering characters

show its alhances with Pandorea, Tecomanthe, Campsis
and Campsidium. It is very well distinguished however
by its induplicate-valvate corolla and the exsert stamens.

In the Paris herbarium I found a sterile New Guinean
specimen, collected by Hollrung (Augusta river. Kaiser
Wilh. Land, 1887), called byK. Schumann: Hausmannia
mollis K. Sch. n. sp.. There were only a few stalks and
some single leaflets. Perhaps the stalks were rhachises.
They were ca. 0.3 cm. thick, soft brown-pubescent, hairs
patent; internodes ca. 5 cm. 1.. On one side between
each pair of leafscars (leaflet-scars?) there was a peculiar
pezizaeformous gland ca. round 0.2 cm. diam., concave,
and placed on small prominent bubble. The leaflets are
nearly sessile or 0.3 cm. petiolate (petiolulate ?); the petiole,
underside of the leaves and nerves above being all brown
pubescent as well as the stalks. Leaf-form oblong rounded
at the base, somewhat oblique, entire, acuminate at the
top. The nerves 6—7 pairs slightly curved are patent
(ca. 45°), prominent and reticulately connected. Reticulations
lax. As I do not know whether Schumann has ever
published this and I did not see any flowers, I suffice with
a description. The leaves are somewhat like those of
Haplophragma, but such glands I saw never before. I do not
believe it to be a Bignoniaceae.

Tall glabrous woody lian, branchlets terete. Leaves
opposite shining at both sides, digitate 3-foliolatc, petiole
rather long 4—6 cm. 1. slightly striate. Leaflets convex,
membranous, entire, articulate at the end of the petiole,
oval or elliptical, narrowed into the short lateral 0.5—1 cm. 1.
petiolules and the 2.5 cm. 1. terminal petiolule, 9—11 X 4—4.5
cm,; the terminal leaflet occasionally confluent with one
of the lateral ones; sidenerves 8—10, curved, patent, reticu-

lately connected, prominent especially below, reticulations
lax. Inflorescences cymobotryoid (never racemes), terminal
or lateral in the axils of the leaves immediate under the
terminal one 5—15 cm. 1. rather few-flowered. Cymes

0.6—2nbsp;cm. 1., opposite, 2—5-flowered; bracts small, lan-
ceolate; pedicels 0.4—0.6 cm. 1.; bracteoles minute or none.
Calyx campanulate truncate 5—toothed, 0.4—0.6 cm. 1.
(incl. lob.) glabrous, rather thick membranous, lobes 5 sub-
equal or equal, broad triangular, 0.2 cm. 1., minutely
acuminate. Corolla purple, tubular, incurved, tube 2—2.5 cm.

1,nbsp;(excl. lob.) slightly dilated upwards above the calyx;
lobes 5 broad ovate induplicate-valvate in bud, ca. 0.5 cm. 1.,
hairy inside, obscurely arranged in 2 lips. Stamens 4, didy-
namous, fifth rudimentary, filaments hairy at their insertions
below the middle of the tube, exceeding the corolla ca.
0.7—1 cm., anther-cells diverging or divaricate. Disk hypo-
gynous, ca. 0.2 cm. 1. cupular. completely enclosing the
ovary when flowering. Ovary short, slightly compressed,
dissepiment transverse; ovules oo in several rows on each
placenta; style with 2 ovate stigmatic lobes. Capsule
locuhcidally opening in 2 concave valves as in Tecoma.,
7.5—11 X 3. 5 cm. acuminate at both ends, smooth; seeds
in several rows, overlaying each other, flat, pyriform,
wrinkled, ca. 1 cm. broad membranous winged. Dissepiment
broad, thick, and firmly attached to the valves or one
of them.

Geogr. distr. Queensland — Seaview Range. Rocking-
ham Bay: Da 11 a c h i, type specimen (P,L sub n. 898, 199...
71); Rockingham Bay: F. v. Mueller (P.L sub n. 898,
199... 70); Stony Creek. Cairns, shoot and loose flowers:
L. J. Nugent (Bailey); Thursday Island, large leaf and
very young shoot: E. Cowley .(Bailey); Freshwater
Creek, Cairns, the first time a fruit was gathered: L. J.
Nugent (Bailey).

The plant seems to be common in the Queensland
tropical shrubs, but fruiting specimens are seldom obtained.
Though all authors thought the genus a Bignoniea, the
celebrated Queensland botanist F. v. Mueller already-
suggested its alliance with several Tecomeae, such as
Tecoma and Campsis. and he was allright herein. Even
Bentham says: quot;I do not see the affinity with the genus
(or section of Tecoma) Campsis, suggested by F. v.
Muellerquot;. And Seemann calls it even a Nyctocalos
and considers the whole genus having an induplicate-vaU
vate corolla, which is absolutely wrong, though he thought
it firstly (Journ. Bot. 3 (1865) 93) allied with Campsis.
Certainly Seemann considered the leaves and the stamens
as important characters within the Bignoniaceae spoken
about, but I am supposing just the contrary.

Its affinities are to be found in Tecomanthe, Pandorea,
Campsis. Campsidium and Tecoma. It is however distin-
guished by its unique aestivation, certainly an important
character and further by its large cupular disk and exsert
stamens, this being of Httle importance.

Without doubt it belongs to the group: Tecoma, Pan-
dorea, Tecomanthe, Campsis and Campsidium, which is
characterized by its climbing habit, B/^noniaceae-flowers
following the usual scheme, loculicidally dehiscent fruits, a
single dissepiment with 2 placenta\'s on each side and flat
winged seeds.

Edinb. Philosoph. Soc. (1823) 263; Seemann. Journ. Bot.
1 (1863) 18,87; — = Tecomaria Bureau Monogr. (1864)
47, t. 13; Benth amp; Hook. Gen. PI. 2 (1875) 1044;
Bâillon Hist. PI. 10 (1891) 41; Engler-Prantl. Pfl. Fam!
IV. 3b (1895) 240; Sprague in Dyer Fl. Cap. 4 (1904) 448.

Tecomaria. Sprague, however, pointed out that this was
a confusion. Stenolobium has free anther-lobes and only
2 rows of ovules in each cell. Thus Tecomaria Bur. ~
Stenolobium; Tecomaria fulva Seem. = Stenolobium fill-
vum Sprague; Tecoma SmithiiBull. Cat. (1889) 8, a reputed
hybrid (T. velutina and T. capensis. Card. Chron. (1894)
64 = Stenolobium alatum Sprague {= Tecoma alata D.C.).

Tecomaria is a -South-African genus and T. capensis
has been introduced into America. Stenolobium is an
American genus. Evidently the genera are allied with
each other.

Stenolobium stans (L) D. Don in Edinb. Phil. Soc.
9 (1823) 88: - = Tecoma stans (L) Juss. in D. C. Prod.
9 (1845) 214: Bot. Magaz. t. 3191; Seemann Journ.
Bot. 1 (1863) 88; Koorders Exc. Fl. Java. 3 (1912) 184;
Merrill Enum. Philip. PI. 3 (1923) 444:

A many-branched small tree or shrub ca. 5 m. high,
glabrous throughout, leaves opposite, pinnate or (in the
var. castaneaefolium Seem, ordinarily) 1-foliolate, folioles
in 3—4 pairs, lanceolate, serrate ca. 5—10 X 1.2—2.5 cm.
Flowers yellow in terminal inflorescences, the calyx small
ca. 0.5 cm. 1., tubular campanulate with 5 broad triangular
short lobes and remarkable glands. The corolla is tubular-
funnel-shaped with 4 fertile included stamens, inserted at
the base and 1 staminodium; the thecae are long pubes-
cent (not occurring in the other Malayan species) and
divergent. The capsule is oblong ca. 10—15 X 0.5—0.6
cm. with many seeds in two rows in each cell, ca. 0.5 X
1.5—2 cm., broad membranous winged.

Geogr, distr. S. E. Asia. Upper Burma — Minbu distr.:
A. T. Gage in Rec. Bot. Surv. Ind. 3 (1904) 86 — Br. India —
Bombay: Tub ber t (L sub n. 898, 196... 301) — Chutia
Nagpur distr.: Wood in Rec. Bot Surv. Ind. 2 (1902) 125.

Malayan Archipelagp. Java — Res. Batavia. Cult, in
Hort. Bog. XI. H. 40a. (L sub n. 922, 70... 403):

Buitenzorg pr. Bantar peté: Boerlage (L. sub n. 908,
339... 33. n. 908, 339... 34. n. 908, 343... 639):
Buitenzorg: H. Hallier n. 244, in a garden pr. Djem-
batan Merah (B): Weltevreden: Backer (B) - Res.
Soerabaya — pr. Prigen: A. Rant (B).

Philippines. Mindanao — Distr. of Zamboanga: R. J.
Alvarez. F. B n. 15434 (L) — Rather widely distributed
in cultivation (Merrill).

Geogr. distr. S. E. Asia — Siam — Bangkok: Zim-
mermann. PI. Siam. n. 69 (U).

Malayan Archipelago. — Java — Buitenzorg. Cult, in
Hort. Bog.: n. 4079 HB (B) — Ternate — pr. Doefa-
doefa: Beguin n. 915 (B).

A native of tropical America from Argent., to Mexico
and now widely distributed and cultivated in the tropics.
According to Koor der s it is in Java very frequently
found in parks and gardens as well as in India and Burma
(Gage, Wood). On the Philippines Merrill notes that
it scarcely naturalises, but in the Malayan Archipelago
even this seems never to occur.

—nbsp;form, but no constant heriditary variety. As I stated, it
is connected with several intermediate forms to the type.

Bull. Soc. Linn. Normand. 7 (1862) 11; — = Diplanthera
Banks et Soland. ex. R. Br. Prod. (1810) 448; idem 1 (1827)
304; Endlicher. Gen. Plant. (1836—40) 676; D. C. Prod.
10 (1846) 299; Bureau. Soc. Bot. France 9 (1862) 163;

—nbsp;= Bulweria F. v. M. Fragm. Phytogr. Austr. 4 (1864)
147; Bureau. Monogr. (1864) 51; F. v. Mueller. Fragm.

Phytogr. Austr. 5 (1865) 72, 214; Seemann. Journ. Bot. 3
(1865) 93; F. v. Mueller Journ. Bot. 5 (1867) 212; Bent-
ham. Fl. Austr. 4 (1869) 540 ; Scheffer, Nat. Tijdschr. Ned.
Ind. 31 (1870) 332; Seemann. Journ. Bot. 8 (1870) 148,
163; Benth. amp; Hook. Gen. Plant. 2 (1876) 1048; F. v.
Mueller. Sec. System. Cens. Austr. PI. Part. I. (1889) 167 ;
Bâillon. Hist. PI. 10 (1891) 44; K. Schumann Engler.
Prantl. Pfl. Fam. IV. 3b (1895) 235; Boerlage Handl.
Fl. Ned. Ind. 2 (1899)591 ; — = Montravelia Montrousier
mss. ex Beauvisage. Ann. Soc. Bot. Lyon. 26 (1901) 89;
Ridley. Fl. Mai. Penins. 2 (1923) 552; van Steenis Nova
Guinea 14 (1927) 293.

Trees with thick branches and large, verticillate or
opposite, simple, ovate or oblong leaves, mostly with large
glands at the base, margin entire, mostly yellow tomen-
tose just as the ultimate branches and the inflorescences,
sometimes glabrous. Thyrses terminal, large, mostly yellow
tomentose. Flowers rather large ; calyx campanulate coria-
ceous, 3-lobed, valvate in bud, posterior lobe entire, lateral
lobes 2-toothed or with 5 subequal lobes, sometimes with
glands, hairy or glabrous; corolla yellow, sub-bilabiate,
5-lobed, dilated towards the throat, tubular-ventricose,
lobes ca. round; stamens 4, didynamous, seldom with a
fifth rudimentary one, distinctly exsert; anthers with dis-
tinct divergent thecae, reflexed in bud; ovary sub-sessile,
bilocular, with 2 placenta\'s in each cell, which are some-
times adnate ; ovules oo. pluri-seriate ; style exsert with a
bilamellate stigma. Capsule 2-valvate, oblong or lanceolate
with strongly-coriaceous or even woody valves; seeds
oo broad finely membranous winged.

Robert Brown, who published the genus in his
Prodromus (1810). founded on one species D. tetraphylla
R. Br., gave it provisionally a place at the end of the

Solanaceae, as he thought it „Solaneis Scrophulareisque
afpnequot;, probably because he had no capsules to his dis-
position, and supposed its alliances among the Gesneraceae
and Sesamae. E n d 1 i c h e r reckoned it to the Salpiglossidae
among the Scrophulariaceae and so did De Candolle.

Later on VieiHard (1862) described another genus
Deplanchea, founded on a New-Caledonian species, D.
speciosa Vieill. He thought it a Verbenaceae, but gave a
remark of its alliance with the Bignoniaceae.

In the same year Bureau (1862) gave a complete
description of this species and recognised the genus as a
true Bignoniacea.

F. von Mueller (Fragm. 4 (1864) 147) described a
third genus Bulweria F, v. M., founded on an Australian
Bignoniacea from the Rockingham Bay in Queensland,
B, nobilissima F. v. M. (= Tecomella Buliveri F, v. M.
in herb.), which Seemann (1865) supposes just to be
the same as Deplanchea Vieill.

It is curious, indeed, that von Mueller (Fragm. 5
(1865) 72), though he adopted the remark of Seemann,
referred Deplanchea Buliveri still to the Verbenaceae.
The same year, however (op. cit. 214), he recognised
Bulweria nobilissima to be synonymous with Deplanchea
Bulweri.

At last he suggests (Journ. Bot. 5 (1867) 212) the sy-
nonymy of Diplanthera and Deplanchea, the first being
the eldest name, though the fruit was still unknown.

Bentham (Fl. Austr. 4 (1869) 540) does not know, what
to say about the identity of the two monotypic genera,
but supposes them to be distinct genera, which he based
on the structure of the androecium. Deplanchea speciosa
was described with a fifth rudiment and a bare space
between two distinct placenta\'s, this not being the case
with Diplanthera tetraphylla.

family, corrected the first point, in saying that this may
not be such an important point within the family, the
same being the case in Nyctocalos and Newbouldia.

F. V. Mueller (Sec. Syst. Cens. Austr. PI. 1 (1889) 167),
reduces Diplanthera as a synonym of Deplanchea. Indeed
he is quite right in doing this ; there was already described a
Potamogetonacea called Diplanthera Thouars so that
this name is wrong for the Bignonian tree in respect to
the priority-principles.

Bentham and Hooker (1. c. 1048) do not indicate
the fifth rudiment in the generic diagnosis, nor say anything
about the placenta\'s, and so did Bâillon and Schumann.

This question remained still unsolved till 1901, when
Beauvisage pointed out, that in D. \'sessilifolia and
D. montana the fifth rudiment occurs in only 20 % of
the flowers, but is absent in general. Besides he suggests
the existence of two distinct placenta\'s and I fully agree
with his remark. Indeed I never found a rudiment. As to
the placenta\'s, this is a mere question of conception. Most
times the septum in the ovary shows 2 enlargements in
each cell, but the ovules are also found on the space
between now and then. Whether there is a bare space
present or not, I do not suppose of high taxonomical
importance. The ovary shows the usual Teconieae-scheme.
The flowers and the foliage are the main characters for
recognizing the species, because the capsules of several
species are still lacking up to this moment.

According to Scheffer and Teysmann it seems
that at least in D. bancana, only the old and full-grown
trees produce capsules, this being the case as well on the
original habitat as well as in the Hort. Bog., where the
three was introduced by the latter,

F. von Mueller described the uppermost leaves of
the Queensland species as opposite, whereas Bentham
pointed this out to be a mistake, which I can affirm here.
Probably this mistake is due to the fact that at the end
of a branch there may appear small axillary shoots with
opposite leaves, as Bentham adds and which I also saw
in the 3-verticillate D. bancana.

Scheffer doesn\'t think the phyllotaxis to be a constant
important character to separate the 3-verticillate D. bancana
and the 4-verticillate D. tetraphylla, but I do, as I never
saw any variations of this character.

Further I will point to a remarkable parallel in habitus
and also in respect to several taxonomical characters with
some Verbenaceae. For instance with Faradaya. This genus
was first described by F. von Mueller as a Bignoniacea
(Fragm. 5 (1865) 21) and based on a species Faradaya
splendida F. v. M. The points of resemblance are striking,
which is also the case with other Faradaya s, and very
interesting from a biological point of view. Both are trees,
possess large, entire, opposite (Faradaya and the uppermost
young shoots of Deplanchea now and then) or verticillate.
simple leaves, sometimes with glands at the base; inflores-
cences and flowers are nearly of the same size, calyx
rather irregularly orquot; split (Faradaya) and glandular at the
base, stamens exsert. Even in geographical distribution
they harmonize, Deplanchea going much further to the
west however. They are, of course, easily recognizable in
their gynaecium.

That on the contrary Deplanchea has been referred to
the Verbenaceae I have already mentioned.

Also with 0.vera (Verb.) there are points of resemblance,
which Schlechter^) mentions. I found in the Paris

1) Schlcchtcr. Die Pflanzengeographische Gliederung der Insel Neu-
Caledonien. Inaug. Diss. Engl. Bot. Jahrb. 36 (1905) 1 — 41.

herbarium a specimen (Schlechter n. 14843), which
even he had called Oxera bignonioides Schltr. n. sp., a
striking proof for the resemblance.

It is very difficult to explain these parallels. They are
truly too striking and especially too common in the
vegetable kingdom as to think they form mere examples
of accident, I suppose there must be a general rule about
it and I do not believe them to be always adaptations to
climatic or other conditions, but I suppose them inhaerent
to the evolutionary tendencies of the forms themselves.
Adaptation can indeed explain much and may be the
principal factor that turned the evolution of some forms
to xeromorphous plants, etc., here and in many other
cases it is another matter. Here is adaptation a hypothesis
explaining nothing.

In my opinion there are 2 main points for a distinction
of 2 groups. First some species possess a short, rather
broad corolla-tube: D. bancana, D. tetraphylla (Fig. 8 j)
and D. hirsuta. The other species D. novocaledonica, D.
tubulosa (Fig. 8 k), D. glabra and D. speciosa (Fig. 8 i)
show a tubular tube not concealed in the calyx.

Now there is another systematical character as to the
structure of the septum and the placenta\'s (Fig. 8a—g).
Some species show 2 rather pronounced placenta\'s, thickened
as in the first group: D. bancana. D. speciosa and D.
tetraphylla; a simple thickened septum (placenta) I have
found in D. tubulosa, D. glabra and D. sessilifolia, which
species compose a second group.

It is a pity I had no well-preserved materials of D.
novocaledonica. nor of D. hirsuta.

The groups based on the corolla and those based on
the placenta\'s seem to differ, e. g. D. speciosa belongs to
the group characterized by a long corolla-tube, but has,
however, more or less distinguishable placenta\'s which

character the other species with a long corolla-tube do
not possess. Thus it seems to me that these characters
can be used to distinguish the species, but I do not know,
whether they are important in phylogenetical respects.

I have tried to find a rather natural taxonomical cha-
racter as to the structure of the ovary, especially in regard
to the placenta\'s and the series of ovules. Indeed some
species possess a rather constant number of series: D.
glabra has 10 series, D. speciosa and D. sessili folia shovf
6—8 series, whilst D. tetraphylla has 10—12 rows per cell.

1.nbsp;Calyx inside yellow-pubescent-tomentose, outside with
few craterimorphous glands . I. D. novocaledonica.
Calyx inside always glabrous..........2

Leaves short or long petioled, not densely tomentose
at the base..................

4.nbsp;Leaves totally glabrous, inflorescence glabrescent,
corolla-tube tubular, not concealed in the calyx

Leaves at least at the underside, just as the inflores-
cence dense yellow pubescent or hirsute.....5

5.nbsp;Leaves narrow-lanceolate in outline, margins repand
and irregularly toothed, petioles short 1.5—4 cm.,
whole plant hirsute, calyx inside with scale-like glands,

Leaves ovate or obovate to oblong, margins entire,
petioles much longer 5—10 cm. 1., plants sub-gla-
brous or tomentose, calyx inside smooth, high
trees.....................6

6. Corolla-tube broad campanulate, short, rather broad,
the largest part concealed in the calyx, ca. 1.5 cm. 1.,
inflorescences large, primary axis long, rather multi-

Corolla tubular, long, rather narrow, much exceeding
the calyx in length, tube ca. 2.5 cm. 1., inflorescences
rather small, primary axis short, 1-3-flowered ... 7
1. Calyx 1.75—2 cm. 1., rather broad ventricose-cam-
panulate (1.5 cm.) sub-coriaceous, leaves rather long
cm.) petioled, obovate-oblong, sidenerves

Calyx 1.5 cm. 1. tubular, rather narrow (0.7 cm. diam.),
firmly chartaceous, leaves rather short (1.5— 2 cm.)
petioled, oblong, sidenerves 13 pairs. • 8. Z). tubulosa.

Cymis 3—5 cm. longe pedunculatis; pedunculus teres,
breviter luteo-tomentosus, basi et apice contractus, tricho-
tomus; flores medianae 1.3—^1.7 cm. longe pedicellatae,
laterales 1.5—2 cm. I. pedicellatae. Calyx coriaceus, tubu-
loso-campanulatus, extus subglabrescens, glandulis crateri-
morphis paucis praeditus, intus dense luteo-brunneus pubes-
centi-tomcntosus, ca. 1.5 cm. I., 5-lobatis, lobis subaequa-
libus, ovato-triangularibus, ciliatis, apice penicillatis. Corolla
4 cm. 1. tubo anguste cylindraceo, 0.5 cm. lato, fauce
dilatato, 5-lobata, bilabiata, lobis orbicularibus, margine
ciliatis.

Type specimen: Vieillard n. 2507).
Geogr. distr. New-Caledonia — Vieillard n. 2507 (L).
In Nova Guinea XIV I have mentioned the inside
pubescent calyx of D. sessilifolia, of which species there
is no description, but I have got an original specimen of
Vieillard. This remark of mine is a mistake, as I have
found that the materials put in the cover on the paper

Fig. 9. Geographical distribution of Deplanchea Vieillard. 1. D. novocaledonica vSts. 2. tetraphylla (R. Br.) F. v. Muell.
3. D. sessilifolia Vieill. 4. D. glabra vSts. 5. D. hirsuta (Bailey) vSts. 6. D. bancana (SchefF.) vSts. 7. D. speciosa Vieill.

do not belong to D. sessilifolia, but represent a well-
defined new species. It is a pity there are no leaves nor
capsules preserved, the latter lacking mostly, however, in
the collected species within this genus.

The species is easily recognizable by its inside hairy
calyx, which character never occurs in the other species
known. It is for this reason that I have described it, to
emphasize the fact, that New Caledonia is such an important
country for the whole genus, showing an endemic develop-
ment (name!).

Diplanthera tetraphylla R. Brown Prod. (1810) 448; F. v.
Mueller Sec. Cens. Austr. PI. 1 (1889) 167: R. Br. Prod,
(ed. Nees). 1 (1827) 304; R. Brown\'s Vermischte Bot.
Schrift. III. 1. 305; — = Bulweria nobilissima F. v. M.
Fragm. Phytogr. Austr. 4 (1863-4) 147; - = Tecomella
Bulweri F. v. M. (nomen) Op. cit. 4 (1863—4) 147; — =
Deplanchea Bulweri F. v. M. Op. cit. 5 (1865—6) 72. 214;
Bentham Fl. Austr. 4 (1869) 540; Scheffer Nat. Tijdschr.
Ned. Ind. 31 (1870) 334; F. M. Bailey Queensl.
Fl. 4 (1901) 1137; F. M. Bailey Compreh. Catalog.
Queensl. PI. (1909) 368 ; van Steenis Nova Guinea 14
(1927) 293; 111. Bot. Cook\'s Voyage. 72, t. 229; C. T.
White Contr. Knowl. Fl. Papua. Proc. Royal. Soc.
Queensl. 34 (1922) 52.

Fig. 8 c, d, j; 9 (2).
Big tree, stem with a diam. often exceeding 1 m.; bark
thick, soft and somewhat corky; the head of the tree
irregular diffuse; the (ca. 1 cm.) thick branchlets, under-
side of the leaves and inflorescence covered with a thick
soft tomentum, often assuming a golden or bronzed hue
and consisting of single or clustered but scarsely stellate
hairs. Leaves crowded at the end of the branches, in
whorls of 4 or the first leaves of the uppermost young
shoots opposite, on short 3—5 X 0.25-0.4 cm. petioles.

which are dilated at the base. Leaves ovate, obtuse or
obovate-elleptical, entire, very large, 30—60 X 30 cm.
or more broad, or those immediately under the thyrse
15—23 X 8—14 cm., the upper surface glabrous or
slightly scabrous, basal part of the midrib with 0—3 large,
brown, smooth glands, cordate or sub-cordate at the base,
firmly membranaceous, at the underside softly dense
tomentose, midrib thickened with 7—9 erect patent side-
nerves, nerves above not prominent, slightly sunken. Leaf-
scars cordate 0.6—0.7 cm. broad. Thyrses terminal, dense,
sub-globose, 15—20 cm. diam., nearly sessile above the
last leaves, common peduncle 5—7 X 0.8—1 cm., primary
axis whorled, terete, 4—6 cm. 1., 1—5 flowered, each one
dichotomously branched with a flower shortly ca. 1 cm. 1.
pedicellate in each fork. Bracts linear, minute, ca. 0.5 cm. 1.
Lateral flowers 1.5—2 cm. 1. pedicellate. Ca/yx coriaceous
ca. 1.3 cm. 1., pubescent when young, 3-lobed, lateral
lobes 2-toothed, triangular-ovate, acute lobes as long as
the tube. Corolla yellow, tube shortly exceeding the calyx,
ca. 2.5 cm 1. sub-bilabiate, outside glabrous, tubular-
campanulate, 5-lobed, lobes as long as the tube, ovate-
orbicular, margin ciliate, ca. 1 cm. 1.; stamens and style
exceeding the corolla by 2.5 cm. or more, very divergent,
inside above the base slightly pilose. D/s/c hypogyn, rather
thick. Stamens 4, didynamous at the base of the corolla
inserted and pilose there, 3.5 and 4 cm. 1., with distinct
divergent cells, in bud reflexed. Capsule 5—7.5 cm. 1., the
valves hard and woody, smooth inside with a longitudinal
line where the thick dissepiment was attached, the placenta
bearing dissepiment not broad and rather thick.

Geogr. distr. Australia — Northern Queensland pr.
Rockingham Bay (Dallachi) ; F. v. Mueller (B, P, U
sub n. 001067, n. 001066): Cape York (M. Gillavry,
Daemel); Endeavour river (Banks and Solander).

magna 20—30 cm. lata, depresso-globosa, multiflora, pe-
dunculo sparse luteo-tomentoso. Ci/mi 7—15-flori, glabres-
centes; folia supra basin 1—7-glandulosa.

Geogr. distr. New Guinea — pr. Okaba :Branderhorst
n. 38 (L, B. D, Kew, U); Thursday Island: Jaheri (B);
Astrolabe Range and Sogere, a common tree (C. T. White).

According to Bailey it is a common tree in the tro-
pical scrubs. The wood is of a whitish colour, close-
grained and firm.

I now suppose the New Guinean specimen, which I
thought slightly different from the Australian materials,
to represent a new variety. The cymes are more branched,
the leaves possess much more glands at the base and the
habit is slightly tomentose with respect to the Australian
specimens. D. tetraphylla is especially allied to D. bancana,
with its rather short, broad corolla-tube.

3. Deplanchea sessilifolia Vieillard ex Guillaumin nov.
spec. Ann. Mus. Col. Marseille (nomen). Ser. II. 9(1911)
42, 203. Cat. PI. Phanér. Nouv. Caled. et Dépend. ; Beau-
visage. Gen. Montrouziana in Ann. Soc. Bot. Lyon 26
(1901) 89.

Rami dense luteo-tomentosi ; internodiis superiis 3 X 1
cm. dense luteo-tomentosis ; folia 3-verticillata, sessilia,
rhomboidea, obovate, utrinque obtusa, vel basi leviter
truncato-sub-cordata, 9—19 X 6—11 cm. supra sub-glabra,
basi 2-vel multi-glandulosa ; glandulis magnis. pezizae-
formibus, 0.2—0.3 cm. diam., infima basi utrinque dense
luteo-tomentosa, subtus praecipue in nervis primariis
nervisque secundariis paululo tomentosa ; thyrsus terminalis,
densus, 3.5 x 7—8 cm., plane globosus, dense luteo-tomen-
tosus, pedunculus 4—5 cm. 1., bracteis violaceis lineari-
lanceolatis, 0.8—1 cm. 1., peduncuH secundarii 1.5 — 1.8
cm. longi, 3—7 flori, bracteis 0.5—0.6 cm. 1., lineari-lan-
ceolatis, pedicellis florum medianorum 0.5—0.7 cm. 1., brae-

teolis 0.3—0.4 cm. 1., Ca/i/.x: tubuloso-campanulatus, viola-
ceus, extus sparse luteo-tomentosus, intus glaber, 1.5 cm.
1., fauce 0.6 cm. latus, 5-lobatus lobis subaequalibus, trian-
gulari-lanceolatis. Corolla tubuloso-campanulata, parte
superiore ventricose-dilatata, 4 cm. 1., (lob. incl.), fauce.
8—1 cm. diam., omni ex parte glabra, 5-lobata, bilabiata,
sub-curvata, intus supra basin 0.5 cm. floccosa, lobis ro-
tundis; stamina 4, didynamia, 4 et 4.5 cm. 1., thecis
hastatis, pendulis in alabastro, divergentibus, quinto rudi-
mento interdum praesente. Ovarium oblongum, disco annu-
lari hypogyno, biloculare, placentis in utroque loculo 2,
non distincte separatis, ovulis 3—4-seriatis. Fructus mihi
ignotus.

Geogr. distr.New Caledonia pr, Gomen :M. Deplanche
n. 480 (L.P): M. Deplanche n. 132 (P).

So far as I know Vieillard never gave a description
of this endemic New Caledonian species; so I give it
here from a well-preserved original specimen of his in
the Paris herbarium.

The species is very well characterized by its sessile
leaves, with remarkable tufts of yellow hairs at the imme-
diate base of the midrib below the glands.

In these characters, the calyx and the tubular corolla,
it differs e.g. immediately from D. tetraphylla.

Now and then 6-merous flowers occur as abnormal
aberrations with a 3-locular ovary.

According to B e a u v i s a g e the fifth rudiment is most times
absent. The species seems to me mosdy allied with D.
speciosa as to its long, rather tubular corolla and D. bancana
and D. tetraphylla, both possessing however a much

4. Deplanchea glabra vSts. nom. nov. — = Diplan-
thera glabra vSts. in Nova Guinea 14 (1927) 293.
Fig. 8f, 1; 9 (4).

Tree with terete branches, glabrous throughout; inter-
nodes of the ultimate flowering branches 1 —1.5 cm. 1.,
0.8 cm. thick; leaves 3-verticillate, coriaceous, obovate-
oblong, 16—19 x 9—11 cm., 2.5—3 cm. 1. petioled, rotun-
date or cuneate-rotundate at the base, midrib at the
basal part above with 3—7 glands, with some scattered
glands along the remaining part. Thyrse terminal,
narrow, short ca. 3 cm. 1. peduncled, ca. 8 cm. 1.,
70-flowered, sub-glabrous, primary axis 1.5— 1.7 cm. 1.,
trichotomous, triflorous, pedicels ca. 0.8—1.2 cm. 1. rubi-,
ginosus. Calyx yellow, cyathiformous, 1.5—1.7 cm. 1.,
totally glabrous, lobes 5, triangular, acute, 0.5—0.6
cm. 1., 0.5 cm. broad at the base, sub-inequal, at the top
short penicillate. Corolla sulphureous, 3—4 cm. 1., cyhn-
drical at the throat ca. 1.2 cm. br., outside glabrous,
inside near the insertions of the stamens pilose, lobes 5,
orbicular, 0.7 cm. 1., ciliate. Stamens 4, exsert, ca. 5 cm. 1.,
yellow, hairy at the base, thecae yellow, divergent, 0.4 cm. 1.
reflexed in bud. Ovary subglobose, bilocular with 2 placenta\'s
per cell. Ovules in 10 rows per cell; style glabrous, ca.
5 cm. 1., stigma bilamellate, yellow. Capsules unknown.

Geogr. distr. — Dutch Northern New Guinea — pr.
Humboldt Bay: Gjellerup n. 136 (L, B, U, Kew, D).

As I mentioned in my original diagnosis, this species
is closely allied to D. speciosa VieilL, but differs in the
much more narrow thyrses, which do not exceed the
leaves; for the rest it is totally glabrous.

5. Deplanchea hirsuta (Bailey) vSts. nom. nov. —
= Diplanthera hirsuta Bailey Depart. Agric. Contrib.
Queensl. Fl. Bot. Bull. 14 (1896) 11; Bailey Queensl. Fl.
Part. IV. (1901) 1137; Bailey Compreh. Catal. Queensl.
PI. (1909) 368.

leaves opposite or in whorls of three, narrow-lanceolate
in outline, but very irregular, the ends of some being
broadly truncate, attaining 50 cm. in length with a breadth
of 16 cm. about the centre, margins repand, crenulate or
deeply and very irregularly toothed, base cordate and much
undulate, petioles 1.5—4 cm., hirsute. Calyx 1.25 cm. 1.,
campanulate, coriaceous, appearing 3-lobed, the 2 sidelobes
emarginate, the inside bearing scale-like glands. Corolla
yellow, ringent, 2.5 cm. 1., spreading to 3.75 cm. wide, the
lobes blunt, longer than the tube. Stamens 4. exceeding the
corolla by about 3.7 cm. ; style about the length of the sta-
mens, the stigmatic lobes ovate-apiculate. Capsules unknown.

Geogr. distr. Queensland — pr. Cairns, Stony Creek,
a shoot and loose flowers collected: L. J. Nugent;
Thursday Island, a large leaf and a very young shoot
gathered: E. Cowley.

Though I saw no materials of this Queensland-species
it seems to me well characterized. Firstly the leaves are
much narrower than in all other species, with repand
irregularly toothed margins. The habitus is evidently hirsute,
and not tomentose. For the rest the calyx bears scale-like
glands inside, a character I never saw in the other species.
It seems to me most allied to D. tetraphylla.

With respect to the other species, I think the opposite
leaves belong to the very young shoots, the tree in
general bearing 3-verticellate leaves.

6. Deplanchea bancana (SchcfF.) vSts. nom. nov.
— = Diplanthera bancana Scheffer. Tijdschr. Ned. Ind.
31 (1870) 334; Hasskarl in Flora 53 (1870) 219; Hooker
Fl. Br. Ind. 4 (1885) 385; Boerlage Handl. Fl. Ned. Ind.
2 (1899) 59; Prain Journ. Asiat. Soc. Bengal. 383; Ridley
Fl. Malay Renins. 2 (1923) 552.

Large tree attaining 30 m.. bark fuscous, rugose, young
parts just as the pedicels and petioles pubescent; leaves

pubescent, later on glabrous except the nerves, long ca.
4—10 cm. petioled; petiole stout, swollen at the base, cor-
date and contracted to the base, obovate-oblong, entire or
sub-repand, firmly membranaceous, at the base of the midrib
with 2—4 concave elliptic glands, midrib strong, promi-
nent beneath, sidenerves 10—12 pairs, distincdy promi-
nent, erect-patent, leaves ca. 45 x 25 cm., the uppermost
under the thyrse shorter, 15—25 x 11 —17 cm., 5—12 cm.
1. petioled, leaves at the underside sparsely yellow-tomen-
tose, the nerves densely tomentose. Thyrse terminal 5—10
cm. 1., 8—15 cm. 1. peduncled, peduncle ca. terete, shortly
yellow-tomentose, rather compact, erect with horizontal
side-axis of the cymes, which have filiformous bracts, infe-
rior axis remote, superior dense, 2.5—3 cm. 1., trichotomous,
pedicels ca. 1.25 cm. 1. Flowers articulated. Ca/yx outside
pubescent, mostly with few glands, scarcely shorter than
the corolla, as long as the pedicels, above the base slighdy
contracted, 5-dentated, teeth sub-equal, when fruiting sub-
spathaceous split. Coro//a-tube inside above the base pilose,
lobes rotundate, inferior the largest, throat glabrous, fila-
ments inserted near the base of the corolla tube and pilose
at their bases; anthers with distinct cells, in bud refiexed,
pendent. Capsule with 2 thick lanceolate valves, when dry
sub-woody ca. 12.5 x 3 cm., with oo seeds; seeds flat, obo-
vate, extremely fine membranous winged, 2 cm. 1. and 4
cm. br., germ flat, sub-orbicular, ca. 0.7 cm. diam.

Geogr. distr. Malay Peninsula — Woods, not very
common (Ridley); Malacca, Bukit Sadanen; Malacca:
Main gay n. 1214 (L): Batu Feringhi (Curtis); Sungei
Hudang (Derry); Negri Sembilan, Port Dickson (Foxwor-
thy); Selangor, Ginting Bidai (Ridley).

Sumatra — Res. Riouw — P. Karimoen pr. k. Sangka,
tree 28 m. high, alt. 10 m.: n. bb 7362 (B,L); P. Temiang:
Teysmann (B).

Banka — pr. Djeboes: Teysmann n. H.B. 9666(B);
Banka: leg.? (B); pr. Djeboes, moist habitat: Teysmann
(B), original specimens of Scheffer.

Billiton — Tandjoeng Pandan: Teysmann n. H. B.
5158 (B,L); van Rossum n. 74 (L).

Geogr. distr. — Malay Peninsula — Penang, Batu
Feringhi and Hill (Curtis) — Sumatra — Gouvt. Sumatra\'s
Oostkust — Bila pr. Aek-boeroe, 80 m. alt. in a young
forest, not seldom, tree 8—15 m. high, flowers somewhat
fragrant: Lorzing n. 11525 (B.L).

This species is somewhat variable and a glabrous form
is mentioned above. This variety is easily recognizable
from D. glabra at first sight by the long peduncled, mul-
tiflorous, much larger inflorescences, which are compactly
contracted at the top of the peduncle and the much longer
petioled leaves, which have more glands at the base.

D. bancana is rather closely allied with D. tetraphylla,
with its short broad corolla-tube, the other species pos-
sessing a much longer, tubular one.

D. bancana is also allied with D. speciosa Vieill., but
differs by its dense inflorescences, the larger flowers, the

glabrate corolla and calyx and the few (1—2) glands at
the base of the midrib.

I must refer here to an imperfect specimen (a leaf, a
single peduncle and some primary axis of a thyrse with
some buds) from Andaj collected by Teysmann and
preserved at Buitenzorg. Whether it may belong to D.
bancana I cannot decide, but it does seem so as far as I
can judge. It is a pity the specimen is indeterminable.

7. Deplanchea speciosa Vieillard. — Bull. Soc. Linn.
Normand. 7 (1862) 96; Bureau in Bull. Soc. Bot. Fr. 9
(1862) 164; — = Diplanthera Deplanchei F. v. Mueli
Journ. Bot. 5 (1867) 213; — = Diplanthera speciosa
{VieilL) K. Sch. in Engler-Prand Pfl. Fam. IV. 3b (1895)
235; Beauvisage Gen. Montrouziana. PI. Nov. Gal. Ann.
Soc. Bot. Lyon. 26 (1901) 89; A. Guillaumin Cat. des PI.
Phanérog. de la Nouv. Caléd. et Dépend. Ann. Mus.
Colonial. Marseille. Ser. II. 9 (1911) 42.

High tree, bark smooth, grayish, leaves sub-erect at
the apical part of the branches, glabrous, leafscars ob-
cordiformous. Leaves 3-verticillate, 3.5—7 cm. 1. petioled,
petioles dilated at the base, leaves puberulous. entire,
obovate, obtuse at both ends, 20—30 x 12—14 cm., pri-
mary (8) and secondary nerves puberulous beneath, for
the rest leaves glabrous, midrib at the base with ca. 2
glands. Thyrse terminal, densely yellow-tomentose, com-
pact, globose, cymes ca. 2 cm. 1. peduncled, trichotomous,
1-3-flowered, ca. 8 cm. diam., bracts and bracteoles linear-
lanceolate violaceis, peduncle ca. 7.8 cm. 1., yellow-tomen-
tose when young. Flowers large 0.6—0.8 cm. 1. pedicellate,
erect, slightly curved, yellow. Calyx glabrous, coriaceous,
campanulate, tubular-ventricose, 2.5—3 cm. 1., 5-lobed,
lobes sub-equal, triangular, 0.7—0.8 x 0.5 cm. at the apex
slightly carinaeformous contracted, resistent. Corolla totally
glabrous, bilabiate, sub-carnosous with 5 sub-equal lobes.

Stamens 4. didynamous 3 cm. 1. exsert, filaments thick,
smooth, twisted, glabrous, truncate at the apex, inserted
below the top of the connective; authers 2-celled, cells
separate, lanceolate, divergent. Ovary conical-oblong, 2-locu-
lar with 2 placenta\'s per cell ; ovules oo 6-8-seriate per cell,
septum naked at the middle. Style filiformous inflexed in
bud. Capsule erect, elliptic-oblong, acute, slighdy curved
11,—14 X 2.5—3 cm., ca. 1.5 cm. thick, shghtly carinate,
valves woody, septum membranous spongy.

Geogr. distr. New Caledonia — pr. Balade (Vieillard) ;
Wagap pr. Bondé in forest at medium altitude, many
specimens, also fruiting ones: Vieillard n. 1036 (L, P);
Deplanche n. ? (P); pr. Ourai, 400 m. alt.: Le cart
n. 65(P);pr. Ourai, fruiting specimen: Petit n. 163 (P);
mountains near Paita, alt. 200 m.: Schlechtern. 14843
(P); pr, Wagap, tree 20 m. high: M. Thiebault n. ?
(P); flowers orange, tree 15 m. high: Herb. Pan cher
(P); South New Caled.: M. Raoul (P); Noumea pr.
Prony: M. Franc n. 214 A (P); pr. Koghis: Mr. and
Mrs. Le Rat n. 1121 (P); Nouv. Caléd. : Petit n. 74
(Beauvisage); Yate (Vieillard); Baie du Sud: F our nier
et Sebert n. 62—1 (Beauvisage); Baie de Prony:
Jeauneney in herb. Pancher (P).

In Nova Guinea I made a confusion in saying that the
calyx of D. speciosa should be pubescent inside, mistaking
it for D. novocaledonica.

This species is allied with D. bancana, but differs in
the midrib and primary nerves being sparsely hairy only
beneath, further in the small compact inflorescence and
the larger flowers, which are totally glabrous.

Now and then there are abnormous flowers, e.g. 6-merous
ones, which possess a 3-locular ovary (Beauvisage in
herb. Paris).

Arbor mediocris, fere 7 m. alt.; cortice brunneo-griseo;
ramulis ultimis fere 1 cm. diam., teretibus; internodiis
basi attenuatis, brevibus. ca. LS—2 cm. longis; cicatrices
foliorum late ovatae. Folia 3-verticillata. petioli breviores.
sub-teretes. basi dilatati, flavo-tomentosi, 1-1.5 cm. longi;
lamina oblonga, integra, basi rotundata vel sub-cordata,
apice rotundata, emarginata vel sub-acuta, sub-coriacea.
supra nitida, obscure viridia, glabra etiam in nervis, basi
glandulis pezizaeformibus praedita; infra glabra, nervis sub-
tomentosis; nervi numerosiores, paralleli, fere recti, ca. 13,
prominentes, reticulatione distincte prominente. Thyrsi termi-
nales, in ramis foliatis. fere 10 cm. longi, sub-tomentosi,
densi, oblongi; pedunculo tereti, sub-glabro; axes laterales
1 —1.5cm. longis, 3-flori; pedicelli 0.7 cm. longi; bracteis
vel lanceolatis, sub-tomentosis. Calyx tubulosus, 1.5—1.6
cm. 1. (lob. incl.), tubo angusto, flavescente, 0.6—0.7 cm.
lato, intus oculo nudo glabro, dense microscopice glanduloso,
extus sparse pubescente, lobis 5 aequalibus, ovatis triangu-
laribus. 0.3 X 0. 25 cm., apice breviter penicillatis. Corolla
glabra, sufflavescente. paullo odorata. tubuloßa. tubo fere
recto. 2. 5—3 x 0. 6 cm.. 5-lobata; lobis inaequalibus oblongis
0.8x0.5 cm., sub\'ciliatis; tubo intus\'apud insertioijem stami-
num dense glanduloso-pubescente. Staminodia nulla; stamina
4. lutescentia. 2 et 3 cm. longe exserta. glabra; antheris
luteis. filamentis 4 et 4. 5 cm. longis. 0.7 cm. supra basin
tubi insertis. Discus annularis, carnosus. Ovarium glabrum.
ovatum. leviter compressum. 2-loculare. placenta in utroque
loculo singula integra, omni ex parte ovulis obtecta; ovula
co , 6—lO-seriata. Stylus glaber, filiformis, 4. 5 cm. iongus.
Capsula ignota.

Geogr. distr. New. Guinea^pr. Hollandia, Humboldt
Bay, N. New Guinea, 700 m. alt., specimen 7 m. high, single

This species is very well distinguishable by its narrow
tubular calyx and corolla, the inside microscopically glan-
dular calyx and the large number of nerves of the oblong
rather small leaves.

Its nearest affinities are to be found in D. speciosa, from
which it differs by its much narrower calyx and corolla,
the inside glandular calyx and its nerves.

(n. 191); one specimen had buds, inflorescence and leaves,
another without label showed the same flowers. He thinks
it different from Deplanchea speciosa and D. tetraphylla
from which it should differ: quot;par les fleurs plus petites,
qui le rapprochent D. montana Vieill. et d\'un échantillon
anonyme de Balansa n. 3299, dont il paraît différer nota-
blement par ses feuillesquot;.

2.nbsp;M. montana Montrouzier (nomen). Beauvisage sup-
poses this species also to ^e a distinct one. »

It is a pity he did not describe these 2 species, but
this is due to his remark „Mr. E. Bureau prépare la
révision de ce genrequot;.

I did not see any materials of these 2 species, nor of a
specimen (B e c c a r i n. 2085) from Borneo, which Bentham
and Hooker suppose to be a distinct species. Perhaps
it is D. bancana.

3.nbsp;Bignonia moluccana D. C. Prodr. 9 (1845) 144; —
= B. discolor Rich. Sert. Astrol. p. XXIX, non RBr.,
Miquel. Fl. Ned. Ind. 2 (1856) 751 —quot; Foha petiolata opposita

simplicia ex basi attenuata elliptica acuta integerrima, supra
glabra, subtus pubescentia, flores terminales lutei bracteis
spicatisquot; — Moluccas.

4. Bignonia comosa Roxb. Fl. Ind. 3(1832)103; D. C.
Prodr. 9 (1845) 144; Spathodea? comosa G. Don. Gen.
Syst. Dichl. PI. IV. 222; Miquel. Fl. Ned. Ind. 2 (1856)
751 — „Partes juniores tomentosae, folia opposita cordata,
integerrima, corymbi terminales breves foliosi, calyx 5-lobus,
corollae tabus longus gracilis, limbus bilabiatusquot;. —Moluccas.

I have not seen these 2 species. The first may be a
Deplanchea, though Deplanchea is not found as yet in
the Moluccas. The second I can not identify either. With
its entire opposite leaves it may only be a Deplanchea.

— —Dolichandta Sect. B. Dolichandrone Fenzl in Denkschr.
Baier. Bot. Ges. Regensb. 3 (1841) 113, 265; Seemann
in Ann. Mag. Nat. Hist. Ser. 3.10 (1862) 31 ; — = Spathodea
RBr. Prod. (1802—05) 471 ; Seemann. Journ. Bot. 1 (1863)
226; ibid. 8 (1870) 379; Bureau. Monogr. (1864) 50, t.
27; Benth. amp; Hook. Gen. PI. 2 (1876) 1041 part. {sect.
Markhamia et Muenteria excl.); Bâillon. Hist. PI. 10
(1891) 48; K. Schumann in Engler-Prantl. Pfl. Fam. IV
3b (1895) 240; Sprague in Kew. Bull. (1919) 303.

Trees with pinnate or single, opposite or scattered
leaves, leaflets elliptic to filiform, entire or denticulate
flowers in terminal racemes or panicles. Calyx spathaceous,
almost arcuate. Flowers fragrant, white, nocturnal, the
lower part of the tube long funnel-shaped, much exceeding
the calyx, limb ca. equally 5-lobed, mostly crispate or

dentate. Stamens 4, didynamous with a fifth rudiment,
inserted at the throat: anthers glabrous, bilocular. Ovary
sessile with oo ovules in many rows. Capsules sub-cylindric
or compressed siliquiform loculicidal, pseudo-quadrilocular
owing to an incomplete false septum. Semina oo in 4—6
rows pro loculo, corky or corky-membranous winged.

Sprague gives an excellent critical enumeration of the
species and also a revision of the essential characters in
comparison with Markhamia. The distinctions between
this two allied genera are to be found in the flowers;
the capsules in both genera are pseudo-bilocular.

The flowers of Dolichandrone are always nocturnal, pure
white, fragant, with a long funnel-shaped tube much ex-
ceeding the calyx, the Hmb being almost actinomorphic;
whereas in Markhamia the cylindric part of the tube is
very short and concealed in the calyx, only the upper
part of the funnel being visible and the limb is conspicu-
ously bilabiate; the corolla is yellow, rarely pink or lilac
or has a yellowish tube spotted with purple.

I suppose there are 2 groups to be distinguished with-
in the genus, according to those which Sprague calls
quot;species australiaequot; and quot;species asiaticae.quot; The former
group is composed of 3 allied species, only occurring in N.
Australia and Queensland, where none of the other
Dolichandrone s is found. They are distinguished from the
others in the following characters: mostly shrubs or some-
times little trees, leaves possessing a particular type of
nervature, being coriaceous with a thick epidermis, whorled,
scattered or almost filiform. I propose for this section the
name Coriaceae, for the other Membranaceae,

So I only suggest that these 3 species show peculiar
xeromorphous habitus and structure, but 1 emphasize that
I do not take into consideration the physiological question,
what xerophytism means for these plants. I feel fully
authorized to speak about xeromorphous adaptation, as

those structures always and only are found in arid coun-
tries, a fact that cannot be denied.

Urban (Ber. Deutsch. Bot. Ges. 34 (1916) 755—6), too,
makes a remark on the heterogeneous character of Do/r-
chandrone as K. Schumann has limited it. Ur ban takes
D. spathacea {L.f) K. Sch.. D. falcata {Wall.) Seem..
D. stipulata {Roxb.) Benth. (= Markhamia stipulata Spra-
gue), in a group agreeing in some respects with my sub-
genus of the „Membranaceae, and the pollen-grains of all
of them showingquot; 3 Spaltenquot;; whereas the other species
(my sub-genus Coriaceae) are very much varying with
quot;3—5 Spaltenquot;, among which also D. crispa {Ham.) Seem.
{-D.arcuata C. B. Clarke Sprague Kew Bull. (1919) 307)
occurs. For the present the most rational point of view
seems to be that of Sprague^), which I completed, viz.
Markhamia and Dolichandrone, the latter genus having
two sub-genera Coriaceae and Membranaceae which pro-
bably may be of generic rank.

1.nbsp;Leaves pinnate; leaflets long, terete, filiform or lan-
ceolate-oblong .................2

2.nbsp;Leaflets almost filiform, terete . . . 1. D. filiformis.
Leaflets broader lanceolate-linear or oblong .... 4

3.nbsp;Leaves ovate mostly scattered . . 2. D. alternifolia.
Leaves lanceolate or linear, mostly whorled (also
pinnate sometimes)......3. D. heterophylla.

4.nbsp;Leaves opposite, ca. 30 cm. 1., 3—4-jugate, leaflets
oblong or ovate, barbate in the axils of the nerves
beneath; calyx 3—4 cm. 1., along the sea-shore

Leaves much shorter, mostly whorled lanceolate or
linear as well as their leaflets. . 3. D. heterophylla.

Frutices vel abores parvi, glabri: folia simplicia vel
1-pinnata, varie disposita, sparsa, opposita vel verticillata;
foliola coriacea, rigida, plerumque lineari-lanceolata vel
filiformia; nervatura facile recognoscenda. nervis lateralibus
sub-parallelis, nervis tertiariis fere aequaliter prominentibus.
Septum angustum, pseudo-septum latum.

It would be interesting to know the cUmatic conditions
of the habitat of the 3 species. It is clear that I suppose
the origin of the species to be mainly due to xeromorphous
adaptation in the arid countries of N. Australia. The
reduced leaf-surface and the xeromorphous structure of the
segments as well as the shrubby character of the small
trees point to this. D. alternifolia would be the species of
semi-xeromorphic places, perhaps being the eldest of the
three. Then follows D. heterophylla with already much
more: reduced laminas and at last D. filifor mis, representing
an undoubtedly true xeromorphous plant.

I do not know, however, whether this forms a case
similar to the famous researches of Diels on the genus
Rhus § Gerontogeae.

The small number of specimens I saw and the lack of
any indications as to climatic conditions make it impossible
for me to work out precisely the hypothesis. Moreover
filiform leaves are not the single indicator for xerophytism.
The number of leaves, the structure of the cuticula and
the total number of stomata on the plant, together with
the moisture of the air arc striking proofs with physiolo-
gical researches on xeromorphous structures.

^ L. Diels. Die Epharmose der Vegctationsorgane bei Rhus L.
§ Gerontogeae Engl. Engl. Bot. Jahrb. 24 (1898) 568—647.

thodea ? filiformis D. C. Prod. 9 (1845) 249; - = Stereo-
spermum filiforme D. C. Rev. in Bibl. Univ. Gen. 1838
sine descript.: -\'—Bignonia filiformis A. Cunn. in Herb.
Cook, ex. D. C. Ann. des Sc. Nat. Ser. II. 9 (1839) 286,
nomen; Bentham. Fl. Aust. 4 (1869) 539; — = Dolichan-
drone filiformis Seem. Journ. Bot. 8 (1870) 383; F. v.
Mueller Sec. Syst. Cens. Austr. PI. (1889) 167; Ewart
and Davies. PI. North. Terr. 250; Sprague Kew. Bull.
(1919) 304.

Small glabrous tree; leaves irregularly 3-verticillate, irre-
gularly opposite or scattered, pinnate, 1—3-jugate; leaf-
lets filiform terete as well as the rhachis, 15—25 cm. 1.
and distant or in other specimens more crowded and shorter,
Racemes terminal pauciflorous 5—10 cm. 1. shorter than
in D. heterophylla-, pedicels elongate 3—5 cm. 1. longer
than in D heterophylla. Calyx glabrous, smooth, spatha-
ceous, 1.5—2 cm. 1. arcuate. Corolla white ca. 5 cm. 1.,
glandular, tube narrow, lobes 5, undulate-dentate. Ovary
glabrous. Capsule sub-cylindric when dry, terete when fresh,
arcuate ca. 25 cm. 1. or much smaller, glabrous. Seeds as
in D. heterophylla.

Geogr. distr. North Australia — Mt. Essington. N. E.
Australia; Leichhardt (P); Victoria river, a fruiting
specimen: F. v. Mueller (P); Copeland Island, North
coast of New Holland: A. Cunningham in herb. Cook.
(Seemann).

This scrubbish tree probably occurs in arid places in
a macqui-vegetation and is similarly adapted to a dry cli-
mate. It is allied with D. heterophylla. It is limited to
North Australia and does not occur in Queensland.

Journ. Bot. 8(1870)340; idem 382, partim; Bailey Class.
Ind. PI. Queensl. (1883) 29; Queensl. Fl. 4 (1901) 1135;
- = D. heterophylla F. v. Muell. Fragm. 4 (1864) 149,

Fig. 11.\' Geographical distribution of Dolichandrone (Fenzl.) Seem. 1. D. alternifolia Seem. 2. D. heterophylla F. v. Muell.
3. D. filifonnis F. v. Muell. 4. D. spathacea (L. f.) K. Sch. A coast-tree!. 5. D. serrulata Seem. 6. D. atrovirens Sprague.

partim; ^ = Spathodea alternifolia Br. Prod. (1810)
472; D. C. Prod. 9 (1845) 209; Bentham Fl. Austr. 4
(1869) 538; Bailey Compreh. Catal. Queensl. PI. (1909)
368; — = Dolichandrone alternifolia Benth. amp; Hook.
Gen. PI. 2 (1876) 1046; Sprague Kew Bull. (1919) 303;
— = Dolichandrone Brunonis F. v. M. in herb. Paris.
Fig. 10 a, e, f; 11 (1).

Tree, branchlets terete, smooth, with scattered leaves
alternate or irregularly opposite, single, ovate or broadly
ovate-lanceolate, obtuse or acuminate. (Those of Thursday
Island measured 8—9 X 4—5 cm. In some specimens I
saw an emarginate top or even leaves with 2 lobes, a
link to compound leaves). Leaves very coriaceous, obliquely
veined, narrowed into a long petiole, never pinnate. Capsu/e
arcuate, ca. 30 cm. 1. or shorter, flat, smooth, slightly
acuminate. Flowers unknown. Seeds rather narrow 3 x 0.5
cm. (incl. wings ca. 1 cm. I.).

Geogr. distr. Queensland — Endeavour river: Banks
and Solander; Burdekin river; F. v. Mueller; Gulf
of Carpentaria (P); Upper Lynd: Leichhardt (P);
Rockingham Bay: Dallachy (Sprague); between Clev.e-
land Bay and Rockingham Bay: Hill (Sprague); Thursday
Island, fruiting specimen: Jaheri (B).

A species evidently allied to D. heterophylla, treated
by Seemann as a variety of Z). heterophylla. Bentham,
F. V. Mueller and Bailey mark it a distinct species,
and so does Sprague, who, however, supposes some
specimens, which Bentham determined as D. hetero-
phylla, to belong to D. alternifolia.

Without doubt it is a semi-xeromorphous plant, occurring
in rather arid vegetation.

Fragm. Phytogr. Austr. 4 (1864) 149 in obs., excl. syn.;
Seemann Journ. Bot. 8 (1870) 382; Bailey Queensl.
Fl. 4 (1901) 1135; Ewart and Davies Fl. North. Terr.

250; — = Spathodea heterophylla R. Br. Prod. (1810)
472; D. C. Prod. 9 (1845) 207; Benth. Fl. Austr. 4
(1869) 538; Sprague. Kew Bull. (1919) 304; Bailey.
Compr. Catal. (1919) 364, fig. 344.

Scrubby glabrous tree 3—5 m. high with a rugged
bark and leaves crowded on the young shoots in dense
masses, mostly in whorls of 3, simple or pinnate, 1 — 3-jugate.
rather varying, sometimes 2 leaflets, oblong-lanceolate to
linear 2.5—7.5 cm. 1., simple leaves mostly lanceolate,
3.5—12 cm. 1. narrowed into the petiolule without articu-
lation, both leaves and leaflets thickly coriaceous with
very oblique veins ca. parallel to the margin. Flowers
white in pauciflorous short terminal racemes 5—10 cm. 1.,
very fragant; pedicels 1—2.5 cm. 1. Calyx ca. 2.5 cm. 1.
Coro//a-tube slender, ca. 3.5 cm. 1. dilated only at the
top; lobes nearly 0.6 cm. diam. broadly rounded with the
margins undulate and crispate. Hypogynous disk thick and
fleshy, the margin forming a short ring round the base
of the ovary. Capsule from 5 cm. to above 30 cm. 1.
compressed when dry; valves slightly concave; dilatations
of the dissepiment rather thick and corky, almost reaching
the margins of the valves. Seeds transversely oblong, the
wing on each side as long as the seed itself, ca. 0.5 x 3 cm.

Geogr. distr. N. Australia and Queensland — Gulf
country (Bailey); Islands of the Gulf of Carpentaria:
R. Brown, Henne (Bailey); Rockingham Bay: F. v.
Mueller (P); Gulf of Carpentaria: F. v. Mueller (P);
Mt. Elliott: F. V. Mueller (P); Upper Lynd, frequent
from the Upper Lynd to Mt. Essington: Leichhardt (P).

Arbores altiores, plerumque pilosi; folia 1-pinnata,
semper opposita; foliola membranacea vel chartacea orbi-

cularia vel oblonga, penninervia, nervis primariis distincte
prominentibus ; dissepimentum 4\'alatum, pseudo-septo septo
sub-aequilato.

(Species 4 mihi cognitae, Africam orientalem, Asiam
austro-orientalem usque ad Novo-Caledoniam incolentes).

4. Dolichandrone spathacea (L. f.) K. Sch. = Big-
nonia spathacea Linn. Suppl. (1781) 283; K. Schumann
in Fl. Kais. Wilh. Land. (1889) 123 ; - == Mir Ponge/ion
Rheede Hort. Malab. 6 (1686) 53. t. 29; - -Lignum
equinum Rumph. Herb. Amboin. 3 (1750) 73, t. 46;

-nbsp;= Spathodea longiflora Vent. Choix (1803) 40; - =
Spathodea Rheedii Spreng, in Syst. 2 (1825) 835 (quoad
syn.); Wallich Cat. (1832) n. 6516; Decaisne in Nouv.
Ann. Mus. Hist. Nat. Paris 3 (1834) 380; Blanco Fl. de
Filip. (1837) 499; Spanoghe Linnaea 15 (1841) 326; —
= Bignonia longiflora Willd. ex D. C. Prod. 9 (1845)
206 (in syn); D. C. Prod. 9 (1845) 206; - = Spathodea
rostrata Span. Linnaea 15 (1841) 326; — = Spathodea
grandiflora Zipp. ex Spanoghe I.e.; — = Spathodea
Loureiriana D. C. Prod. 9 (1845) 209; - = Spathodea
luzonica Blanco Fl. de Filip. ed. II (1845) 350. ed. III. 2
(1878) 284. t. 242; Wight Ic. 4 (1850) t. 1339; - =
Spathodea Diepenhorstii Miquel Fl. Ned. Ind. 2 (1856-59)
754 ; — = Spathodea Rheedii Miq. Ann. Mus. Lugd. Bat.
1 (1864) 201 ; — = Dolichandrone Rheedii Seemann
Journ. of Botany 8 (1870) 380; F. Villar. Nov. App.
(1880) 151; Vidal. Synopsis. Atlas (1883) 35, t. 73. f. D.;
C. B. Clarke in Hooker f. Fl. Br. Ind. 4 (1885) 379;
Warburg Pfl. Papuas. Engl. Bot. Jahrb. 13 (1890) 418;

-nbsp;= Dolichandrone longissima K. Sch. Engler-Prantl. Pfl.
Fam. IV. 3b (1894) 240; Koorders and Valeton Bijdr. Booms.
1 (1894) 69; Trimen. Fl. Ceylon. 3 (1895) 282; Boerlage
Handl. Fl. Ned. Ind. 2 (1899) 600; Schimper Indo-Mal.
Strandfl. 129; Ridley Transact. Linn. Soc. 3. 327; Ridley
Journ. As. Soc. Straits, n. 33. 120; Gamble Man. Ind. Timb.

ed. 2. 512 (1902); King and Gamble Mat. Fl. Mai. Penins.
Gamopet. 377 ; Brandis. Indian Trees (1906) 494 ; Watt. Diet.
Econ. Prod. Ind. 3 (1891) 174; Koorders. Meded.\'s Lands
Plantentuin 14 ( 1897) 552; Bourdillon. For. Trees. Travancore.
275; Retz. Obs. Bot. fasc. 5. 5; Willdenow. Spec. Plant.
3. 304; Beddome. For. Man. 168; Kurz. Repert. Veg.
Andaman. Isl. 43; Kurz. Journ. As. Soc. Beng. 45. 142;
Kurz. For. Fl. Br. Burma. 2. 234; Pers. Syn. 2. 173; K.
Schumann u. Lauterbach. Fl. D. Schutz. Geb. Sudsee (1901)
540; Prain. Rec. Bot. Surv. Ind. 2 (1903) 246, 247, 326;
Gage. Rec. Bot. Surv. Ind. 3 (1904) 86; Whitford. PhiUp.
Journ. Sc. 1 (1906) 674; Merrill. Philip. Forestry Bur. 1 (1903)
52; Ridley. Kew Bull. (1910) 203; Ridley Journ. As. Soc.
Straits. (1911) n. 59. 40. 146; Guillaumin in Ann. Mus.
Col. Mars. Ser. II. 9 (1911) 204; Merrill Fl. Manila.
(1912) 429; Koorders Exc. Fl. Java. 3 (1912) 184; Koor-
ders-Schumacher Syst. Verzeichn. 1 § 1. Fam. 258. 25;
Koorders and Valeton Atlas. Baumart. Java 2 (1914) t.
357; Merrill. Interpr. Rumph. Herb. Amboin. (1917)469;
Foxworthy Indo-Malay. Woods. Philip. J. Sc. 4 (1909)
557-9; Heyne. Nuttige PI. Ned. Ind. 4(1917) 166; Merrill
Spec. Blancoan. (1918) 349; Sprague Kew Bull. (1919)
304 (excl. syn. Spathodea macroloba Miq.) ; Merrill Journ.
Straits Br. Roy. Asiat. Soc. (1921) 525; Diels in Engl.
Jahrb. 57 (1922) 500; Merrill. Enum. Philip. PI. 3 (1923)
444; van Steenis Nova Guinea 14 (1927) 293; A. Guil-
laumin Catal. de PI. Phanérog. de la Nouv. Caléd. et
Dépend. Ann. Mus. Colonial. Marseille. Ser. II. 9 (1911) 42.
Fig. 31; 10 d; 11 (4).
Tree on the average 15—20 m. high, leaves opposite
ca. 20—30 cm. 1. minutely puberulous or glabrous, impa-
ripinnate, mostly with 3—4 pairs of leaflets, rhomboid
7—16.5 X 4—7 cm. often unequal at the base, petiolulate,
ovate-lanceolate, acuminate, entire nigrescent when dry,
barbate in the axils of the primary nerves at the under-

side. Racemes terminal, pauci- {2—6) florous, flowers shortly
peduncled; calyx deciduous with obscure ribs, sub-
coriaceous, spathaceous, arcuate, 3—4 cm. 1.; corolla fra-
grant, white ca. 15 cm. 1., glabrate, with a long funnel-
shaped tube 10—12.5 cm. 1. campanulate near the mouth,
sub-equally 5-lobed, lobes flmbriate-crispate. Capsule sub-
cylindric, acute, mostly straight or Httle arcuate ca. 30—
60 X 1.5—3 cm., glabrous, not ribbed; seeds including the
wings 0.6—0.8 X 1.8 cm. rectangular, wings and germ corky.

Geogr. distr. S. E. Asia—Br. India; P er r o 11 e t n. 35 (P);
Br. India; Wight n. 2333 (P)—Malabar(Rheede, Seemann,
Beddome); Travancore, banks of rivers (Bourdillon); Madras
pr. Pondichery (L); pr. Pondichery: Perrottet? (P); Sun-
dribuns, general but never common from the nothern forests
down to the coast (Seemann, Gage); Nwamadaung Hills
(Gage) — Ceylon — especially in mangrove swamps (Trimen).

Lower Burma (Ridley) — Bengal — East Bengal: Herb.
Griffith n. 4066 (P); aestuary of the Ganges: Wallich
n. 302 (P) — Andamans (Ridley); southern Andamans:
K u r z (P) — Nikobars (Seemann) — Merqui Archip. (See-
mann) — Cochin-China (Loureiro) — Indo-China — differ-
ing specimen, banks of the Me-Khong (Laos): Harm and
n. 3 (P) — Malay Peninsula — in tidal swamps common
(Ridley); Singapore, Bajon, Malacca. Ayer Panas, Pahang,
Kwala Bruas, Perak, Matang Penang, P. Betong, Setul,
Perlis, Kanga (Ridley).

Malayan Archipelago. — Ins. Simaloer: Achmad n.
368 (L). — Sumatra — Gouvt. Atjeh. — P. Bras pr. Laping:
Kds. 10654 /? (B); Sumatra (Miquel); Sumatra: Korthals
(L). — Res. Sumatra\'s Westkust — Priaman: Diepen-
horst (U) — Res. Sumatra\'s Oostkust — pr. Medang:
Bruinier n. 116 (B) — Res. Palembang — tidal forests
pr. Moearabaroe: Scheffer (B).

V.n.: koedo-koedo-oewi (Simaloer: Achmad); koeda
koeda (Priaman: Diepenhorst); oedjong pangassang (Kort-
hals); toewé ej (Atjeh: Koorders); ki arak (Scheffer).

Java. — common in W.,M. and E. Java, sometimes
plenty, e.g. in tidal swamps pr. Poeger (Kds. and Val.);
Java: Zollinger n. 2905 (P) - i^es. Bantam - (Kds. and
Val.); Tjamara. pr. Tjiringin: Kds. 210 ß (L,B). — Res.
Batavia — Tandjong Priok: Hallier, f. n. 158 (L,B);
Batavia: Vorder man (L,B); pr. Antjol near the creeks:
Backer (B); Tandjong Priok coast-creek-shore: Backer
n. 23153 (B) — Res. Preanger — Palaboeanratoe: Kds.
222 ß (L,B); pr. Tjiratjap: Backer n. 17366 (B); Tji-
kadin: M einders ma n. 4 (B) —ßan;oemas — Noesa
Kambangan: Kds. 26855 ß (L), Kds. 215 ß (L), Kds.
20232 ß (L), Kds. 213 ß (L), Kds. 214 ß (L), Kds.
24565 ß (L), Poerwokerto cuh.: Kds. 29586 ß (L,B);
north of the Kinderzee: Backer n. 31484 (B) - Res.
Pekalongan — Soebah-seashore: Kds. 211 ß (B), Kds. 212
ß (L,B), Kds. 36791 ß (L) — Res. Semarang - (Miquel);
Waitz (L); pr. Karang Anjar: Docters v. Leeuwen
n. 85 (B); Djapara, Pasokan-Doekoseti-seashore: Kds.
35302 ß (B), Kds. 35407 ß (B) — Res. Soerabaya -
Boedoeran pr. Sidhoardjo: J es wiet n. 60 (B) — Res.
Kediri — pr. Prigi, tidal swamps: Backer n. 11994 (B)
.-Res. ßesoe/c/— Poeger Watangan: Backer n. 17902
(B); Kds. 216 ß (P,B), Kds. 217 ß (P,L), Kds. 218 ß
(L), Kds. 30259 ß (L), Kds. 39819 ^ (B); Karimondjawa:
Kds. 42197 ß (B); Banjoewangi pr. Gradjakan: Zo 11 inger
n. 2905 (B); Banjoewangi: Vor derm an. n. 133 (B).

V.n.i kapal (Kds. Exc. Fl.); (Mal. or Jav.) djaran
(Preanger, Bantam, Noesa Kambangan, Batavia, Besoeki);
(Jav.) kajoe djaran (Preanger, Pekalongan, Bantam); (Jav.)
kadjeng kapal (Banjoemas, Besoeki, Preanger, Bantam);
(Jav.) kajoe pelok (Pekalongan); (Soend.)/cicf/aran (Bantam);
djaran pelok (Pekalongan: Kds.); (Mal.) koeda koeda

(Batavia: Hallier); (Jav.) djaranan (Banjoemas: Kds.): (Mad.)
kadjoe djavan binek (Besoeki).

Macioera — pr. Bangkalan: Backer n. 19323 (U, L, B.)
Timor — Timor (Miquel); Spanogue (L); Timor, 6
specimens: Spanoghe n.? (P) — P. Sepandjang: Backer
n. 28788 (B); Kali Sangka: Backer n. 26996 (B); Pabean:
Beguin n. F. 3 (B).

V. n.: (Mad.) kadjoe djharan (Kangean, Madoera).
Borneo — Korthals (L); Plante bornensi: O. B e c c a r i
n. 1745 (P.); pr. Pembliangan: Hallier f. n. 805 (L);
Labuan (Seemann); Modey (Merrill); Pontianak pr. k. Padang
Tikar: de Jong n. bb. 8295 (B); Pembliangan: Amdjah.
n. 805 (B); W. Koetei pr. k. Djembajan: E n d e r t n. 1469 (B);
Sarawak Mus. n. 188: Foxworthy n. 496 (B).
Ceram: Kornasi n. 436 (L); Agama n. 543 (B).
V.n.: (Mai.) toewi (Pontianak); kelajoe (Koetei); pokon
kajoe koeda (Ceram).

Palawan Archip. — (Merrill) Su/u Archip. (Merrill) —
Philippines — Along the sea-shore and tidal streams from
Union Prov., Luzon, to Mindanao (Merrill); Malaban,
Luzon: Merrill. Spec. Blancoan. n. 514 (L); Lucaena,
prov. of Tayabas: Whitford n. 583 (P): Sibuyan
Elmer n. 12497 (L).

V. n.: pata (Ilk.); tangas (Tagb.); tanhas (C. Bis.); tanghas
(P. Bis): tewi (Mbo); tiwi (Tag., Bik., C.Bis); fua(Tag.);
tui (Tag.).

Talaud yArc/iip. — Ins. Salikaboe pr. Marange: Lam n.
3225 (B) — Celebes — Saleyer pr. k. Benteng: Walangitang
n. bb. 3683 (B.L); Idem.: Teysmann n. 13861 (B);
Menado, sea-shore: Kds. 16250 P (B, L); Minahassa
pr. Pokoe oere; Kds. 16252 /? (L.B); sea-shore pr.
Kajoe watoe: Kds. 16249 /3 (B): Makassar (Rumphius):
pr. k. Panamboengang: N oer kas n. 55 (L, B); Menado:
de Vriese (L); Gorontalo: Riedel (B).

V.-n.: (Makass.) djaran (Makassar); kajoe pelompong
(Menado); kajoe djaran (Minahassa).

Halmaheira — pr. SinganoU : F o r s t e n (L) — Ternate
pr. Sasaketjil: Be g u in n. 1135 (L) — Amboina — common
in Amboina (Rumphius); Plant. Rumph. C. B. Robinson
n. 68 (L.B); Zip pel (L).

V. n. : djamè, kajo koeda, djodjame (Moluccas : Heyne);
ansarangi (Talaud: Lam); kati Araff(Amboina: Robinson).

New Guinea. — Loreniz river, pr. Zwaluwbivak: Ver-
steeg n. 1804 (L. B, U. D, Kew); pr. Merauke: Jaheri
(B); idem: Branderhorst n. 7 (L. B. D, U, Kew); Mam-
beramo pr. Teba: Moszkowski n. 32; Constantinhafen
(Warburg) ; Hatzfeldhafen : H o 11 r u n g n. 354, S40 ; Alexis-
hafen: Wiesenthal n. 43; Brit. N. Guin. (Warburg).

New Caledonia — pr. Balade: Vieillard n. 1001
(Seemann, P); tree 5—8 m. high, loosing its foliage in Sept..
plain of Naketi, on brookbanks: M. Balansa n. 1836 (P).

This common species is in a flowering state easily recogniz-
able by thepauciflorous(l—8) inflorescences of large fragrant
white flowers with a long narrow tube infundibuliformous
enlarged into the throat. The sub-coriaceous calyx is in bud
already arcuate at the apex and splits at one side; when
dry it appears nigrescent.

The narrow capsules ca. 25—50 cm. 1. are generally
distinctly arcuate and open first in the middle. The coria-
ceous or corky septum is narrow with 2 broad secondary
perpendicular septa and has a fragil structure easily dropping
to pieces.

When sterile the 1-pinnate (2)-3-4-(5) jugate leaves are
easily recognizable; the folioles possess an almost oblique
base and have short tufts of hairs in the axils of the primary
nerves; moreover these are enlarged at their immediate
base near these tufts. Very young leaves are seldom 2-
jugate. Now and then there are 5 pairs of leaflets.

These latter forms Miquel (1. c. 201) called the forma c,
quot;ioliolis usque 5-jugatis, acuminatisquot;. — Borneo in distr.
Bandjermassing (Korthals). I have seen his original specimens,
but I think this of no important difference.

Miquel also described a broad-leaved form as forma a
quot;foliolis late ovatis et ovato-oblongis abrupte acuminatis;
capsulis pede plerumque brevioribus, arcuatis.quot; This form
Miquel first described as a distinct species Spathodea
Diepenhorstii Miq. on a specimen of Sumatra collected by
Diepenhorst. I saw his original materials, but I can only
see herein an expression of the variability within the species
and so I consider the differences as being individual.

The same I think as to the third form of Miquel with
narrow leaflets, forma b. quot;foliolis angustioribus, magis
oblongis, sensim non abrupte acuminatis; capsulis pede
longioribus.quot; The specimen was collected in Amboina by
Zippel which the latter called provisionally Spathodea
grandiflora Zipp. (herb, nom.)

The leaflets are generally very variable, not only in
length but also in form. E. g. in the specimen: Kds.
20232 /?, some were 16 X 6 cm. with the largest breadth
in the middle, others however 8.5 X 5 cm. with the largest
breadth ca. -J- above the base.

Probably these variations are due to the age of the tree
and the place at the branches. As to the first cause den-
tate leaflets are but to be found on very young specimens,
this being already stated by Koorders and Vale ton.

The structure of the leaflets is always papyraceous,
wrinkled nigrescent when dry. This was not the case with
a specimen from the seashore in the Res. Preanger (K d s.
222 /?) which possessed thick-papyraceous, leaves, pale-brown-
green when dry,

The slighdy arcuate capsules vary in length and the
measures of Koorders and Valeton are too small. On
the average they are ca. 30 cm. 1., sometimes shorter.

but often longer ca. 50 cm. 1. and 2-2.8 cm. broad; the
seeds vary 0.6—0.9 x 1.5—2 cm.

The species is very common and scattered, known from
Malabar to the Solomon Arch, and even to New Caledonia.
So it is described in the eldest botanical works e. g. by
Rheede in the Hort. Malab. and by Rumphius in his

Dolichandrone spathacea is absolutely limited to the
sea-coasts in salt or brackish water from the mangrove up
to the estuaries of the rivers as far as the flood asserts its
influence (See e. g. Linné f. Spec. Plant, ed. 4.3 (1800) 304 ;
quot;Habitat in Javae, Malabariae, Zeylaniae, sylvis prope
aquasquot;. The remark of Kurz from the interior of Burma
is a mistake for D. serru/ata 5eem. according to Sprague.

The appearance in the rice-fields naar Kanga-village
in Lower Siam explains Ridley (Sprague 1. c.) quot;the tree
is about 80 feet tall and the most common one in the
paddyfields. Doubtless it is one of the relics of the time
when the whole of this country was a tidal swamp, gradually
filling up after the disappearance of the sea, which overlay
all this area. There are several more seashore-plants still
scattered over the paddy-fields such as the sandspunge.
Euphorbia Atoto.quot;

The means of spreading, Prain marks in his Flora of
the Sundribuns (1. c. 254) as the wind — quot;the only swamp-
forest tree for which introduction by wind seems aequi-
vocal is Dolichandrone Rheedii, which has seeds with large
membranous marginal wings.quot;

Koorders and Valeton call the wind and the water
as being due for spreading. I agree with Sprague when
he values the wind as sufficient only for a local spreading,
but regards especially the sea-torrents as the most impor-
tant means by which Dolichandrone spathacea obtained its
enormous present area. Indeed the seeds do not have a thin
membranous margin as usually is the case with the Bigno-

niaceae but the wings are of a corky structure and
certainly peculiarly fit to float.

I do not agree with Koorders when he indicates it
as a salt-plant. As above remarked Dolichandrone spatha-
cea is not limited to saltish soil, as Ridley found it in
not-saltish paddy-fields.

Beauvois Fl. Owar et Ben. (excl. fruct.) 1 (1805) 46;
Ventetat Choix, n. 40 (1803); Fenzl Denkschr. Baier.
Bot. Ges. Regensb. 3 (1841) 242, 265; Benth. ô Hook.
Gen. PI. 2 (1876) 1045; K. Schumann in Engler-Prantl.
Pfl. Fam. IV. 3b (1895) 240; Bâillon Hist. PI. 10 (1891) 46.

Spathodea campanulata P. Beauvois — Fl. Owar. et
Ben. I (1805) 46, t. 27; D. C. Prod. 9 (1845)208; Hooker\'s
Niger Flora (1849) 461; Seemann Journ. Bot. 3 (1865)
332, t. 40; Bot. Magaz. t. 5091; Benth. 6 Hook. Gen.
PI. 2 (1876) 1046; K. Schumann in Engler-Prantl. Pfl.
Fam. IV. 3b (1895) 240, fig. 92E; Treub Ann. Jard.
Bot. Buitenz. 8 (1889) 39 (with table); Kds. and Val.
Bijdr. Booms. Java. 1. (1894) 65; Koorders Ann. Jard.
Bot. Buitenz. 4 (1897) 72, fig. 138, 142-4 ; Koorders Exc.
Fl. 3 (1912) 184; Merrill Fl. Manila (1912) 429; idem
Enum. Philip. PI. 3 (1923) 447.

Tree 15—20 m. tall with opposite, 5—6-jugate leaves,
leaflets entire, nearly sessile, the younger parts yellow-
tomentose, later on glabrous except the calyx ; calyx large
4—6 cm. 1.. densely yellow-tomentose, spathaceous, arcuate ;
corolla large 7—8 cm. 1., broad ventricose-campanulate,
tubular portion short 1 — 1.5 cm., corolla totally exeeding
the oblique calyx; capsule oblong 15—20 cm. 1. with

Geogr, distr, Java — In Java common along allies and
in parks, but never spontaneous selfsown (Koorders);

Weltevreden: Backer (B); Cult, in Hort. Bog. sub n.
XI H 29b. (B); idem n. C 30b: Hallier. f. (L sub n.
29, 926... 193) — Kangean Archip, — pr. Pandeman in
the garden of a missionary: Dommers n. 251 (B) —
Borneo — Br. N. Borneo: Agama n. 460 (L) — Philip-
pines — cult, in Manila for ornamental purposes (Merrill).
V. n,: djati Manda (Kangean).

A native of tropical Africa but now cultivated through-
out the tropics. Evidently it never occurs selfsown outside
its natural area in Africa.

Linnaea. 7 (1832) 720; — = Dipterosperma Hassk,
Cat. Hort. Bog. (1844) 152; Endlicher Gen. PI. (1836-
40) 4131; D. C. Prod. 9 (1845) 210 ; Bureau Adansonia. 2
(1861) 191, t. 4; Bureau Monogr. (1864) 50, t. 29; —
n sub-genus Eu-Stereospermum Benth, cS: Hook, Gen.
PI. 2 (1876) 1047; Bâillon Hist. PI. 10 (1891) 49; K.
Schumann in Engler-Prand. Pfl. Fam. IV. 2gt;h (1895).

Trees, leaves 1—2-pinnate entire or dentated; panicles
large, lax, terminal or sometimes lateral ; calyx ovoid, open
or closed in bud. during the bloom mostly truncate or
shortly unequally lobed, seldom tubular. Corolla tubular-
ventricose, white, yellow or pale rose, lobes 5 nearly
equal, round, crispate. toothed or laciniate ; capsule elongate,
terete, sub-compressed or obscurely quadrangular, loculi-
cidally 2-valved; septum thickened sub-terete, corky, notched;
seeds few in 2 series per cell, trigonous with membranous
wings, embedded in the notches of the septum.

(Species ca. 20 in Africa, S. E. Asia and Malaya. Type
species: Stereospermum Kunthianum Chamisso in Op. cit.).

I agree with Schumann, who states the differences
between Stereospermum and Radermachera of generic rank.

The remarkable structure of the seeds and the notched
septum, next to the 2-seriate ovules and few seeds are
sufficient to keep the original conception of Chamisso
as to the genus, in comparison to the characters of
Radermachera.

The geographical distribution of Stereospermum Cham.
is limited to Africa, Madagascar, British India, Indo-China,
Burma and Malaya (Malay Peninsula, Borneo).

I did not study many African and Indian species. I know
from Africa: St. Kunthianum Cham, (common in trop.
Africa), St. acuminatissimum K. Sch. (Kamerun), St. cinereo-
viride K. Sch. (White Nile), St. senegalense Miq. (perhaps
syn. with St. Kunthianum Cham), St. integrifolium A.
Rich. (Abyssinia), St. Ar^uezana A. Rich. (Abyssinia), St.
molle K. Sch. (Mittu, White Nile), St. Harmsianum Sprague
(Angola), St. Zenkeri K. Sch. (Africa); from Madagascar
the doubtful St. euphorioides D. C.; from S. E. Asia: St.
chelonoides D. C. (common from Br. India to Borneo), St.
neuranthum Kurz (Burma), St. fimbriatum D. C. (Burma to
the Malay Peninsula). St. tetrami^m D. C. (N. Brit. Ind.),
St. suaveolens D. C. (throughout moister India to Tenasserim),
St. angustifolium Haines (Br. India), St. annamense A. Cheval
(Indo-China), St. grandiflSrum Cub. et Smith. (Burma). St.
cochinchinense Bur. (Cochin-China). St. Harmandi Bur.
(Cochin-China) and the doubtful St. Ghorta C. B. Clarke
(Ind. or.) and St. Wallichii C. B. Clarke (Burma); the
latter 2 species may belong to other genera, perhaps
Radermachera or Haplophragma.

The distribution states the close affinities of South East
Asia (British India) with Madagascar and Africa.

In the Dutch Malayan region only St. chelonoides D. C.
was gathered by Hasskarl in Borneo, without further
indication of finding-place. Perhaps it will be found in
N. Sumatra, as it occurs in the Malay Peninsula.

Plant hirsute, with glandulous rather densely branched
inflorescences; leaves sub-coriaceous, stout, leaflets 7—9,
rhomboid acuminate.........I. St. smveolens.

Plant glabrous throughout, except the flowers; inflores-
cences rather loosely branched, genuine, leaves membra-
naceous, leaflets 7—11, caudate. ... 2. St. chelonoides.

1. Stereospermum suaveolens D. C. Bibl. Univ. Genève
17 (1838) 124; Prod. 9(1845)211;— = Tecoma suaveolens
G. Don. Gen. Syst. 4.244; — = Bignonia suaveolens
Roxb. Fl. Ind. 3 (1832) 104; Wight. Ic. 4 (1850) t. 1342;
Miquel Fl. Ned. Ind. 2 (1856) 756; Beddome For. Man.
196; Brand For. Fl. 351; Kurz For. Fl. 2.231; - =
Heterophragma suaveolens Dalz. amp; Gibs. Bomb. Fl. 161;
Benth amp; Hook. Gen. PI. (Sub. Sect. 1. Xylocarpeae) 2
(1875) 1047; G. B. Clarke in Hooker. Fl. Br. Ind. 4
(1885) 382; Trimen. Handb. Fl. Ceylon. 3 (1895) 284;
K. Schumann in Engler-Prand. Pfl. Fam. IV. 3 b (1895)
243; J. J. Wood. Rec. Bot, Surv. Ind. 2 (1902) 125;
Burkill. Op. cit. 4 (1910) 123; J. G. B. Beumée Florist.
Anal. Onderz. Korte flora djati-opstand. Dissertation.
Wageningen (Holland) (1922) 33.

Tree 10—20 m. high, innovations viscous-hairy, leaves
stout, ovate to oblong 30—50 x 15—25 cm.; petiole
10—20 X 0.2—0.3 cm., terete with obscure ribs and covered
with lanceolate whitish lenticels, petiolule short ca. 0.2—
0.3 cm. 1., rather thick, hairy; leaflets 7—9, rough, hirsute
when young, later on nearly glabrous, the lowest pair is
the smallest ca. 6—7 cm. 1., the terminal leaflet the largest
ca. 13—15 cm. 1., the internodes ca. 5—8 cm. 1., leaflets
nearly sessile, ovate (the lowest) to rhomboid-oblong,
shordy acuminate or acute at the top, cuneate to
the base, ca. 20 X 10 cm. entire or a little den-
tated, little shining grayish green above (when dry)

with distinct not-prominent puberulous midrib and 8—9
primary nerves, genuine reticulate, the underside dull-
green when dry, with sparsely hirsute strongly prominent
midrib, primary nerves and reticulations. Inflorescence
many-branched, rather dense, viscous-hairy ca. 25
cm. 1., the peduncle at the base with some sterile
bracts, pedicels knob-like thickened by 2 short bracteoles.
Calyx viscous-hairy 0.6—0.8 cm. 1. campanulate with

3—5nbsp;broad triangular lobes which are ciliate apiculate
contracted at the top; corolla pale or dark purple, tube
narrow tubular glabrous below, inside with small glands,
towards the limb viscous-hairy, dilatate, 2.5—3 cm. 1.,
lobes obovate, entire, ca. 1.3 cm. 1., long pubescent-floc-
cose at the base and inside the throat. Stamens 4 with a
fifth rudiment, towards the base with small glands, inserted
just above the narrow part of the tube; anthers divaricate,
obtuse-oblong; ovary 2-celled lanceolate glabrous with 4
ribs, sometimes sparsely glandular, ovules ca. 30—40,
2-seriate in each cell. Capsule linear, terete or obscurely

4-costate,nbsp;45 x 1.5—1.8 cm., woody, shghtly rough with
lenticels; septum spongy 0.6—1.2 cm. br.; seeds0.8 X 3.6
cm. notched at the middle, sub-rectangular.

Geogr, distr, — India — Throughout moister India from
the Himalayan Terai to Travancore and Tenasserim
(Clarke) — Ceylon — Occasionally planted (Trimen).

Malayan Archipelago — Java — Cultivated in the
Botanical Gardens at Buitenzorg — Cult, in Hort. Bog.
sub n. XI. H. 14 (B); idem, ex Bengal n. H. B. 6637 (B)
— Res, Rembang — Forest-district East-Toeban. on red
vulcanic soil, only local spreading and then socially on
ca. 0.5 H. A.: B. Kruyne n. ? (B) — Res. Soerabaya-
Forest-complex Takis, S. Soerabaya. in teakforest on vul-
canic soil single groups of shrubs and some trees ca. 15 m.
high, locally forming a vegetation: J. G. B. Beumee

Geogr. distr. — Java — Res. Soerabaya — Forest-district
Takis, only one specimen: J. G. B. Beumee n. 2480 a (B).
V. n.: kajoe teken or djati teken (Soerabaya).

The specimens found in the East-Javanese teakforests
differ a little from the Indian materials I examined.

Tree 8—15 m. high, stem ca. 20 cm. diam., habitus
more stout, folioles gray at the upper side and with
smaller flowers, the tube ca. 2 cm. 1., the limb bilabiate,
two posterior lobes ca. orbiculate 0.7—0.8 cm. 1., three
anterior ones ca. 1.4—1.5 cm. 1. irregularly dentated. The
sweet-scented flowers appear before the leaves on naked
branches, whilst other sterile branches are already leafy.
The colour is described as follows: tube inside yellow,
limb browny-purple, the tube darker, yellow towards
the base.

According to Beumee (I.e. p. 33), Ten Oever and
C a r t h a u s, Tectona grandis is not endemic in the Malayan
Archipelago, but must be considered as introduced by the
Hindu\'s from British India whilst their domination.

In connection with this it may be that Stereospermum
suaveolens was imported with soil amongst Tectona-sQeds
or the Hindu\'s perhaps first cultivated the species.

Indeed in other countries it is cultivated, e.g. in Ceylon
especially about Buddhist temples (Trimen) and in Chutia
Nagpur, where it is a roadside tree (Wood).

It is remarkable that, whereas the species flowers very
well in the teakforests. one never found ripe capsules in
Java, neither in the teakforests. nor at Buitenzorg.

In East-Java it probably propagates asexual and gives
rise to new specimens by means of roots.

endemic-teakforests, together with the asexual propagation
and the very local spreading, show that Stereospermum
suaveolens is not endemic in Java, buth there represents an
anthropochorous species. It is a native in the east as far
as Tenasserim; it lacks in the Malay Peninsula.

2. Stereospermum chelonoides D» C. — Bibl. Univ.
Genève 17 (1838) 124; Prod. 9(1845)210; padn, Rheede.
Hort. Malab. 6 (1736) 26; non syn. Bignonia chelonoides
Linn. f. Suppl. 282; — = Bignonia caudata Miq. in PI.
Hohenack. n. 182; Wallich. Cat. n. 6501 ; — = Stereo-
spermum caudatum Miq. Ann. Mus. Lugd. Bat. 1 (1864) 200;

—nbsp;= Stereospermum Hasskarlii ZolL et Mor. Syst. Ver-
zeichn. (1854) 54; — = Stereospermum Hasskarlii Teysm. in
Miquel Fl. Ned. Ind. 2 (1856) 756; Wight. Ic. t. 1341;

—nbsp;= Dipterosperma personatum Hassk. in Cat. Bog. ; idem.
Flora 25 (1842) II Beibl. I, 28; idem. PI. Jav. Rar.(1848)
507; idem. Diagn. Nov. 112: — — Heterophragma chelo-
noides Dalz. et Gibs. Bomb. Fl. 160; Kurz For. Fl. 2.
230; Brandis For. Fl. 352; Beddome Fl. Sylv. t. 72;
Roxburgh Fl. Ind. 3 (1832) 106; C. B. Clarke in Hook.
Fl. Br. Ind. 4 (1885) 382; Gammie Rec. Bot. Surv. Ind.
1 (1895) 63, 83, 72; Gage Op. cit. 3 (1904) 87; Fischer.
Op. cit. 9 (1921) 132; Burkill Op. cit. 10 (1925) 331;
Trimen Fl. Ceyl. 3 (1895) 283; Haines Kew. Bull. (1922)
121; J. S. Gamble Fl. Madras V. Kew. Bull. (1924) 237.

Tree 10—20 m. high, glabrous except the flowers, leaves
30—45 cm. 1, leaflets in 3—5 pairs, elliptic, caudate 9—
12 X 4—5 cm., petiolule 0.8—1.2 cm., nerves at the un-
derside sometimes scarcely puberulous. Inflorescences ter-
minal or sometimes lateral 20—40 cm. 1. many-flowered
loosely branched, branches glabrous, genuine, main axis
with small linear lenticels; calyx campanulate ca. 0.6 cm.
1 shortly 3—5-lobed. Corolla rose, lobes yellow or pink,

crispate; capsule 30—50 x 0.8—1.2 cm. obscurely qua-
drangulate, nearly woody; septum sub-terete ca. 0.4 cm.
diam., seeds trigonous 0.5 x 0.5 cm. ca. 0.5 cm. br. mem-
branous winged at both sides, embedded in the notches
of the septum.

Geogr. Distr. India — Throughout moister India, from
the Terai of Oudh and Assam to Ceylon and Pegu and
Ava (Clarke) — Punjab — Saharanpur: M. Buysman
n. 481 (U) — Bombay — Bombay Presid.: Gibson (L
sub. n. 898, 199 . . 148) — Malabar pr. Mangalore:
Hohenacker n. 182 (U,L); Malabar: Stocks, Law amp; C.
(L sub n. 898, 199 . . 145); Malabar, Concan: Hook.
6 Thomson n. 185 (P);— Cey/on — Common (Trimen)

—nbsp;S. Br. India Mt. de Cottalam: Leschenault n. 157
(L, P) — Madras — Madras prov. Anaimalai Hills (Fischer) —
Bengal - Mt. Nilghiri: Thomson (L sub n. 898, 199. .
151); Sikkim: J. D. Hooker (Lsub n. 898, 199 . . 150);
Cult, in Hort. Calcutt.: Kurz. (B); Cult, in Hort. Calcutt.
(Lsubn. 898, 199 . . 146, P); Penins. Orient.: Wight
n. 2337 (L,P); idem (L sub n. 898, 199 . . 143); India:
F. von Mueller (L sub n. 898, 199 . . \\A\\) - Assam

—nbsp;Mt. Khasia, 300—1000 m. alt.: J. D. Hooker and
Thomson (U,L sub n. 898, 199 . . 147, P); Assam:
Dr. Kings collector n.? (U); Sibsagar Assam: S. Peal n.
69 (L) — Slam — Siam: A. E. G. Kerr n. 1167 (U,L).

V. n.: (Sanskr.) lunu-madala (Ceylon: Trimen); dunu-
madala (Ceylon: Trimen); padri (Trimen, Rheede).
Malayan Peninsula — Woods, not common (Ridley).
Malayan Archipelago — Java — Cult, in Hort. Bog.
sub XI. I. 4a. (L sub n. 921, 41 . . 227, n. 921, 41 . .
228);? Cult. Hort. Bog. n. 153 (B,U); From Coromandel,
cult, in Hort. Bog. n. 6635 (B);? Cult, in Hort. Bog. (U
sub n. 032845); ? Java cult, in Hort. Bog. (L sub n. 898,
200 . . 5, n. 898, 200 . . 6, n. 898, 200 . . 7) -
Borneo - Korthals (L sub n. 898, 200 . . 2, n. 898,

Haines (1922) supposes Stereospermum chelonoides
Linn. ƒ. not being identical with that of De Candolle.
The first type specimen of the younger Linné was col-
lected by König near Tranquebar in South India. It is
a very pubescent, even hirsute plant, with the young
leaves tomentose, the petiolules being shorter and rather
stout and the panicle being closely pubescent. These are
remarkable points of difference so that the specimen of
Linné f. may appear to be a distinct species. Those I
saw, possessed always the characters belonging to Stereo-
spermum chelonoides D. C.

As to the spreading, it occurs through moister India
(Clarke). The specimens cited from ? Java are most proba-
bly collected in the Gardens at Buitenzorg. However
those of Korthals from Borneo are as far as I can
judge not cultivated there. One might suppose a mistake
could have taken place with the labels, but I do not
believe this happened. It is a pity Korthals did not
indicate where the material was collected. I accept that
indeed Stereospermum chelonoides D. C. is found in Borneo,
though perhaps very rare. Besides it is remarkable that
nothing is known from Sumatra as it occurs in the Ma-
layan Peninsula here and there, though not common.

In the centre of its area, e.g. in Upper Assam, it is a
characteristic tree of the evergreen woods (Gammie). The
natives there make canoes from the stems and the timber
is moderately used for most purposes ; so in Bengal
(Burkill).

Zoll, et Mor. Syst. Verz. Pfl. Ind. Arch. 3 (1855) 53;
Bureau in Adansonia 2 (1862) 192. t. 2; - = Lagaropyxis

Miq, Ann. Mus. Lugd. Bat. 1 (1863) 198; Miquel. Op.
cit. 3 (1867) 250; Bureau Monogr. t. 28; Seemann Journ.
of Botany 8 (1870) 145; Benth. 6 Hook. Gen. Pl. 2 (1876)
1047; Bâillon Hist. Pl. 10 (1891) 49; K. Schumann in
Engler-Prantl. Pfl. Fam. IV. 3b (1895) 243; Boerlage
Handl. Fl. Ned. Ind. 2 (1899) 595; — = Radermachera
Hassk. Merrill. Philip. Journ. Sei. (1908) 335; Merrill
Enum. Bibliogr. Bornean Pl. Journ. Straits Branch Roy.
Asiat. Soc. (1921) 525.

Descr. emend.. Frutices vel aramp;ores,/o/iïs oppositis, 1-2-3-
pinnatis vel biternatis ; inflorescentiae terminales vel axillares
paniculiformes vel corymbiformes, e cymis compositis ;
calyx parvus, campanulatus, truncatus vel 2—5-lobatus,
saepe glandulosus ; corolla vulgo infundibuliformis vel
sub-campanulata, 5-lobata, bilabiata, tubo parte inferiore
angustato, alba, rosea vel luteo-alba, glabra vel pubescens ;
stamina 4, didynamia cum quinto rudimentario vel 5 fertilia
sub-aequilonga, basi saepe pubescentia vel glandulosa ; stylus
filiformis, stigmate bilamellato; ovarium elongatum disco
annulari carnoso basi cinctum, 2-loculare, uno loculorum
antico, altero postico, placentis in utroque loculo 2, ovulis
3^4,seriatis, id est ovulis in utroque loculo 6—8-seriatis
quinconcialibus anatropis horizontalibus, micropyle externa ;
capsula linearis, cylindrica, glabra, 2-valva, saepe spiraliter
torta, valvis coriaceis primum medio dein apice dehiscentia,
post dehiscentiam complanata, septo demum libero spongioso
crasso, capsulam totam implente valvis contrario; semina
numerosissima horizontalia parva compressissima lateribus
septi in utroque loculo inserta (medio septo nudo), corpora
sub-discoidea ala cincta lateraliter producta.

(Species ca. 19. Malayenses, Philippinenses, Chinenses
et Asiam austro-orientalem colentes.

ZolL (= R. glandulosa (Bl). Miq.); Bureau (Adansonia l.c.)
did not know the description of Zollinger and M o r i t z i
and gave (1862) a new description of the genus, founded
on the same species, a specimen of which he found in the
herbarium at Paris from Java: Zollinger n. 3141. At
the same time he gave a description of a second species
from the Philippines, which he called R. hanaihana Bur.:
Luzon, prov. Calauan: Gallery n. 50. He did not know
that Blanco (Fl. de FiHp. 501) had already described
this species as Millingtonia pinnata Blanco.

The next year Miquel described a new genus Laga-
ropyxis (1863) on 2 Malayan species, Blume called
Spathodea: L. gigantea Miq. and L. glandulosa Miq. as he
did not yet know the diagnosis of Bureau. In 1867
Miquel had seen the article of Bureau and placed the
species in Radermachera.

At last it appears that R. stricta Zoll. et Mor. and R.
glandulosa (Bl.) Miq. are identical, which Zollinger
already supposed to be the case with Spathodea glandulosa Bl.
Miquel\'added a new species R. Lobhii Miq. allied with
R. gigantea Miq., wich he had first identified with Lagaropyxis
gigantea Miq. (1863). So in 1867 there were 4 species
known: R. hanaihana Bur. (Philipp.), R. gigantea (Bl) Miq.
(Malaya), R. glandulosa (Bl) Miq. (Malaya) and R. Lobhii
Miq. (Malaya).

Seemann i) remarks that R. glandulosa Miq. always
possesses 1-pinnate leaves, but that it is said that Milling\'
tonia pinnata Blanco has 2—3-pinnate leaves, notwith-
standing that Blanco had described it with 1-pinnate
leaves.

Seemann considers R. hanaihana Bur. manifestly the
same as R. pinnata Blanco. He transfers Millingtonia
pinnata Blanco to R.pinnata (Blanco) Seem, and reduced

R. banaibana Bur. as a synonym. So he did with the second
species of Blanco viz. Millingtonia quadripinnata Blanco,
which he ahered into R. quadripinna Seem.

The difficulty was, as Merrill remarks, to determine
which were the plants, Blanco described, Blanco\'s
descriptions of both being imperfect. For the rest Blanco
did not make a herbarium. So it is very difficult, e.g. as
to M. quadripinnata, none of the Philippine species posses-
sing 4-pinnate leaves.

Later on several botanists have worked on the Philip-
pine Radermachera\'s which have principally given raise to
a number of synonyms. F. Villar transferred the
Blancoan species to Stereospermum; Rolfe-) gave a des-
cription of a new species Stereospermum Seemanni Rolfe,
based on a specimen of Cuming n. 996, a very frag-
mentary one, which had been referred by Seemann to
R. quadripinna Seem.

Merrill®) gives the history and a review of the Phi-
lippine Radermachera\'s and I accept here his interpretations,
as his conclusions have been based on considerably field
knowledge.

Since his review, a lot of new species have been des-
cribed by him, Elmer and others, so that I thought it
would be interesting to give a review of the whole genus.

Radermachera Zoll. et Mor. is undoubtedly allied with
Haplophragma P. Dop. and less with Stereospermum Cham.,
though Radermachera Zoll. et Mor. as well as Stereospermum
Cham, possess a terete septum.

Leaves 2-3-pinnate, sometimes only the lowest pinnae
2-pinnate, 3-5-foliolate............6

2.nbsp;Calyx strongly longitudinally costate. i?. coriacea.
Calyx not costate..............3

3.nbsp;Inflorescence very short ca. 4 cm. 1., leaves deUcate,
10—17 cm. 1., petiole thickened at the base, capsule

4.nbsp;Flowers small, corolla ca. 2.5 cm. 1. 3. R.Whitfordii.
Flowers larger, 3—5 cm. 1...........5

Folioles glandular at the underside of the base,
corolla ca. 3—4 cm. 1.....5. R. glandulosa.

6.nbsp;Capsule rather sturdy, hnear-lanceolate, valves ca.
2.5 cm. broad, rough with large tubercles, woody

Capsule rather elegant, linear, at most 1.25 cm.
broad, not rough, but smooth, lepidote or lenti-
cellate, sub-coriaceous to coriaceous......7

7.nbsp;Inflorescence very short ca. 4 cm. 1., leaves dehcate,
10—17 cm. 1., leaflets small 3—5 cm. 1., lanceolate-
caudate or acuminate, capsule very short ca. 6 cm.
1., the upper leaves sometimes 1-pinnate ....

2. R. brachybotrys.
Inflorescence, leaves and capsules much larger,
leaves 2—3-pinnate.............8

8.nbsp;Leaves biternate, leaflets thick, coriaceous, obtuse
or short obtuse-acuminate, inflorescence pauciflorous,

Leaves not biternate, inflorescence longer peduncled
and many-flowered, capsules 30—50 cm. 1. . . . 9

rig. IZ. oeograpnicai aismomion uinbsp;juvu. i.nbsp;w.^... f, unbsp;--------------. onbsp;11 n i

Miq. 6. R. xylocarpa (Roxb.) K. Sch. 7. R. biternata Merr. 8. R. mindorensis M^j^ «orsogonensis Flm. 10. R. pinnata (Blanco) Seem. Jl. R. sinica

11.nbsp;Leaves 3-, seldom 2-pinnate, lanceolate or oblong-
lanceolate. caudate-acuminate, 8—11 X 2—3.5 cm.

8. i?. mindorensis.
Lowest pinnae 3-5-foliolate, the upper single, leaflets
oblong, apiculate or acuminate towards the top,
11 —14 X 5—6 cm.......3. i?. Whitfordiu

12.nbsp;Calyx 2-lobed, corolla 2.5—3 cm. 1., lobes gla-
brous, inflorescences ca. 20 cm. 1.. Leaflets rather
small 5—9 cm. 1. lanceolate, light-green when dry

9. R. sorsogonensis.
Calyx (3)-4-lobate, corolla ca. 1.5 cm. 1., lobes
ciliate, inflorescences ca. 30—60 cm. 1......13

14.nbsp;Calyx with minute whitish glands and 10 furrows
ca. 3 cm. 1., 5-lobate, lobes mucronate, corolla

15.nbsp;Corolla-tube sub-cylindrical, 5.5 cm. 1., white;
shrub, sub-glabrous, leaflets shining, margin curled

16.nbsp;Stamens 5 fertile, flowers yellow, leaves great to
1 m. 1., capsule ca. 1 m. 1. . . 13. i?. pentandra.
Stamens 4 fertile, fifth rudimentary, leaves much
smaller................ ... 17

18.nbsp;Leaflets obtuse, obovate-elliptic, 4—7 cm. 1., at
the base puberulous and glandulous ca. 8 pairs of

14a. R. fenicis var. acuminata.
Leaflets oblong-lanceolate or lanceolate, short to
long acuminate, 8—13 cm. 1., glabrous and with-

out glands, above shining, flowers crowded at
the ends of the terminal branchlets 15. R. acuminata,

19.nbsp;Corolla dark spotted, puberulous towards the top,
petiole longitudinally striped, puberulous at the
base, leaflets with microscopical points

Corolla not spotted, glabrous or indistinct pube-
rulous towards the top, petiole glabrous, smooth 20

20.nbsp;High tree to 35 m., folioles with small glands at
the underside of the base, corolla pinkish-white

Lower trees 6—12 m. high, leaflets at the under-
side without glands, flowers white, yellowish-
white or pink................21

21.nbsp;Stem slender with few short branches, leaves
crowded towards the ends of these, narrow lan-
ceolate ca. 10 cm. 1., margin little recurved, ob-
tuse-acuminate; inflorescence rather short ca. 20
cm. 1. few-branched, corolla pink or whitish, deep
yellow near the throat . . . 18. i?. sibuyanensis.
Many-branched trees, branches in dense masses;
leaflets longer and acute acuminate....... 21

22.nbsp;Little tree ca. 6 m. high, leaflets with 10—12
straight parallel side-nerves; petioles and main-axe
of the inflorescence densely lenticellate, corolla large
ca. 6 cm. 1. pink, not fragrant . . 19. R.elmeri.
Tall tree 12 m. or higher, many-branched, petioles
and peduncle with few lenticels, folioles sharp
arcuate-acuminate with 7—9 curved sidenerves.
corolla yellowish-white, fragrant, 5—6 cm. 1.

1. Radermachera coriacea Merr. in Philip. Journ.
Sci. Bot. 3 (1908) 333; Enum. 3 (1923) 445.
Fig. 12 (1).

A tree, glabrous throughout; branches terete, brown,
densely lenticellate; leaves 1-pinnate ca. 20—30 cm. 1.,
leaflets 5, oblong or elliptical-oblong, firmly coriaceous,
7—14 X 3—4 cm., very shining above; the lower surface
shghtly paler and somewhat shining, densely punctate-
glandular, the base acute, the apex obtuse or shortly and
obscurely blunt-acuminate the margins rather strongly
recurved; nerves about 13 on each side of the midrib,
anastomosing, slightly more distinct than the rather lax
reticulations; petiolules of the lateral leaflets about 1 cm. 1.,
that of the terminal leaflet 2.5 cm. 1.; panicles at least
15 cm. 1.; calyx 1.8 cm. 1., narrow, strongly longitudi-
nally costate with 5 or 6 ridges, cleft down at one side
nearly to the middle, 3-toothed at the apex; corolla 4 cm.
1., the tube rather narrow, slightly enlarged above, the
lobes about 1 cm. 1., obtuse; capsules ca. 16 cm. 1., the
valves ca. 0.5—0.7 cm. wide, shining, coriaceous, glabrous,
blunt or acuminate at the apex; seeds unknown.

Geogr, distr. — Philippines — Luzon — Prov. of Tayabas
(east-coast) (Principe), Baler. Merrill n. 1099 (B).

A very characteristic species, not only in its simply
pinnate leaves and very coriaceous leaflets, but also in
its clieft and strongly ridged calyx. It is the only known
Philippine species possessing the latter character. In forests
at low altitudes (Merrill).

Litde tree, branchlets terete, 0.35 cm. diam., bark
grayish, rugose, puberulous. Leaves opposite, crowded at
the end of the branchlets, 2-pinnate and immediately below
the inflorescence 1-pinnate, glabrous, 10—17 cm. 1., slender;
petiole, rhachis and petiolules elegant; petiole 3.5—4 cm. 1.,
incrassate at the base, sub-sulcate above; leaflets lanceo-

late, caudate-acuminate 3.5—5 cm. 1., acumen obtuse, ca.
1 cm. 1.; lateral petiolules 0.3—0.4 cm. 1., terminal ca.
1 cm. 1.; lamina narrow cuneate, sub-glabrous and opaque
above, midrib distincdy prominent beneath, primary nerves
5—7 slightly prominent. Inflorescences thyrsiformous, very
small 4—5 cm. 1.; peduncle very short, 1.5 cm. 1., pauci-
florous. side-axes puberulous, trichotomous. Capsule short
6 X 0.4 cm., glabrous, terete, obtuse; valves sub-coria-
ceous; septum ca. 0.2 cm. thick, terete. 5eelt;fs minute, 0.15
X 0.3 cm., at both sides 0.25 cm. broad membranously-
winged.

Geogr. distr. — Philippines — Leyte — Mt. Abucayan:
Edano n. B. S. 41688 (L, B).

This dwarf species is very well characterized by its
genuine habitus, showing in many respects nana-characters
in regard to the other species within the genus.

3. Radermachera Whitfordii Merr. — PhiHp. Journ.
Sci. Bot. 7 (1912) 352; Enumer. 3 (1923) 447.

A small tree glabrous throughout, the young parts more
or less resinous; ultimate branches somewhat compressed
with few scattered lenticels; leaves 1—2-pinnate, 25—35
cm. 1., the basal part of the petiole more or less lenti-
cellate; leaflets 5—7 on the simple leaves, when 2-pinnate
the lower jugae have 3—5 leaflets, oblong-elliptic, sub-
coriaceous, sub-equally narrowed at both ends, the base
acute, apex acuminate, 11 — 15 X 3.5—6 cm., the underside
of the base with small glands, lateral nerves spreading,
about 12 on each side, the reticulations lax; petiolules of
the lower leaflets 1—2 cm. 1., the upper ones half as long
or shorter; panicles longer than the leaves, up to 40 cm.
in length, rather lax, the branches distant, spreading, the
lower ones up to 15 cm. 1.; flowers rather few, 2.5 cm. 1.,

the corolla sUghtly pubescent externally in the upper part;
stamens 4 didynamous with a fifth rudiment, glandular-
hairy at the base; calyx 0.8-0.9 cm. not ribbed; imma-
ture follicles 25 x 0.4 m., somewhat compressed.

Geogr, distr,^PhilippineS\'-Mindanao^Distr. of Cota-
bato—Lebak, in dry river-bottoms at low altitudes: Whit-
ford n. F. B. 11817 (D) - Distr. Lanao-Ponce n. F.
B. 23367 (D); Mt. Urdaneta: Elmer n. 13975 (L.U) —
Camaguin /s/ant/— Mambajao: Elmer n. 14213 (L.U) —
Bukidnon. Davao. Cotabato, Lanao. Zomboanga: So-
monte n. 25686. Ramos and Edano n. B. S. 37386,
Wester n. 112 (Merrill).

V,n,: banoi-banoi (Bag.); bunglai (Buk.); hali-hali (Sul.);
kutokong (Sub.); labayanan (C. Bis).

This endemic species of Mindanao occurs in thickets,
borders of forests, etc. at low and medium altitudes.

The leaves are not always simply pinnate as I saw in
the type specimen of Whitford. The species is most
closely allied with R. elliptica-Merf.. but with very differ-
ent leaves and inflorescence and smaller flowers (Merrill).
For the rest the leaflets show another nervature and form,
and the inflorescences are , more lax and not so many-
flowered.

Merrill has the two numbers collected by Elmer,
referred to R. pinnata {Blanco) Seem., but after a close
study of these specimens I suppose Merrill to be
mistaken.

4. Radcrmachera elliptica Merrill. Philip. Journ. Sci.
Bot. 3 (1908) 334; Enumer. 3 (1923) 445.

A tree glabrous throughout; branches terete, brown,
strongly lenticellate; leaves simply pinnate, about 35 cm.
1.; leaflets 5 elliptical or obovate-elliptical, 12—15 x 7—9
cm. coriaceous, shining, the base acute or somewhat acu-

minate, the apex broad, rounded short acute-acuminate,
or very shortly and broadly obtusely acuminate; nerves
about 9 on each side of the midrib, distinctly anastomo-
sing, the reticulations lax, petiolules about 1.5 cm. 1., that
of the terminal leaflet short, but the rhachis produced
about 5 cm. beyond the upper pair of leaflets; panicles
axillary, about 10—15 cm. 1., peduncled, densely fiowered
more or less resinous and shining; flowers white; calyx
about 2 cm. 1., closed in bud, obliquely split in anthesis,
not toothed, sub-membranaceous, smooth; corolla 5 cm. 1.
the tube somewhat abruptly enlarged where it emerges
from the calyx, about 1.5 cm. in diam. above, the lobes
broadly ovate, rounded, 1 cm. 1., somewhat hairy inside
near the insertions of the anthers; filaments glabrous; capsules
20—25 c.m. 1., nearly cylindrical, slightly compressed,
glabrous, shining, 0.7—0.8 cm. in diam., the apex somewhat
acuminate; seeds numerous, including the wings 1.3 cm.
broad.

Geogr. distr. — Philippines — Luzon — Prov. of Bulacan,
Angat: Aguilar n. F. B. 11141-Prov. of Rizal (Merrill)
-Merrill n. 9823 (B); Ramos n. B. S. 29412.

This endemic species of Luzon is well characterized
by its pinnate leaves, its large elliptical, coriaceous leaflets
and large flowers; not closely allied to any other known
Philippine species (Merrill).

5. Radcrmachcra glandulosa (Bl.) Miq. ^ = Spathodea
glandulosa Bl. Bijdr. Fl. Ned. Ind. (1825) 763; D. C. Prod.
9 (1845) 207; — = Bignonia Porteriana Wall. Cat. 6509;
Miquel. Fl. Ned. Ind. 2 (1856) 751; D. C. Prodr. 9 (1845)
165; — = Radermachera stricta Zoll. et Mor. Syst. Ver-
zeichn. 3 (1854) 53; Miquel. Fl. Ned. Ind. 2 (1856) 755;
— = Stereospermum glandulosum Miq. Fl. Ned. Ind. Suppl.
(1860) 240, 565; ^ =: Lagaropyxis glandulosa Miq. Ann.

Mus. Lugd. Bat. 1 (1863) 199 : Bureau snh R, stricta Zoll
et Mor, Adansonia 2 (1861) 193; Bureau. Monogr. (1864)
t. 28; Miquel. Ann. Mus. Lugd. Bat. 3 (1867) 250; See-
mann. Journ. Bot. 8 (1870) 147; Clarke in Hook. Fl. Br. Ind.
4 (1885) 383; Kds. and Val. Bijdr. Booms. Java. 1 (1894)
74; idem. Atlas. 2 (1914) fig. 356 L-M; Koorders Exc.
Fl. 3 (1912) 185; Koorders—Koorders-Schumacher. Syst.
Verz. Fam. 258. 28; K. Schumann in Engler-Prantl. Pfl.
Fam. IV. 3b (1895) 243; Heyne. Nuttige Planten Ned.
Ind. 4 (1917) 167; Ridley. Fl. Malay Renins. 2 (1923)
550; Merrill. Journ. Straits Branch. Roy. Asiat. Soc. (1921)
525; A. T. Gage. Bot. South Lushai Hills. Rec. Bot.

Fig. 3 k; 11 (5).
Small crooked tree 10—12 m. high, ca. 25—30 cm. diam.
glabrous throughout; stem disorderly branched already
± immediate above the surface of the soil; bark light-
gray, terminal branchlets thin; leaves 1-pinnate, 3—4-jugate;
leaflets entire, puberulous wherf young, distictly glandular
at the mostly convex basal part of the underside, elliptic-
acute or -acuminate with acute acumen 15—30 X 7—17 cm.,
lateral nerves 6—8 (—10), prominent, sometimes lanceolate
short acuminate 42 X 13.5 cm.; inflorescence pauciflorous to
rather many-flowered 10-15 (-20) cm. li calyx pale purple,
truncate, persistent; corolla white with pale rosa 3—4 cm. 1.;
buds dark purple; pedicels ca. 0.8 cm. 1.; capsules on ca.
1 cm. 1. pedicels, pendulous, linear 15—25 x 0.5 cm., valves
coriaceous, septum terete 0.3-0.4 cm. thick, thinly wrinkled;
many minute seeds, flat, orbicular, ca. 0.18 cm. diam. at

Geogr, distr. — Burma — pr. Ft. Lungleh. Lushai Hills
(Gage) — Tenasserim - Moulmein: Lobb (Clarke in Hook.
Fl. Br. Ind.) — Malay Peninsula — Malacca (Maingay);
Forests often on stream banks (Ridley): Batang Malacca
(Derry); Negri Sembilan. Tampin; Bukit Payong and Bukit

Kandang (Candey); State of Pahang pr. Bentong: Burkill
amp; H a n i f f n. 16600 (B); Kwala Tenok, Tahanriver (Ridley);
Selangor: H. N. Ridley n. 8537 (B); Ulu Gomba; Batu
Caves; Perak, Temengoh; Tapak (Wray); Penang (Walhch);
Sungei Penang. Curtis (Wallich); Fl. of Pahang: H. N.
Ridley n. 189 (P).
V.n.: bunga pawang.

Malayan Archipelago—SumatraSumatta: Korthals
(L sub n. 898. 199... 180. n. 898. 199... 178. n. 898.
199... 182); Sumatra: H. O. F o r b e s n. 2663 (L); pr. Siboe-
langit: Lörzing n. 5545 (L); Priaman; Diepenhorst
n. 2920 H.B.(U); Padangpr. AjerMantjoer: O.Beccari
n. 811 (L).

V. n.: (Mal.) toewie gadang (Diepenhorst).
Krakatau — (Docters van Leeuwen in Ann. Jard. Bot.
Buitenzorg 31 (1921) 126. 138. A tree scattered in the
ravines from 50—500 m. above the level of the sea.)

Java — Zollinger PI. Javan. n. 3141. perhaps type
specimen of R. striata Zoll (P) — Java: van Hasselt
(L sub n.898, 199... 177); Java: Herb. Dr. Blume n. 1365
(L sub n. 898. 199... 176. n. 898. 199... 175); Java:
Teysmann (U sub n. 032846); W. Java pr. Tjitjaringin:
Hasskarl (L sub n. 908. 332... 1031); Java: ?(L sub
898. 199... 183. n. 898. 199... 184); primary forest pr.
Oengaran: Junghuhn (L sub n. 898. 199... 191 - Res.
Bantam — pr. Pandeglang. 1 specimen on a riverbank:
Backer n. 7396 (B); Loehoer Tjaja pr. Tjipanas south
ot Sadjira: Backer n. 1966(B) — Res. Batavia — G. Rad-
nararang pr. Djasinga on abrookbank: Backer n. 10156
(B); Cult, in Hort. Bog. subXI. I: Backer (B); G.Salak
ca. 600-1000 m. alt.: Kds. 24174 ß {L) - Res. Preanger
- Palaboeanratoe; Kds. 33046 ß (B). Kds. 261 /?(L.B),
Kds. 259 ß (L). Kds. 260 ß (L.P); pr. Tjisolok: Kds.
262 ß (L). Kds. 12259 ß {L); Kds. 13250 ß (L); Mt.
Gede: timber-species from the Cede n. 336 (L) — Res.

Cherihon - Cheribon: ? (L subn. 898, 199... 192, n. 898,
199... 190) — Res, Banjoemas — Forest-distr. Noesakam-
bangan on loam upon sandstone: Kds. 269 P (B.L);
Forest-distr. Pringombo 700-1000 m. alt.: Kds. 266(L).
Kds. 267 P (L), Kds. 268 P {L) ^ Res. Pekalongan--
G. Slamat-Simpar: K d s. 263 /? (L); Soebah-distr. pr. Soerdjo,
G Praoe on vulcanic soil: Kds. 264 P (B), Kds. 36767 P (L),
pr. Gringsing: Kds. 265 P{B.L). Kds. 13434 (B, L)-
Kds. ■ 13435 (B.\'sL) — Res. Madioen — Ngebel-distr. 800 m.
alt.: Kds. 29689 P (L) — Res. Kediri — pr. Gadoengan
Pare: Kds. 22786 P (L); Forest-distr. Tjoeramanis: Kds.
12833 P (L. B), Kds. 270 P (L, B, P) - Res. Besoeki^
Banderan pr. Bondowoso: Backer n. 9547 (B).

V. n.: (Soend.) kihapit (Preanger, Banjoemas); (Jav.)
bedali, pedali, padali (Pekalongan, Banjoemas, Preanger);
poedang (Kediri); (Jav.) ambal (Pekalongan); (Jav.) godong
ambal (Junghuhn); (Jav.) lambal (Preanger); klajoe (Madioen);
kibako (Preanger); (Soend.) kisikap (W. Java); (Jav.)
bangkongan (Pekalongan); (Jav.) bangkong (Pekalongan);
(Mad.) sekar pote (Besoeki): kilanghit (Cheribon).

Borneo-Borneo: Korthals (L sub n. 898, 199... 181);
Sarawak: Beccari n. 811 (Merrill).

This species is easily recognizable by the large 1-pinnate
leaves, the leaflets of which are remarkably glandular at
the base. The leaflets vary much in form and length. In
this characters it differs totally from the other common
Malayan Radermachera gigantea (BL) Miq.; for the rest
the seeds of R, glandulosa, (Bl.) Miq. are smaller ^and
more genuine than those of the other species. Indeed the
leaflets of R. gigantea (Bl.) Miq. are also sometimes a
little glandular at the underside, but never possess such
characteristic crowded pezizaeformous often dark glands.
On this glands I saw often a black layer which is caused
by a fungus I couldn\'t determine, because only a mycelium
was present.

R. glandulosa (BL) Miq. occurs in thickets and forests
at low and medium altitudes up to ca. 900 m.; it is
often found on riverbanks; it is never mentioned as a
social species as far as I know. It is ecpecially found in
close-shadowy forests. The wood is only used as small
timber in S. W. Preanger.

It produces a lot of minute seeds, which are easily
spread by the wind. This is well demonstrated by the
fact that Docters van Leeuwen found the plant in
Krakatau (1919); this is undoubtedly due to wind-spreading.

The vernacular names are not always to be trusted as
Koorders and Valeton remark. As I cannot judge
this, I mentioned the names I found on the labels.

6. Radermachera xylocarpa (Roxb.) K. Sch. — =
Bignonia xylocarpa Roxb. f^ort. Bengal. 47; Fl. Ind. 3
111 (1832) 108; K. Schumann in Engler-Prantl. Pfl. Fam.
IV. 3 b (1895) 243; — = Stereospermum xylocarpum Wight.
Ic. (1850) t. 1335-6; Wallich Cat. 6511; D. C. Prod. 9
(1845) 169; Beddome Fl. Sylv. t. 70; Dalz. and Gibs.
Bomb. Fl. 159; — = Tecoma xylocarpa G. Don. Gen.
Syst. 4,225; — = Spathodea xylocarpa Brand. For. Fl.
349, t. 43; Clarke in Hook. f. Fl. Br. Ind. 4 (1885)383;
Bentham and Hooker. Gen. PI. 2 (1876) 1047.

Tall elegant tree, 10—20 m. high; bark tuberculate-
lenticellate, rather spongy, brown when dry. Innovations
pubescent, mature glabrous. Leaves opposite, 2-pinnate,
glabrous, crowded at the end of the branchlets, ca. 30—
100 cm. 1., rhachis minutely lenticellate. Leaflets short
ca. 0.2—0.7 cm. 1. petiolulate, entire, base oblique, rounded
or cuneate, ovate to elliptic-oblong, acute or acuminate
9^12 X 3—6.5 cm.; nerves 4—6, reticulations lax. F/owers
in rigid terminal ovate thyrses, totally yellowish-brown-
pubescent, small, when young ca. 3—4 cm. 1., later on
ca. 15 cm. 1.; common peduncle short, erect, firm. Calyx

1.2—1.6 cm. 1., pubescent, mature glabrous, lobes 3—5,
short, broad. Corolla much like that of H. macroloba,
ventricose from near the base, sub-glabrous, fragrant, ca.
3.7—5 cm. 1., white, tinged yellow; lobes rounded, crispate,
nearly equal. Filaments hairy below. Capsule linear-lanceo-
late, stout, sub-terete, sub-straight or strongly curved,
rugged, 25—75 x 2.5-3J cm., ca. 2.5 cm. thick: septum
terete, 0.7 cm. diam. shining-yellow, corky, bubbled, per-
pendicularly to the woody valves. Seeds (incl. wings)
0.6x3 cm. thinly discoid: germ orbicular, broader than
the wings.

Geogr, distr. Deccan Peninsula — Common, extending to
Sarpura Range (Brandis); Calcutta: Cult, in Hort. Calc.
(P, L sub n. 898.200 ...49); Cult, in Hort. Calc.:
Wallich n. 299 (P): Concan: Hook. f. amp; Thomson
(P); Prov. of Madras, Neelgherry Hills, fruiting spec.:
Herb. Wight, n. 3342 (Kew); Santavery Ghat Bobab,
ca. 1200 m. high: A. Meebold n. 9444 (P): Anaimalai
Hills, Prov. of Madras (Fischer. Rec. Bot. Surv. Ind. 9
(1921) 132).

7. Radermachera biternata Merrill. — Philip. Journ.
Sci. 1 (1906)Supp.3.238; ibid. 3 (1908) Bot. 334; Enumer.
3 (1923) 445.

Descr. emend. A small tree about 8 m. high, quite gla-
brous throughout with biternate leaves, elliptical ovate,
usually obtuse leaflets and few-flowered panicles much
shorter than the leaves, the flowers about 5.5 cm. 1.. Branches
gray or brownish, the younger parts black when dry.
Leaves 20cm. 1. or less, opposite biternate; leaflets 5—9
X 2.5—5 cm., coriaceous, shining, the apex rounded obtuse
or broadly acute, the margins revolute, sometimes with
little glands at the base; midrib strongly prominent, prim-
ary nerves about 10 on each side rather distinct beneath,
the reticulations netted, rather close; petiolules of the

lateral leaflets 1.5 cm. 1. or less, of the terminal ones about
3 cm. 1.; inflorescence much reduced, the rhachis 3 cm. 1.
or less, the branches very short or none at all, or some-
times (in a fructiferous state) longer, rhachis 15—20 cm. 1.:
side-branches 3-florous, 3—4 cm. 1. always few flowers;
calyx about 1 cm. 1., closed in bud, in anthesis unequally
3-lobed, the lobes short, acute; corolla 5—5.5 cm. 1., the
tubular portion less than 1 cm. 1. about 3.5 cm. in diam.,
enlarged-ventricose above, the lobes about 1.5 cm. 1.; style
2 cm. 1.; capsules 10—18 cm. 1., not pendulous, ca. 0.5 cm.
br. terete, narrowed at both ends acuminate to the apex,
valves coriaceous, smooth, indistinctly ribbed in the middle,
seeds numerous kidney-shaped pointed above ca. 0.6 cm.
br. and 0.2 cm. 1. at both sides 0.6 cm. br. winged, flat.

(Description of the capsules and
seeds of E1 m e r n.12836 (U);
type specimen; Merrill n.
568 from Culion (D)).
Geogr. distr. Philippines. — Culion: E. D. Merrill n.
568 (D); Merrill n. 9602 — Palawan - Mt. Pulgar:
Elmer n. 12836 (D,L.U,B) — Busuanga — Cur ran n.
F. B. 3491.

This endemic species of the Palawan-group grows in
open grassy valleys slightly above the sea-level and in
secondary forests at low altitudes ascending to 500 m.
and is well characterized by its biternate, coriaceous leaves.
The inflorescence is reduced, though not so much reduced
as Merrill states.

8. Radermachera mindorensis Merr. — Philip. Journ.
Sci. 3 (1908) Bot. 338; non Millingtonia pinnata Blanco;
= Stereospermum pinnatum Rolfe Journ. Linn. Soc. Bot.
21 (1884) 314; Vidal. Rev. PI. Vase. Filip. (1856) 203;-
= ? Stereospermum quadripinnatum Naves in Fl. Filip. ed.
3. t, 252; Merrill. Enum. 3 (1923) 446.

A tree glabrous throughout, about 20 m. high; branches
terete, brown or gray, lenticellate; leaves 3-pinnate rarely

2-pinnate.nbsp;40—50 cm. 1.. rhachis lenticellate; leaflets lan-
ceolate or oblong-lanceolate, chartaceous. somewhat shining.
8^11 X 2—3.5 cm., the base acute or somewhat acuminate,
acumen about 2 cm. I, acute; nerves 12 on each side of
the midrib, anastomosing, slightly more distinct than the
secondary ones and reticulations; petiolules of the lateral
leaflets about 0.5 cm. 1., those of the terminal leaflets
1—2 cm. l; panicles terminal, glabrous, diffuse, equalling

flowers light purple; calyx somewhat campanulate 0.4-0.5
cm. 1., closed in bud, in anthesis shortly and irregularly

3—5nbsp;toothed; corolla 1.5—1.8 cm. 1. the portion within
the calyx slender, tubular, then abruptly enlarged and
tubular-campanulate, somewhat pubescent on the outside,
irregularly lobed; capsules 45 cm. 1., 0,4-0.5 cm. in
diam., somewhat compressed; seeds, including the wings,
about 1.3 cm. br.

(Type specimen: Merrill 893).
Geogr, distr. — Philippines — Mindoro — pr. Calapan;
Merrill 893; Pola: Merrill 2240, 2473; Bongabong
river; Whitford n. 1387; Baco river: Mc. Greg or,
257; Cuming 1517; Fenix n. B. S. 15166; Ramos
n. 39650; Merritt F. B. 3731; Merrill-Luzon-Prov.
of Nueva Vizcaya: R. J. Alvarez. F. B. 18402 (L.).
V. n.: hanai banai (Tag., P. Bis.).
Allied to R. pinnata (Blanco) Seem, but with more
diffuse panicles and much smaller flowers. In his Enum.
Merrill remarks that this species of secondary forests
at fjow altitudes is perhaps better to be reduced to R.
pirinata. Indeed a specimen collected by R. J. Alvarez,
by Merrill determined as R. pinnata I am obliged to
refer here.

Folia 2-3-pinnata, subviridia (in sicc.), glabra, firmiter
papyracea; petiolus insignis ad 20 cm. 1., 0.3 cm. crassus
basi et apice incrassatus, basi articulatus, lamina ca. 20 cm.
longa, late triangularis, rhaches laterales anguste sulcatae;
foUola lateralia fere sessilia, terminalia ca. 0.5—1 cm.
longe petiolulata; foliola lanceolata 5—9 x 1,6—3.2 cm.,
integerrima, basi cuneata sensim in petiolos angustata,
margine subrecurvata, acumine 1—2 cm. longo acuto vel
obtuso, nervo mediano subtus valde prominente, nervis
lateralibus vix prominentibus, reticulatione indistincta;
inflorescentia terminalis, glabra, 20 x 6—10 cm., pedunculo
ca. 8—10 cm. 1. ca. nonnisi basi paulo lenticellato, axibus
lateralibus 2.5—4 cm. 1. trichotomis; flores modesti 2.5—3
cm. 1.; calyx ca. 1 cm. 1., 2-lobatus, microscopice albido-
lepidotus praesertim ad basin, ceterum laevis, glaber, cam-
panulatus, lobis ca. 0.5 cm. 1. inaequalibus altero integer-
rimo, late ovato, altero 2—lobato: corolla tubuloso-infun-
dibuliformis. 2.5—3 cm. 1., glabra vel apice puberula;
ovarium griseum, lineari-lanceolatum 2-loculare, glabrum
vel apice puberulum, ovula oo , 8-seriata in utroque loculo;
capsula ca. 20 cm. 1. anguste linearius 0.4 cm. lata, glabra,
septo subtereti.

Geogr. distr. Philippines — Luzon — Mt. Bulusan pr.
Irosin: Elmer n. 15337 (L, B, U).

10. Radermachera pinnata (Blanco) Seem. — =
Millingtonia pinnata Blanco. Fl. de Filip. (1837) 501; ed. 2
(1845) 351; ed. 3. 2 (1878) 285; Seemann. Journ. Bot. 8
(1870) 147; Merrill. Philip. Journ. Sc. 3 (1908) Bot. 336;
D. C. Prod. 9 (1845) 182; Miquel. Fl. Ned. Ind. 2 (1856)

753- — = Millingtonia quadripinnata Blanco I.e. c. 501,
351.\' 2.286; Miquel. Fl. Ned. Ind. 2 (1856) 753; - =
Radermachera hanaihana Bureau. Adansonia 2 (1861) 194;
Seemann. Journ. Bot. 8 (1870) 147; Merrill. Philip. Journ.
Sc 1 (1908) Suppl. UA-, - = Stereospermum hanaihana
Rolfe Journ. Linn. Soc. 21 (1884) 314; Rolfe Rev. PI.
Vase. FiHp. (1886) 203; Vidal Ph. Cuming. Philip. (1885)
132; ^ = Stereospermum Seemanni Rolfe. Journ. Linn.
Soc.\' Bot. 21 (1884) 314; Vidal. Phan. Cuming. Philip.

(1885)nbsp;132; Vidal. Rev. PI. Vase. Filip. (1886)203;- =
Stereospermum quadripinnatum F. Vill Novis. App. (1880)
151- Vidal. Sinopsis. Atlas. (1883) 35. t. 73. fig. A; Rolfe
in Journ. Linn. Soc. Bot. 21 (1884) 313; Vidal. Phan.
Cuming. Philip. (1885) 132; Vidal. Rev. PI. Vase. Filip.

(1886)nbsp;202; — = Radermachera quadripinna Seem. Journ.
Bot. 8(1870) 147; — = Stereospermum pinnatum F. Vill
Nov. App. (1880) 151; Rolfe. Journ. Linn. Soc. 21 (1884)
314; Vidal. Phan. Cuming. Philip. (1885) 132; Vidal PL
Vase. Filip. (1886) 203; Merrill. Philip. Journ. Sc. 3 (1908)
Bot. 336; idem. Spec. Blancoan. (1918) 350; idem. Enum. 3
(1923) 446.

Leaves 25—40 cm. 1.. 2—3-pinnate. rhachis not lenti-
cellate, pinnae with 2—3 pairs of leaflets, leaflets lanceo-
late narrowed into an acute or sub-obtuse acumen, 6—10
X 2.5—3.5 cm., smooth and glabrous above, the under-
side with distinct midrib, sidenerves indistinct just as the
reticulations. Thyrses 30-60 cm. 1. rather dense, main
axis ca. 0.25—0.4 cm. thick, without Ienticels, terete,
shortly tomentose. especially near the base more densely
pubescent, secondary axis thin 3-4 cm. 1. Calyx mostly
4-(2-3-) lobed, ca. 0.6 cm. 1., lobes short, obtusely rounded,
tube sometimes a little glandular; corolla pink or pale
purple and marked with yellow inside infundibuUformous,
the inferior part ca. 0.5 cm. 1. narrow cylindrical then

enlarged tubular-campanulate, ca. 2 cm. 1. as the lobes
to the apex puberulous; lobes ca. orbicular, ciliate, ca.
0.5 cm. 1.; stamens mostly 4 with a fifth rudiment or 5
fertile ones inserted ca. 0.5 cm. above the base, in the
basal part glandular-hairy, ca. 1.5 cm. 1., thecae divergent,
connective elongate. Capsule ca. 60 cm. 1., seeds imbri-
cate, winged.

(Description of Merrill spec. Blancoan. n. 834).
Geogr. distr. Philippines — Luzon — Cagayan to Camarines
(Merrill); Calauan: Mc. Gregor n. B. S. 12399 (L);
Prov. Laguna: F. M. Amarillas n. F. B. 24674 (L);
Rizal Prov.: Merrill, spec. Blancoan. n. 834 (L, B);
Aroroy, Prov. Masbate: Whitford n. 1696 (B); Lu-
zon: G. P. Ahern n. 61 (B) — Mindoro (Merrill) —
Mindanao (Merrill) - Samar (Merrill) - Biliran (Merrill)
— Negros (Merrill) — Philippines without number, etc. —
Plant. Cumingian. (L sub n. 27, 926.. . 194, n. 27, 926... 195,
n. 27, 926...196(; pr. Samar: Rosenbluth n. F.
B. 12848 (P).

V.n.: anagep (Ibn.); ansohan (Bis.): atiatip {lg.);badlan
(Bis.); banai-bdnai (Tag., Bik., S. L. Bis.); banai-banai-
laldki (Tag.); banaibayan (Pang.); 6am-6dm (Tag., Sbl.);
barangauan (Ilk); bunlai (Mbo.); kalapuing (Tag.); lanunisi
(Ibn.); lasilak (Ibn.); pagalayan (Bon.); paiton (Pang.);
paling-udk (Bik.); pata (Pang.); salai (Tag.); tuing-hulo

Geogr. distr. Philippines — Range and habit of the spe-
cies; endemic in the Philippines (Merrill) - Luzon -
Mindanao.

Merrill remarks in his revision of the genus Rader-
machera in the Philippines (1908) that it is the most common
species of the genus in the Philippines, being somewhat
variable and its synonomy rather complicated, primarily

due to Blanco\'s imperfect descriptions and to several
later interpretations of these. The leaves are never simply
pinnate, but 2- and 3-pinnate, frequently on the same spe-
cimen. No one of Blanco\'s imperfect descriptions corres-
ponds with this common species, but Merrill states that
this becomes comprehensible, as it seems very evident that
Blanco had no herbarium, and so the probability of
repetition increased. The flowers vary in size from 2.5—
3 cm., but on all the specimens cited by Merrill (1908)
the flowers are uniformly described by the collectors.

As I consider some specimens, cited by Merrill as
R, pinnata, as belonging to R, mindorensis, I did not note
the cited specimens in his revision (1908), from which I
did not see any materials.

I examined several flowers and discovered the remark-
able fact, that here within one inflorescence occurred flowers
with 4 and 5 stamens.

11. Radcrmachera sinica (Hance) Hemsley — = Ste-
reospermum sinicum Hance. ]ourn. Bot. 20 (1882) 16; Hems-
ley in Hooker. Icones PI. 2728.

Small tree, leaves opposite, 2-pinnate, with 4 pairs of
pinnae, leaflets ovate-lanceolate, entire, rounded at the base,
obtusely acuminate to the top, membranaceous, pale beneath,
at both sides minute impressed-punctate, ca. 5 X 2—2.5 cm.
lateral petiolules 0.25 cm. 1., terminal 1.5 cm. 1. Thyrse ter-
minal, erect, lax; pedicels 0.75—1 cm. 1.; calyx cupular-
campanulate with minute whitish glands ca. 3 cm. 1. slightly
10-sulcate; lobes 5 triangular, mucronate sub-equal ca.
0.75 cm. I. ; corolla glabrescent pale sulphureous, infundi-
buliformous, ca. 7.5 cm. 1., lobes rotundate crispate ca.
2.5 cm. 1. ; stamens 4 didynamous, glabrescent, hardly exsert,
with a fifth rudiment; ovary bilocular with 2 placenta\'s

in each cell and 8-seriate ovules; capsule pendent, sub-
terete, pluri-sulcate, acuminate ca. 40 X 1 — 1.25 cm. minute
whitish-glandular; seeds 0.5 X 1.4cm. inclosed the acute
wings, cotyles flat.

Geogr. distr, China, — Canton — near the river Lien-
Chan: C. Gerlach n. 20797.

Hance considers the species to belong to the sub-genus
Radermachera of Hooker and to be especially allied to
R, hanaihana Bur, (= R, pinnata (Blanco) Seem,) Though
I have not seen any specimen it seems to me rather iso-
lated in taxonomic respect. The seeds are sown by Dr.
Gerlach in the Botanical gardens at Hongkong; after 2
years and a half the specimens attained 3 m. in height
and flowered already.

12. Radcrmachera palawanensis Merrill — Philip.
Journ. Sc. 3 (1908) Bot. 336; idem. Enumer. 3 (1923).

A shruh nearly glabrous, or the branches, rhachises of
the leaves and panicles slighdy pubescent; leaves about
20 cm. 1., bipinnate, the lowest pair of pinnae with 5
leaflets, the next with 3 leaflets, the upper ones single;
leaflets oblong-elliptical or lanceolate-elliptical 3.5—8 x
1—2.5 cm., coriaceous, glabrous, shining on both surfaces,
the margins rather strongly recurved, the base acute, the
apex more or less acuminate, sometimes apiculate and
rarely with one or two irregular teeth at the apex; lateral
nerves about 8 on each side of the midrib, not very
distinctly anastomosing; petiolules of the lateral leaflets

0.3—0.8nbsp;cm. 1., that of the terminal one longer. Panicles
as long as the leaves, lax, sparely flowered; calyx sub-
cylindrical, narrowed below, obscurely lobed, about 1 cm.

1.;nbsp;corolla white, 5—5.5 cm. 1. the portion within the calyx
very slender, tubular, then abruptly enlarged forming a

broadened tubular portion 2—2.5 cm. 1., the limb spreading,
about 3 cm. in diam., the lobes broad, rounded: capsules
very slender, about 20 cm. 1., the valves at least 0.3 cm.

Species evidently endemic on Palawan occurring on rocky
slopes along streams about 1000 m. high.

Tree ca. 9 m. high, ultimate branches thick, densely
lenticellate; leaves very great ca. 1 m. 1., or perhaps
longer, with a firm petiole, 2-pinnate or sometimes 3-
pinnate at the basal part, pinnae circ. 4, remote, with 3—7
leaflets; leaflets petiolutate, coriaceous, ovate-lanceolate,
without the petiolules 10—15 (—22) cm. 1.. entire, acute-
acuminate, cuneate or rounded at the base, glabrous, sub-
shining above, pallid beneath, with 8—10 primary nerves
on each side. Flowers yellow ca. 6.25—7.5 cm. in diam.
in terminal, ca. 30 cm. 1. rigid, lax thyrses, flowers on
short slender pedicels; calyx campanulate circ. 2.5 cm. 1.
and br., irregularly lobed, lobes acute; coro//a broad cam-
panulate glabrous except near the insertions of the stamens ;
5-lobed. lobes subequal broad, entire, rounded, reflexed;
perfect stamens 5, nearly equal, hardly longer than the
corolla-tube ; disk thick fleshy cupuliformous shortly lobed,
when fruiting enlarged; ovary glabrous, narrow cylindrical
ca. 1.25 cm. 1.; ovules oo in many rows; capsu/e sub-terete
ca. 1 m. 1. primarily lepidote, valves sub-coriaceous 0.6—
0.8 cm. br.; seeds 1 — 1.4 cm. br. with flat cotyles.

Allied to i?. sinica {Hance) Hemsley but much larger in
all parts and further differing in the broader corolla and
the 5 perfect stamens.

Though I have not seen this Chinese plant it represents
a well-marked species after the excellent description.

A small tree 3—5 m. high, glabrous throughout: branches
terete, grayish-brown; leaves opposite, about 20 cm. 1.,
the lowest pinnae 3-foliolate, the others of single leaflets;
leaflets oblong-elliptical to obovate-elliptical, 4—7 x 1.5—

4nbsp;cm. rather thin, shining, the apex obtuse, acute or
somewhat acuminate, the base cuneate, the lower surface
minutely punctate with few little glands, margin sub-re-
curved ; lateral primary nerves about 7—9 on each side of
the midrib, anastomosing, scarcely more distinct than the
secondary nerves and reticulations; petiolules 0.5 cm. 1.
or less, that of the terminal leaflet 1 — 1.5 cm. 1. Panicles
terminal, narrow, about as long as the leaves, the brac-
teoles linear-setaceus, about 0.4 cm. 1.; calyx somewhat
campanulate, epunctate with microscopical glands, 1 cm.
1., 2-lobed. one lobe with 2, the other with 3 small teeth;
corolla white about 4 cm. 1. the first 0.5 cm. slender tubu-
lar, then abruptly enlarged and campanulate. 3 cm. br.
above, the lobes broad rounded; stamens 4 glabrous except
near the glandular-hairy base; capsu/es somewhat compres-
sed, about 11 cm. 1. and 0.6 cm. thick, glabrous; seeds
including the wings 1 X 0.3 cm., apiculate.

Geogr, distr, Philippines — Batanas Islands — Santo
Domingo de Basco: Fenix n. B. S. 3583 — Babuyan
Islands — pr. Dalupiri: R. C. Mc. Gregor n. B. S.
10124 (D), n. 10194 — Luzon — Prov. of Ilolocos,
N. Luzon: M. Ramos n. B. S. 27510 (B), n. 32868;

var. acuminata vSts. nov. var. Foliola lanceolata,
non oblanceolata aut obovata, acuminata, nervi laterales 6;
juga inferiora 7-foliolata; foliola papyracea; inftorescentia
brevis (an semper?) ca. 7 cm. 1.

Malayan Archipelago — Celebes — Makassar, pr. Pang-
kadjene: Teysmann n. 12160 (B); pr. Balchangia:
Teysmann n. 12750 (B); Boeton, P. Moena pr.Wang-
kolo, 20 m. alt.; n. bb. 3925 (L, U); Minahassa: Kds.
16256 /? (L).

A species well characterized by its small obovate-obtuse
or short obtuse-acuminate leaflets, greatest breadth above
the middle, its comparitively short capsules and its blunt-
acute or only shortly acuminate leaflets.

Merrill notes already the Mindoro-specimen as a sepa-
rate form and though the materials of this one are rather
imperfect I think it a distinct variety, the differing char-
acters not being sufficient as to a diagnosis of a new
species.

15. Radermachera acuminata Merrill — Philip. Journ.
Sc. 3 (1908) 335; — = Radermachera quadripinnata Rolfe.
Journ. Linn. Soc. Bot. 21 (1884) 313 nec Millingtonia
quadripinnata Blanco nec Radermachera quadripinna Seem.;
Merrill. Enumer. 3 (1923) 445.

A tree glabrous throughout or the inflorescence obscurely
puberulent; leaves 2-pinnate, about 40 cm. 1., the lower-
most pinnae with 5 leaflets, the next with 3 leaflets
and the upper ones simple; leaflets oblong-lanceolate or

lanceolate, 8—13 x 2.5—4 cm., the base acute, the apex
slenderly long-acuminate, coriaceous, the margin recurved
(when dry), beneath at the base and towards the apex
with little glands, slightly shining, lateral nerves about
12 on each side of the midrib, not prominent, anastomosing,
the reticulations fine, indistinct; petiolules 0.8—1.2 cm. 1.,
that of the terminal leaflet 2.5 cm. 1., many-flowered;
thyrses terminal about 25 cm. 1., the primary branches
ca. 5 cm. 1. many-flowered; flowers crowded at the ends
of the branches; calyx closed in bud, in anthesis campa-
nulate, about 1 cm. 1., sometimes 2-lobed (Ramos n. 11029)
with minute glands; corolla 4 cm. 1., the portion within
the calyx slender tubular, then abrupdy enlarged and
campanulate, about 2 cm. br., the lobes rounded, broad;
capsules ca. 40 cm. 1., 0.5 cm. thick, cylindrical, spirally
twisted, ca. straight, 2-valved, valves coriaceous nigrescent
when dry; seeds minute including the thin membranous
wings (0.4 cm. br.) 0.2 X 1 cm..

(Type specimen: Gam mi 11 n. F. B. 277; description
of the capsule and seeds after Mc. Gregor
n. B. S. 32586).

Geogr, distr. Philippines — Luzon — Prov. Pangasinan,
Prov. Pampanga (Merrill); Prov. Zambales: V. Elgin-
colin n. F. B. 28305 (L) — Guimaras: Gam mill n. F.
B. Ill — Masbate (Merrill) — Panay — Prov. Antigue:
R. C. Mc. Gregor n. B. S. 32586 (L) — Cebu — Ramos
n. B. S. 11029 (D) — Mindanao — Camaguin: M. Ramos
n. B. S. 14464 (D), Exscritor n. 2495,. Mc. Gregor
n. 32367 - Basi7an- D.P. M iranda n. F. B. 18966 (L,B).

Merrill refers also 3 fragmentary specimens, namely:
Cuming n. 1003 from Luzon, Whitford n. 1696 from
Masbate and Foxworthy n. 1949 from Luzon.

This endemic species occurs on forested slopes at low
and medium altitudes, according to Merrill.

Arbor; rami ca. 0.6 cm. er. cortice subfusco puberulo,
dense lenticellis orbicularibus 0.05—0.1 cm. diam. obtecto;
folia 2—\'3-pinnata: petiolus teres 8—12 x 0.2—0.3 cm.
sparse lenticellatus, fuscus, indistincte longitudinaliter striatus,
basi puberulus, interdum ut inflorescentia et foliola micros-
copice griseo-lepidotus; foliola elliptico-lanceolata, 8—12
x 3—4 cm., integerrima, laevia, glabra, opaca, utrinque
praesertim subtus ad basin subtiliter glanduloso-punctata,
longe obtuse vel subacute acuminata, acumine —2.5 cm.
1., sensim in petiolos cuneato-attenuata, lateralia 0.5 cm.
terminale 1.5—2 cm. longe petiolulata, petiolulo supra sul-
cato ; nervi laterales 11 — 13, indistincte prominentes, nerva-
tione grosse areolata absolute prominente ; thyrsus terminalis,
firmus, sublaxiflorus, 25—40 cm. 1., pedunculo 7—8 x 0.25
—0.3 cm., sparse lenticellato, ramis primariis 7—8 cm. I.
ut pedunculo, calyceque grisee (microscopice) lepidotis;
flores in apice ramulorum conferti; calyx 2—lobatus ca.
1.5 cm. 1. (incl. lob.), laevis, glaber, lobis ca. 0.5 cm. 1., late
rotundatis ; corolla ca. 5 cm. I., basi 0.5—0.6 cm, anguste
cylindracea, intus glanduloso-pubescens, subito dilatata ven-
tricoso-infundibuliformis, fauce ca. 2.5 cm. lata, ubique
punctata, extus apicem versus pubescens lobis rotundatis
1.5—1.8cm. diam. ciliatis; stamina fertiha 4, didynamia,
1.5—1.8 cm. 1. quinto sterile, basi pubescentia ; thecis diver-
gentibus, connectivo elongato; ovarium 2-loculare, ovulis
00, 6—8-seriatis in utroque loculo; capsula mihi ignota.

Geogr. distr. Philippines — Luzon — pr. Irosin: Elmer
n. 14817 (L,U); idem: Elmer n. 16066 (L,U,B),

acuminata, but differs in the 6ne grayish lepidote leaves,
inflorescences, etc. and the dark-spotted corolla, which is
even to be seen when dry.

17. Radermachera gigantea (Bl.) Miquel. — = Spa-
thodea gigantea Blume Bijdr. (1825) 761; Miquel Ann.
Mus. Lugd. Bat. 3 (1867) 250; — = Bignonia amoena
Wallich PI. Asiat. Rar. 2 (1831) 78, t. 183; D. C. Prod. 9
(1845) 207; Miquel. Fl. Ned. Ind. 2 (1856) 755; - =
Stereospermum hypostictum Miq. Fl. Ned. Ind. Supp. I
(1860) 240, 565; — = Lagaropyxis gigantea (Bl.) Miq.
Ann. Mus. Lugd. Bat. 1 (1863) 198; - = Radermachera
Lobbii Miq. Op. cit. 3 (1867) 250; Clarke in Hook. Fl.
Br. Ind. 4 (1885) 384; — = Stereospermum amoena A.D.C.
Prod. 9 (1845) 208; — = Spathodea Lobbii Teysm. et
Binnend. Nat. Tijdschr. Ned. Ind. 25 (1863) 413; - =
Bignonia amoena Wallich Cat. n. 6512; — = Raderma-
chera amoena Seem. Journ. Bot. 8 (1870) 146; Brand. For.
Fl. 349: Kurz For. Fl. 2. 232; Koorders and Valeton. Bijdr.
Booms. Java. 1 (1894) 72; idem. Atlas Baumart. Java. 2(1914)
t. 356 A-K; Koorders. Exc. Fl. Java. 3 (1912) 185; Koorders-
Schumacher. Verzeichn. 1 § 1 Fam. 258 (1913) 26; K.
Schumann in Engler-Prantl. Pfl. Fam. IV. 3b (1895) 243;
Koorders. Bot. Tabellen Djatibosschen Java. ed. 2(1908);
Heyne. Nuttige PI. Ned. Ind. 4 (1917) 167.

High tree, full grown 40 m. high and 80 cm. diam.,
in some countries trees, 20—25 m. high are common fKds.
and Val.); stem straight, mostly terete, first branches ca.
6—15 m. above the surface of the soil; bark 1.5—2 cm.
thick, dark-gray, ultimate branches slender; leaves 2-pinnate
(on very young and young specimens the leaves are 1-resp.
2-pinnate with sharply serrate leaflets) ca. 35 cm. 1., shining
above, opaque beneath; leaflets membranaceous to sub-coria-
ceous, entire, lanceolate, base and apex acuminate ca. 8— 11.5
X 3—4.5 cm. (5.5 X 1.5 cm. on young leaves), at the basal

part of the underside mostly with few scattered or dense
small glands, midrib prominent at the underside, sidenerves
indistinct. Thyrses many-flowered, terminal, erect, ca. 20—
40 X 10—20cm., bracts and bracteoles deciduous; calyx
pale green, ca. 2—2.5 cm. 2-labiate with 3—5 short teeth,
deciduous; corolla infundibuliformous, 5—6 cm. 1. (incl. the
lobes), fragrant, whitish with pale rosa or sub-purple, lobes
unequal, crispate, ciliate, narrow tubular portion near
the insertions of the stamens dense glandular-hairy; stamens
4, didynamous, with a fifth rudiment, filaments at the base
dense-glandular-hairy, white; anthers yellow, thecae diverg-
ent, connective apiculate; ovary with oo ovules, 6-seriate
in each cell; capsules pendulous 20—60 cm. 1., green when
young, later on gray; seeds oo, 0.85—1.1 x 0.25—0.3 cm.,
sometimes narrower 1.05 x 0.25 cm..

This common Malayan Radermachera differs at first sight
from the other common Malayan species R. glandulosa
(BL) Miq. by its 2-pinnate leaves, the leaflets of which are
never remarkably glandular at the base and are much smaller
and lanceolate, the tree attaining a greater height. For the
rest the seeds are distinctly larger, as well the germ as
the wings, which vary little, perhaps due to the place in
the capsules.

Junghuhn (Java. I. 352) says of it: quot;this tree raises its
top far above the forests, so that one is hardly able to
distinguish with the naked eye its long siliquiformous fruits
and flowers, howsoever magnificent and handsome coloured
the latter may bequot;.

In Java it occurs from the seashore up to 1500 m. and
in some districts of Java it is very common. According
to Heyne, Koorders and several collectors it gives a
good timber, which is used for houses and bridges.

such a lot of materials as I studied, already mentions the
rather large variability.

First there is R. Lobhii Miq, this being synonymous
to Spathodea Lobhii, founded by Teysmann and
Binnendijk on a cultivated specimen from the botanical
gardens at Buitenzorg, they received from Th. Lobb of
Singapore.

Later on Miquel joined R, gigantea and R. Lobhii,
the latter being a synonym. The botanists at Buitenzorg
however, meaning this being wrong, wrote to Miquel,
that indeed R. Lobhii represented a distinct species and
gave him the principal differences (Miquel. 1867 1. c.), these
being the following: „small tree ca. 7 m., firm leaflets
which are obtuse-apiculate, small flowers ca. 2.5 cm. 1.,
the limb being not crispate, the tube white with 9 yolk-
yellow stripes inside, the white lobes yellow spotted near
the base, capsule long, ca. 50 cm..quot;

It is a pity that flowers fail on Miquel\'s original
specimens from Buitenzorg in the Leiden herbarium, so
that I was not able to examine this important point.

Before I am going to write about the value of the
species, I must refer here to two other varieties Miquel
distinguishes of R, gigantea, namely:

f, sumatrana Miq. — foliolis plerisque sublanceolatis,
capsulis plerumque ultrapedalibus.

f. borneensis Miq. — foliis saepe totis fere bipinnatis,
foliolis crassioribus.

The first form he had four years ago described as
Stereospermum hypostictum Miq. on an imperfect specimen
of Teysmann from W. Sumatra near Padang.

Further we have to deal with R. amoena Seem., which
C. B. Clarke calls synonymous to St. hypostictum. This
species was first described by Wallich; the author says
it must have been introduced (1818) by C. Telfair as
a native of the Mauritius and that it even should be

endemic in Madagascar, thriving luxuriantly in Bengal,
forming a small but truly ornamental tree on account of
its flowers and foliage, the former of which being exquisite
fragrant. That R. amoena should be indigenous in Mau-
ritius or Madagascar I ignore, fully agreeing with See-
mann (I.e. 146). The latter supposes, it was introduced
from Ava, as Wallich notes in his Catalogue. The
excellent table and description of this botanist show that
the flowers are ca. 5 cm. 1. with a large 2-labiate calyx,
the corolla being not crispate, the white ciliate lobes and
tube being striped with yellow, the tube outside subrosa and
the leaflets lanceolate, acuminate, coriaceous, 7.5—12.5 cm. 1.

From the Malayan Archipelago Ridley only notes
R, amoena, synonymous to St. hypostictum; therefore I
think he does not consider R, gigantea and R. amoena
being the same.

Howsoever I have tried to come to a conclusion about
the value of these five species and varieties. I have not
been able to solve the question.

Indeed I have found one Sumatranan specimen (Lörzing
n. 11238) justly conformable to R, Lobbii,

As to the ƒ. sumatrana and ƒ. borneensis, of which I
saw M i q u e I\'s original specimens, these always possess
coriaceous leaflets, but much larger flowers than R. Lobbii.

As to R. amoena, of which I saw some materials from
Malacca, this species has larger flowers than R. Lobbii,
but also coriaceous leaflets, however, this is a much larger
tree (up to 27 m.. Ridley), but corresponds with R. Lobbii
by its yellow-striped corolla, this character as far as I
know never occurring in R. gigantea.

Besides. I examined a lot of materials from Sumatra.
Borneo and Celebes, collected in order of the Forestral
Experiment-station at Buitenzorg. This extensive collection
is without a single exception badly collected, the specimens
being nearly always sterile and also very incomplete as

to the notes on flowers and fruit, so it did not help me,
but did make the question only more troublesome. I must
add here that for the genus Radermachera the leaves are
often of litde value as to principally specific characters.

I have come to the following poor preliminary con-
clusion: R, amoena Seem, and R. gigantea {BL) Miq,
are perhaps not synonymous with each other. R, Lobbii,
R. amoena Seem., St. hypostictum Miq., R. gigantea {Bl.)
Miq. f. sumatrana and f. borneensis form in many respects
a distinct group, perhaps a distinct species, not only in
phytographical respects, but also in geographical distri-
bution. as I never saw such specimens from Java.

Nevertheless I hesitate to come to a conclusion for the
present, as I have never seen any Radermachera alive, though
I studied a considerable collection herbarium materials.
The exact conclusion I leave to the future, as I have
seen the trouble which Merrill had with R. pinnata and
R. quadripinnata. owing to conformable conditions. Boer-
lage (Handl. Fl. Ned. Ind. 2 (1899) 600) supposes all the
species mentioned above to be synonymous with R. gigantea
(Bl.) Miq,

The specimens I preliminary refer to the above mentioned
group I suppose existing, I have marked with a note of
interrogation,

For the rest another question may be discussed here,
namely that possibly R. gigantea may occur in the Philip-
pines. Though Merrill described a lot of species from
this Archipelago he does not mention this one. Only one
specimen (Ahern n. 61) was first determined as R. gigantea,
this being later on reduced to R. pinnata. a truly distinct
species, though very variable in habit and certainly not
the same as R. gigantea, In geographical respect it would
be peculiar, that whereas R. gigantea shows such a large
area and represents a common Malayan species it would
not occur in the Philippines, in which country the

Moreover the Philippine Rademachera\'s have most
probably originated from Malayan species, (p. ...),

Prof. E. D. Merrill wrote me dd. 30 JuH 1927 that
he also met with many puzzling points as to some species.
In the future it will be necessary to make an extensive
study about Radermachera.

Geogr. distr. India—Tenasserim—1 Tavoy (Wallich) —
Malay Peninsuld—1 Malacca and Singapore, common in
woods (Griffith, Maingay, Ridley); Singapore: A. C.
Maingay n. 1212 (B);? cult, in Hort, Gale. (P).

Malayan Archipelago—Sumatra—Gouvt. Atjeh — P. Waj:
K d s. 10656 ß (B), K d s. 10655 /? (B); Gajoe and Alaslanden
pr. Tampang:v. Daalen n. 316(L, B)—Tapanoeli—1
W. Karolanden pr. Laoe Balang, 250-500 m. alt., tree
6—10 m. high, flowers white inside with several yolk-
yellow stripes, calyx purple, in secondary forests; Lörzing
n. 11238 (B); N. Bataklanden, 1000 m. alt.: Junghuhn
(L sub n. 898, 200... 17);? Batang Toroe, type specimens
of Miquel\'s f. sumatrana: Teysmann n. H. B. 1177
(B, U); — Simaloer—Achmad n. 164 (L), Achmad
n. 811 (L, U)—Gouvt. Sumatra\'s Oo5f/cusi—Semeloengan
pr. Masekat Hoeba, 700 m. alt.: n. bb 4882 (L); Karo-
landen pr. Simboekan Goenoeng: Galo engi n. 466 (B);?
Karolanden pr. k. Lao Soloe 2-300 m. alt.: n. bb 9300 (L);
Wampoe valley pr. k. Pajangpen Goenoeng 500 m. alt.:.
Galoengi n. 442 (B); Simeloengan n. bb 3088 (L);?
Karolanden, 900 m. alt.: n. bb 6253 (L);? Karolanden pr.
k. Tongkoh, 1500 m. alt.: n. bb 6632 (B) - Res. Suma-
tra\'s Westkust—1 Fort de Kock, 920 m. alt.: A. V. Theu-
nissen n. 6 (L); ? Pajacombo pr. d. Moenggoeng, 1320 m.
alt.: n. bb 6620 (B); Padang Pandjang pr. k. Tambangan;
n. bb 6697 (B); Priaman: Diepenhorst n. H.B. 2084
(B) — Riouw en Onderhoorigheden—1 P. Soekoe pr. Tand-

joeng Binang: Korthals (P. Lsub n. 908. 266... 127)
— Res. Bengkoelen —? Kroë pr. d. Wai Goenoeng Kemala
alt. 700 m.: n. bb 7844 (B) - Res. Palembang^250 m.
alt.: W, Greshoff n. 278 (B) — Res. Lampongsche
Distr. — Kalianda pr. Tetaan, 60 m. alt.: n. bb 9322 (L); ?
Telokbetong pr. Redjosari 100 m. alt.: M. J. Dirksen
n. bb 1789 (L, B); ? Kalianda, forest-reserve, way Pirong
in a primary forest: De Witt n. bb 1753 (L, B);? pr.
Tandjong Karang: Endert n. E 1309 {L) ^ Banka
Blinjoe distr. on dry clay-soil: Greshoff n. 37 (B)—
Tandjong Pandan: Teysmann n? (B); Sumatra:
Korthals (L sub n. 898, 200.. .31); in sylvis pr. Gudarim
Baru: Junghuhn (L sub n. 898, 200... 18).

V. n.: (Mal.) toewi (Sum. W. K., Palembang); (Mal.)
toewi batoe (Heyne); (Mal. or Minangkab.) toei (Sum.
W. K.); koedo koedo, koedo koedoe pajo (Simaloer); toewik
(Banka); kajoe lallah, kajoe deling (Karolanden;) (Batak.)
sandoerlangit, sindoerlangit or soendarlangit (Karolanden);
kekapoeng, kapoeng toewi, talas (Lampong. Distr.); djamatan
simaisaloedang (Sum. O.K.); koetang (Atjeh).

Java—]ava (L sub n. 898, 200.. .29, n. 898, 200.. .26,
n. 898, 200... 24); Zollinger Pl. Javan. n. 2229 (P);
Java; Zippel (L sub n. 898, 200.. .30, n. 898. 200.. .25);
Java: v, Hasselt (L sub. n. 898. 200. . .32. n. 898.
200... 33. n. 898, 200.. .34); Java in Mt. Muroh: Waitz
(Lsubn. 898, 200...23): Java: Reinwardt (Lsub n. 898,
200.. .27, n. 898, 200.. .22, n. 898, 200.. .28); Ungarang,
north inclination primary forest 1000—1600 m. alt. (L sub
n. 898, 200.. .20, n. 898, 200... 16); Java(U. sub n. 032851,
n. 032852) — Res. Bantam — G. Poelosari pr. Pandeglang:
Kds. 236 /3 (L); Tjimara distr. pr.. Batoehideung: Kds.
237 /3 (L).
V. n.: (Soend.) ki padali; (Jav.) pedalie.
Res. Batavia — Buitenzorg: Boerlage (L sub n. 908,
352... 629, n. 908, 352... 642);? Cult, in Hort. Bog.

ex Singapore (R, Lobbii Miq.?) L sub n. 898. 200...
42. n. 898, 200... 43, n. 898, 200... 44); Buitenzorg
pr. k. Babatlan, Tjiliwoeng: Hallier (B): Buitenzorg:
Arsin n.? (B); G. Salak: Kds. 24155 ß (B).

Res. Preanger — Garoet pr. Wanaradja: Kds. 235 ß
(L); Palaboeanratoe pr. Soekaboemi: Kds. 232 ß (L). Kds.
34308 ß (B). Kds. 38493 ß (B); G. Gedeh: timber-species
from the Gedeh n. 93 (L); Tjadasmalang pr. Tjidadap,
1000 m. alt.: Winckel n. 1840(U. L). n. 1753/? (L); pr.
Tjiratjap. Djompong Koeion: Kds. 234 ß (L, B); pr. Sang-
grawa: Kds. 233 ß (B); P. Noesagede in the lake of
Pendjaloe, Kds. 276 ß (B); pr. Tjidadap: Tjibeber: R. C.
Bakhuizen v. d. Brink n. 2601 (B); pr. Batoe Karoet-
Tjidadap: W. F. Winckel n. (B); Tjiloeloempang pr.
Tjadasmalang: W. F. Winckel n.. \\26amp; ß (B); Indi-
hiang: C. A. Backer n. 8536 (B).

V. n.: (Soend.) ki padali or ki padali betoel.
Res. Banjoemas — Pringombo, Bandjernagara 700—
900 m. alt.: Kds. 11162 /? (L), Kds. 27096 ß (L), Kds.
34065 ß (L, B), Kds. 37049 ß (L). Kds. 249 ß (B);
Tjiratjap pr. Djampang Koeion, 50 m. alt.: Backer n.
17390 (B); Noesa Kambangan: Kds. 40250 ß (B), Kds.
251 ß (B, L).

Res. Pekalongan — Soerdja distr. G. Prahoe, primary
forest, 800 m. alt.: Kds. 246 ß (L, B), Kds. 247 ß (L);

Res. Kedoe — G. Andong, young secondary forest,
1200 m. alt.: Kds. 36606 ß (B); G. Oengarang. rest of a
primary forest, 1200 m. ah.: Kds. 36633 /3(B); Keditan,
S. of Telamaya: Docters van Leeuwen n. 202 (B).
V. n.: (Jav.) kedali = bedali.

Res. Semarang — G. Telemojo. Oengaran pr. Ambarawa:
Kds. 245 ß (L): Kedoengdjati: Kds. 243 ß (L). Kds.

238 (L. B), Kds. 10127 /? (L. B). Kds. 24989 /? (L. B),
Kdc 239 iS (L). Kds. 240 (B. L). Kds. 10128 fi (L);
Karangasem. teakforest: Kds. 34141 /3 (L). Kds. 244 /?
(B); N. of Wirosari: Kds. 250 (B), in teakforest Kds.
252 (B, L); pr. Sagoong: Docters van Leeuwen
n. 1064 /? (B): W, Moeria. 700 m. alt., dry slope:
Docters van Leeuwen n. 872 (B).

V. n,: (Jav.) bedali, kedali, kajoe poetih, kokok kejok.
Res, Soerakarta - Wonogiri pr. Djanglot, planted by
the natives: H. Burger n. 5614 (B); Wonogiri pr.
forest-complex Eromoko-Pratje. 400 m. alt.: H. Burger
n. 2196 (B).
V. n.: berdali.

Res, Rembang - Ngandang Sedan. 150 m. alt.: Kds.
36387 P (L); Ngandang. G. Boetah: Kds. 36415 P (B);
pr. Getas. Randoe Blatoeng in teakforest: L. Kalshoven
n. 1623 (L. B); Tjabak: Kds. 42353 P (B); G. Pandan

pr.Djatiswara in a teakforest: A. Thorenaar n. 167 (B).
V. n,: sekar petak; (Jav.) dali.

Res, Madioen - Ngebel — Ponorogo. 1275 m. alt.:
Kds. 241 P (L). Kds. 34162 P (L). Kds. 38625 P (L),
Kds. 29175 P (L). Kds. 29174 P (L). Kds. 242 P (L);
G Pandan. pr. Saradan. 500 m. ah.: Kds. 12417 ^ (L);
Patjitan-Toelahan: Backer n. 2969 (B): Lawoe Sido
Ramping. Gandong Valley. 13-1400 m. alt.: Elbert
n 9 (L sub n. 908. 307... 174. n. 908. 307... 173):
Ponorogo: Heringa n. 6527 (B): pr. Djagaraga. 900 m.
alt., G. Lawoe. alang-vegetation and bushes: C. A. B a c k e r
quot; n. 6709 (B): forestcomplex Ngidjo, P. Laosan, forest-distr.

Res, Kediri — Gadoengan-Pare pr. Soekaradja: Kds.
23030 P (L). Kds. n. 22787 P (L).
V. n.: (Jav.) dali, bldali.

38162 /? (L); G. Saower Waderan pr. Loemadjang; Kds.
257 ß (B); pr. Prigen, forest: Dr. A. Rant (B).
V. n,: (Jav.) bedali, gedalL

Res. Besoeki — Rogodjampi pr. Banjoewangi: Kds.
28876 ß (L); Pantjoer-Idjen pr. Sitobondo: Kds. 20525
;3(L). Kds. 28520/5 (L). Kds. 20524 ß (L), Kds. 14900 ß
(L), Kds. 253 ß (L); pr. Pradjekan; Kds. 12838 /? (L. B,
P). 1000m. alt. Kds. 14361 ß (B); Tjoeramanis-forest-
complex pr. Simpolan: Kds. 12705 ß (L, B). Kds.
256 ß (B); Poeger Watangan pr. Djember: Kds. 258 ß
(L), Kds. 254 ß (L). Kds. 12845 ß (L). Kds. 281 ß (L).
Kds. 255 ß (L); G. Idjen pr. Kajoemas: Backer n. 30715
(L); Banderan pr. Bondowoso, dry alang-alang vegetation,
700 m. alt.: C. A. Backer n. 9549 (B).

V. n.: (Mad.) potian, poetian; (Mai.) kar poteh; (Mad.)
sekar poteh; (Jav.) dali, bedali; (Mad.) kadfoe raras.
Bali - G. Pale, 540 m. alt.: Sarip n. 270 (L).
Flores — pr. Larantoeka: Teysmann n. 8816 (B).
Borneo — ? G. Pamatton, type specimen of Miquel\'s
f. borneensis: Korthals (L sub n. 898, 200... 39, n, 898,
200...40, n. 898, 200...41); ? pr. Tewingan, S. E. Borneo
n. bb 382 (B); ? Ben Dajak pr. k. Teroesan, S. E. Borneo,
stem 42 m. high: n. bb 9889 (B); ? Martapoera pr. k.
Dajak Moerai, white flowers, fragrant: n. bb 2490 (B).

Celebes — pr. Tjamba: Teysmann n. H. B. 12757
(B); Menado, primeval forest pr. Kajoewatoe: Kds. 16257
ß (B); Minahassa: Kds. n. 16255 /? (L), Kds. 16254/3(L).

V, n.: kajoe ma ätoes; (Alf. Minah.) koetoe-koetoe, maätas,
mahatoes, wowohan.

18. Radermachera sibuyanensis Elmer — Leaflets
of Philip. Bot. 4 (1912) 1485; Merrill. Enumer. 3 (1923) 447.
Fig. 3g; 12 (18).
An ascending tree, stem 10 m. high, 20 cm. diam., its

main branches arising from below the middle, ultimately
slender; twigs relatively short, few, thick, yellowish gray
lenticellate; whood odorless, very bitter, moderately soft,
reddish-white; bark brown, checked longitudinally, that
on the branches yellowish brown. Leases crowded towards
the ends, ascending, alternate, 2-to 3-pinnate, 20—40 cm. 1.,
petiole properly 10 cm. 1., swollen at the base, terete,
glabrous, the first internode about as long, the second
one shorter and grooved along the upperside; petiolules
of the 2 lateral leaflets 0.5—3 cm. 1., glabrous, grooved,
the terminal one also greatly varying in length but rela-
tively longer; hmbs 10 X 3 cm., the entire margins sub-
involute, flat, ascending, sub-coriaceous, deep green above,
paler green beneath, glabrous, fusiform or gradually ta-
pering from the middle towards both ends, obtuse-acumi-
nate at the iipex, curing dull brown on both surfaces;
the terminal usually the largest, midrib nearly black when
dry, bold beneath, canaliculate above, glabrous, lateral
leaflets from 5—7 pairs, strictly oblique, tips united, very
fine, usually with alternating secondary nerves, extending
from the sub-marginal line towards the midrib, reticu-
lations very few, the total underside fine microscopically
punctated, towards the base mostly with few small scat-
tered glands. Inflorescence terminal or sub-lateral 20 cm. l.,
sparingly paniculate; pedicels few to several cm. I., slender
as are all other stalks; calyx 1.25 cm. 1., glabrous, whitish,
split one half down on one side, the 2 or 3 other lobes
obtuse or rounded, 0.5- cm. across at the base, gradually
expanding; corolla constricted towards the base and striate,
nearly 5 cm. 1., the saccate portion somewhat flattened
and strigosely hairy on the inside, the tubular portion
white, the 5 lobes pink or whitish, deep yellow about the
throat, the lobes broadly rounded and sub-ovately sprea-
ding, 1.5 cm. across, the saccate tubular portion 1.5 cm.
thick; stamens 4 in 2 unequal pairs with a fifth rudiment.

all included in the second portion and inserted upon the
shoulder of the constricted portion; filaments whitish,
curved, sub-terete, the longer ones 2 cm., the shorter
0.5 cm. less in length, glandular-hairy at the base; anthers
divaricate 0.3 cm. 1. laterally dehiscent; ovary 0.3 cm. 1.,
glabrous but giving a grayish impression, sub-cylindric;
at the base surrounded by a glabrous disk, gradually
tapering into the slender glabrous 2.5 cm. 1. style; stigma
dilated into 2 lobes.

Geogr. distr.—Philippines—Sibuyan^Mt. Giting-Giting,
Magellanes, Prov. of Capiz: Elmer n. 12060 (L,B).

This endemic species of Sibuyan is rather well distin-
guished by its leaflets; it is very closely allied with R. pala-
wanensis. It occurs on fertile wooded banks pf the Patoo-
river at 170 m. alt.

19. Radermachera Elmeri Merrill. Govt. Lab. Pubhc.
Philip. 29 (1905) 48; idem. Philip. Journ. Sc. Bot. 3(1908)
334; idem. Enumer. 3 (1923) 445.

A small tree about 6 m. high, with 2-pinnate glabrous
leaves, and pendent paniculate inflorescence, the pink flowers
ca. 6 cm. 1., the branchlets, axis of the inflorescence and
the leaf-rhachis glabrous, usually with many small white
excrescences. Leaves about 35 cm. 1., the leaflets elliptical-
lanceolate, glabrous, long acuminate, the base acute at
the underside with small glands, 6—10 X 3—5.5 cm.,
nerves ca. straight, about 10 pairs, not prominent, the
reticulations lax, obscure; petiolules 0.5—1 cm. 1., glabrous.
Panicles about 30 cm. 1., the branches spreading, 5—6 cm.
1., dichotomously branching, few-flowered; pedicels slender,
about 1.5 cm. 1.; calyx campanulate, glabrous 1.5—1.8 cm.
1., 3 —4-toothed, the teeth broadly ovate, acute or sub-
obtuse or sometimes rounded 0.4—0.5 cm. 1., 0.5—1 cm.

wide. Corolla 6 cm. 1., pink, sub-campanulate the portion
included in the calyx tubular, 1 cm. 1.. then dilated, the
mouth 5—6 cm. wide, entirely glabrous except near the
insertions of the stamens where it is more or less fer-
rugineous pubescent inside, the margins minutely ciliate,
5-lobed, the lobes rounded, about 2 cm. wide, 1.5 cm. 1.,
sub-equal. Stamens inserted in the tube near the top of
the tubular portion of the corolla, the filaments about
2 cm. 1., glabrous above, more or less ferrugineous pubes-
cent below with capitate hairs. Ovary glabrous, narrow
lanceolate, the disk annular, thickened, the style glabrous
2.5 cm. 1.; ovules oo in 6 rows. Capsule (immature) cylindrical,
ca. 40 X 0.4 cm., glabrous, seeds including the wings
0.8 X 0.2 cm.

Geogr. distr. Philippines—Luzon—Prov. of Benguet pr.
Sablan: Elmer n. 6179(D); Prov. of Benguet: M. Ramos
n. B. S. 5370 (L); pr. Baguio, prov. of Benguet: S. La ray a
n. F. B. 27954 (D.L); idem: Elmer n. 8940 (L.B) -
B o n t o c — Mt. Masapilid: Ramos and E d a n o n. B. S.
37903 (D); Prov. of Ilocos: Paraiso n. F. B. 25454 (B).

According to Merrill this Luzonan species occurs in
forested ravines ca. 400—1600 m. alt. It is closely allied
with R. acuminata and R. fragrans. As to the former it
differs by its dense lenticellate branchlets. petioles and
inflorescences, its smaller leaflets, its rather pauciflorous in-
florescences and its larger flowers. As to the latter it
differs by its not fragrant, pink flowers, the different ner-
vature of the leaflets, the little height of the tree, and
the leaflets, which are not crowded towards the ends of
the ultimate branchlets. Further the inflorescence is pendent
and the leaves are ordinarily shorter.

The Palawan specimens, Merrill gives in his revision,
differ from the type as he remarks. First they possess some-
what smaller flowers. described by Foxworthy as white

and fragrant. He arranged them in the var. fragrans, as
he could not find any valid characters to warrant the
separation as distinct species in the material, he had at hand.

I have found however some additional differences which
might justify the separation.

20. Radermachcra fragrans (Elm,) vSts. nov. spec. —
Radermachera Elmeri Merrill var. fragrans Elmer
Leafl. of Philip. Bot. 3 (1915) 2561 ; Merrill Enumer. 3
(1923) 446.

Fig. 3i: 12 (20)
Sub-erect tree, stem ascending 45 cm. diam., 12 m. high
or higher, crooked; main branches erect, arising from above
the middle, ultimately rebranched and forming dense masses,
bark quite thick, brown on the stem, gray on the branches,
the epidermis flaking into small pieces: branchlets yellowish
and brown lenticellate, the young apical portions green and
glabrous. Leaves opposite ascending, ternately compound,
all the stalks green except the swollen purple joints,
varying from 30—80 cm. 1., leaflets flat, smooth and
glabrous, sub-coriaceous, much darker green on the upper
side, grayish brown on both sides when dry, variable in
size from 6—15 X 2.5—5 cm., elliptic, but with abruptly
caudate-acuminate and strongly recurved points, base obtuse
to cuneate, entire; petiolules 1 cm. 1. or considerably shorter,
canaliculate, appearing articulate or glandularly swollen at
their distal ends, the terminal one, when solitary much
longer; midrib sunken along the upper side, rather con-
spicuous beneath: lateral nerves 6—8 on each side of the
midrib, curved, divaricate faint tips reticulately united,
reticulations evident under a hand lens; thyrse as long as
the leaves, terminal, ascending, glabrous, green except the
sweetly fragrant flowers; pedicels 1 cm. 1. and usually
provided towards the base by one or more very minute
bracts; calyx stipitate towards the green base and falsely
articulate there, whitish above the middle, 2 cm. I. and

1.5 cm. br. across the irregularly short lobed top, turbinate
greenish towards the base, where it seems to separate
transversely when mature, the lobed portion thinner in
texture and whitish; corolla 5—6 cm. 1., glabrous except
the minutely puberulent upper portion, 5 cm. wide across
the throat, the basal 1 cm. constricted and 0.6 cm. wide,
otherwise campanulate and finely veined, pendent, caducous,
light creamy white except the yellowish red inside, some-
what compressed; the 5 lobes on the average 1.5cm. 1.,
and as wide across the base, broadly rounded at the apex;
stamens 4 included of 2 slightly unequal parts; filaments
nearly white, inserted upon the throat of the constricted
portion of the corolla or 1 cm. from the base, the shorter
pair at least 2 cm. 1., the longer pair almost 1 cm. longer,
gracefully curved towards each other, all towards the base
flattened and glandularly hairy along the upper side, other-
wise slender and glabrous; anthers similar to a spear head
when in anthesis, brownish tinged, 0.4 cm. 1., the apical point
blunt and membranous, attached at the distal angle of the
divergent cells, laterally dehiscent; ovary green.

Geogr. distr. Philippines ^ Palawan ^ Prov. of Palawan,
Brooks Point: Elmer n. 12681 (U,L,D,B); Palawan: A.
M anal on F. B. 5190 (D); Palawan: Foxworthy n.
B. S. 584 (D).

V.n.: agtap {Tagh.); barangau-a-nolobaga{l\\k.); pamaya-
bayen (Ilk.); sayo (Ig.); tantangan (Tagb.).

The main differences with the nearly allied preceding
species are: higher tree, the leaves crowded towards the
ends of the ultimate branchlets, white fragrant flowers,
the leaflets usually narrower and caudate, recurved at the
top, with curved side-nerves, inflorescence not pendently
elongated. It occurs in forests along streams and swamps
at low altitudes, and along the sea-shore.

P. Dop 5has made an interesting study on this genus,
whilst he was» working out the Indo-Chinese Bignoniaceae,
He discusses the false septum of the capsule, which he
mentions as the main character in dividing the genus
Heterophragma into Heterophragma and Haplophragma.

For the same reason Clarke (Hook. Fl. Br. Ind. 4. 379)
joined Markhamia and Dolichandrone, both possessing a
false septum. The main problem is, what is the systematical
value of that septum. Sprague (Kew Bull. 1919. 302)

says, that whereas Markhamia and Dolichandrone both
possess a false septum, more external characters exist,
which oblige him to separate these genera. In both the
corolla-tube is more or less funnel-shaped; in Dolichandrone,
however, the lower and cylindric part of the tube is
greatly developed and much exceeds the calyx and the
limb is almost actinomorphic; whereas in Markhamia the
cylindric part of the tube is very short and concealed in
the calyx, only the upper part of the funnel being visible,
and the limb is conspiciously bilabiate. In addition Doli-
chandrone has pure white, fragrant nocturnal flowers, whilst

those of Markhamia are yellow, pink or lilac, etc. and
expand in the day-time. It is probable that Dolichandrone
has a pollination by moths. Thus Sprague.

These external morphological characters and also the
biological ones seem to me indeed sufficient and of principal
rank together to justify the separation Sprague undertook.

Now about Heterophragma (Fig. 13c). This genus is
founded on H, Roxburghii D. C., an Indian species. The
others have been H. sulfureum Kurz, H. vestitum Dop and
H, adenophylla Seem.

Dop remarks in general as to the classification of the
Bignoniaceae-genera : „ces genres sont classés par la plupart
des auteurs... d\'après le calyce. Ce caractère ne m\'a pas
paru tenir compte des affinités réelles, qui ne peuvent être
basées que sur la structure d\'un organe plus central, l\'ovaire
et le fruit qui en dérive.quot;

I cannot agree wit this remark. Seemann, Bentham,
Hooker. Kurz, Clarke and Schumann indeed sup-
posed several external characters, such as the calyx, etc.
of importance, which they have expressed in their keys,
but at the same time they used ovary, etc. as well for the
great divisions as for generic separations. Radermachera
and Stereospermum ate principally separated by the fruit.
That they are treated as sub-genera in the Genera Plantarum
is explained by the fact that Bentham and Hooker
usually considered more or less large generic diagnosis,
whereas nowadays mosdy much more narrow ones are
distinguished. It is therefore that they treated the concer-
ning groups as genera, certainly a matter of conception.
The differences between the two are not altered. For the
rest agreeing with Sprague, I also think that one should
have a general outlook at the external morphological and
biological characters. Besides D o p in- his key of the Indo-
Chinese genera also makes use of the external characters.
So e.g. in Ferdinandia and Haplophragma. the first possessing

a cupuliformous disk and axillar inflorescences, the latter
having an annular disk and terminal thyrses.

With the foundation of Haplophragma I fully agree.
Clarke gives in the diagnosis oiHeterophragma: „septum
flat or 4-angular.quot; The 4-angular septa are found in the
type H, Roxhurghii and other species; a flat septum is
only present inH, adenophyllum. In this respect Schumann
made a mistake in his key (Pfl. Fam, IV. 3b. 229).
Boerlage (Handb. Fl. Ned. Ind. 2 (1899) 596) already
notes that it seems rational to him to separate Het. adeno-
phylla Seem, from the other Heterophragma\'s.

Now there is an other Sumatranan species, described
by Miquel as Spathodea macroloba, which Sprague
(Kew Bull. (1919 305) calls synonyms to Dolichandrnone
spathacea. I am sure he has not seen the original specimen
of Miquel. Indeed it is the second species which is to
be arranged in Haplophragma. The septum is as flat as
that of Hapl. adenophylla, of a shining yellow, slightly
bubbled, the seeds having the same structure, that is to
say they are also flat, have a shining yellow relatively
large germ and narrow membranous wings. The capsule
of Hapl. adenophylla is linear, strongly curved and ribbed,
that of Hapl. macroloba is slightly smooth and sub-straight.

The inflorescenses however differ totally; that of Hapl.
adenophylla being very stout and terminal, whereas Hapl.
macroloba possesses small lateral ones.

As to the affinities of Haplophragma. Dop indicates the
African genus Ferdinandia, which some authors suppose
to be synonymous with Heterophragma. As I have mentioned
above, Dop marks the difference between Haplophragma
and Ferdinandia, the former possessing terminal inflores-
cences and an annular disk, the latter showing axillar
inflorescences and a cupuliformous disk. Whereas I have
already said that Hapl. macroloba is undoubtedly a Ha-
plophragma and has axillar inflorescences; the only difference

remaining is the form of the disk, certainly not an impor-
tant point. As I did not study any African species I am
not competent to judge in this question.

It is a remarkable fact that Haplophragma macroloba
reminds of Radermachera xylocarpa. The thyrses show a
striking resemblance, and the capsule is much like that of
H. adenophylla. It is rough with tubercules, curved or
sub-straight, and much broader than that of most of the
Radermachera\'s, which possess a linear, terete, smooth
capsule. The septum, however, is the terete one, as usual
in Radermachera, though the structure is the same, it is
bubbled (caused by the press of the seeds), shining, yellow
and corky like that of Haplophragma, but terete. The
seeds of R. xylocarpa are also differing from those of
Haplophragma, are broader, membranously winged, and
thinner than those of Haplophragma. I suppose Rader-
machera and Haplophragma to be allied with each other,
especially with respect to R. xylocarpa, the latter being
more or less isolated in Radermachera.

Descr. emend. Arbores, folia magna opposita, 1-pinnata.
Thyrsi terminales vel axillares, multiflori, flavescenti vel
fuscescenti, pubescentes vel tomentosi. Calyx campanulatus,
irregulariter 3-4-5-lobatus. Corolla campanulata-infundibu-
liforma, tubo angusto apicem versus dilatata, extus pubes-
cente, lobis 5, sub-aequalibus, undulatis vel crenulatis.
Stamina 4, didynamia sub-exserta, antheris divaricatis,
quinto rudimentario. Discus annularis. Capsula magna,
cylindrica, leviter vel valde curvata, costata vel laevis,
bilocularis, septum planum, latius, pseudo-septum abest.
Ovula 00, multi-seriata in placentis pro loculo 2, margina-
libus. Semina oo, applanata, anguste membranaceo-alata.

Inflorescences lateral rather small, pubescent when young,
capsule sub-straight, smooth, lenticellate.. H, macroloba.

1. Haplophragma macroloba (Miq.) vSts. nom. nov.
— = Heterophragma macroloba (Miq.) Backer in Herb.
Bog. Nomen — = Spathodea macroloba Miq. Fl. Ned.
Ind. Eerste Bijvoegsel (1860) 565; Heyne Nuttige Plant.
Ned. Ind. 4 (1917) 167, 2 (1927) 1371; Cordes Tijdschr.
Ind. Mij. van Nijverh. Ö Landb. 14. 193.

Descr. emend. Arbor procera 25—40 m. alta parte in-
feriore ca. 15—20 m. ebranchiata; rami teretes, cortice
tenuissimo glabro griseo tenuiter costulato demum ramas
agens longitudinales; folia opposita imparipinnata, 3—5-
jugata 30—40 cm. longa, petiolus ca. 5—9 cm. 1. supra
plana vel subconcava margine acuto; foliola integerrima
subsessilia vel 0.3—0.5 cm. longe petiolulata obovata-
oblonga vel oblanceolato-oblonga apice breviter vel medio-
criter acuminata basi cuneata 14—18 x 5.5—6.5 cm., juve-
nulia parviora, nervis lateralibus 8—11; jugis inferioribus
minoribus, foliolis ca. 4 cm. longis ovatis vel ovato-orbi-
cularibus; inflorescentiae laterales (an semper?) in ramis
junioribus, cymis in racemos dispositis, parvae compactae
pauciflorae 4—7 cm. longae 3—4 cm. latae fusco-tomen-
tosae, pedunculis brevibus 0.5—1 cm. longis, pedicellis ca.
0.5 cm. 1.; calyx campanulatus extus sparse pubescens
(lobis incl.) ca. 1.2 cm. 1., 2-lobatus, lobis ovatis. altero
bifido, altero bilobo. apice ciliatis. Corolla lutea ca. 3—
3.5 cm. 1. hypocraterimorpha tubo angusto. extus sparse
pubescente. limbo 5-lobato. lobis fere glabris crispus;
stamina 4 fertilia quinto sterile, fere 0.75 cm. supra basin

Fig. H. Haplophragma P. Dop. H. macroloba (Miq.) vSts. a. Type
specimen of Miquel (U). b. Flowering specimen (B. cult, sub n. XI c.
106a). c. Part of the septum (U). d. Seed. (U) H. adenophylla (Wall.)
P. Dop. e. and f. Seeds out of the same capsule (Wallich n. 298).
Magnif. X.

corollae inaequaliter inserta; antherae biloculares, divari-
catae; ovarium lanceolatum dense luteo-tomentosum bilo-
culare, ovulis in utroque loculo circ. 6-seriatis; discus
hypogynus annuliformis, minutissime pubescens; Capsula
linearis, magna. 50-60 X 1.4 cm., teres arcuata sufflava
lenticellis lineari-lanceolatis dense obtecta. pedicellus brevis
3 cm. longus 0.4—0.5 cm. crassus. bivalvus; valvis sub-
coriaceis septo coriaceo-suberoso ca. 0.8 cm. lato, sufflavo.
nitido. iniquo; semina sufflava. nitida. cotyledonibus con-
vexis ca. 0.8 cm. longa. 2 cm. lata ala tenuiter membra-
nacea 0.2 cm. lata inclusa.

(Description founded on herbarium-materials (branches,
leaves and capsules) H. B. n. 2353 (Type specimen of
Miquel) and the inflorescences and flowers of the cultivated

Geogr, distr, Sumatra-Gouvt. Atjeh-Uio Nga pr. Glé
Tjoemo: Beumée n. bb. 7368 (L); Afd. Tamiang pr. k.
Ppropnoek: boekoen n. bb. 9797 (B); P. Bras pr. Oelepaya
Kds. 10523/? (B). Kds. 10524/3 (B), Kds. 10525/3 (B).
V.n.: thoee (Beumée); toewi (Afd. Tamiang: Soekoen).
Res, rapanoe/i—Onder-afd. Angkola and Sipirok pr. k.
Saromatingi: n. bb. 3141 (L); idem pr. k. Sajoermatinggi:
Taris n. bb, 6449 (B).
V.n.: (Batak) radja.

Res. Sumatra\'s Westkust—Ptiaman: Diepenhorst n.
H. B. 2353 (Type specimen of M i q u e 1) (U.B); Padangsche
Bovenlanden pr. k. Sidjoendjoeng: Kds. n. 10276/3 (B);
Kds. 10657p^ (B); Karolanden pr. k. Marteloe: L. P.
Thijssen n. bb. 9738 (B); pr. k. Sidjoendjoeng Moedra:

V. n.: (Mai.) soengki, soengkai, soengkè; (Mal.) soengkè
tjirit; (Mal.) soengkai rimbo; (Batak.) toehi (Karolanden).

cribed the inflorescences and the flowers of the flowering
specimens cultivated in the Botanical Gardens at Buiten-
zorg.

The tall Bignonian tree is limited in distribution to the
primary and secondary (?) forests at low and medium alti-
tudes up to 350 m. in N. W. Sumatra; at the east-coast
it does not seem to appear. The frequency within its area
is never rare, although mostly scattered, sometimes occur-
ring sub-socially.

The wood produces an excellent timber; so e.g. it is
used by the natives for making proa\'s; they call it (Mai.)
soengkè, meaning the quot;large and best speciesquot;, owing to
its height and other qualities.

In Buitenzorg C. A. Backer stated.\'it as being a i/efero-
phragma, but never published this; Boer lag e in his
Handleiding does not mention it.

2. Haplophragma adenophylla (Wall.) P. Dop. — =
Bignonia adenophylla Wallich Cat. 6502 P. Dop Contrib.
à l\'étude des Bignon. Bull. Soc. Bot. Fr. 72 (1925) 887;
— = Spathodea adenophylla A.D.C. 9 (1845) 206; Wight
111. Ind. Bot. (1840) 160; Kurz For. Fl. 2.236; - =
Heterophragma adenophylla Seem. Journ. Bot.; Benth. amp;
Hook. Gen. PI. 2 (1876) 1047; C. B. Clarke in Hook. f.
Fl. Br. Ind. 4 (1885) 381; Foxworthy Indo-Malayan
Woods. Philip. Journ. Sc. 4 (1909) 558; Ridley Fl. Malay
Renins. 2 (1923) 551.

Tree ca. 10—20 m. high, innovations tomentose. Branch-
lets ferrugineous tomentose, leaves and inflorescences stout.
Leaves opposite 30—45 cm. I., 1-pinnate, 2—3-jugate;
petiole ca. 4 cm. 1. Leaflets entire, obovate. obtuse, acute
or even acuminate 17—20 X 12—15 cm., brownish pubes-
cent beneath, glabrous above except on the 9—11 pairs
of nerves, lateral leaflets sub-sessile, rhachis stout, sub-terete,
striate, brdwn-tomentose. Leaves immediately under the

inflorescence single, entire, sessile, sub-orbicular ca. 4 cm.
diam. Ca/yx ventricose-campanulate, rusty-tomentose, 2—2.5
X 1.5cm. (incl. lob.); lobes ovale, subequal ca. 0.6cm. 1..
Thyrse stout, lax-flowered, totally ferrugineous tomentose,
erect, terminal; peduncle ca. 8 cm. 1., flowering part ca.
10—12 cm. 1.; cymes 3—7-flowered, bracts narrowly lan-
ceolate or triangular, decideous, ca. 1 cm. 1.. Corolla brown-
yellow, darkest inside, densely woolly-tomentose, when
expanded, ca. 5 cm. diam. at the mouth, lobes hardly
crispate or crenate. Anther-cells nearly separate, pendulous.
Capsule very large sub-terete 50—100 x 2.5 cm., cylindric.
cork-screw-like, on a 0.7 cm. thick pedicel, not hairy,
strongly curved, rough with Ienticels and with several strong
ribs on each valve. Septum flat, shining yellow, slightly
bubbled, ca. 1.4 cm. broad with corky opaque margin
(placenta) ca. 0.15 cm. broad. Seeds 0.8—1.2 x 3.7 cm.

Geogr. distr. India — from Assam and E. Bengal to
Tenasserim and Andamans frequent (C. B. Clarke); Ava
(C. B. Clarke); Ripoe Irawadi: Wallich n. 6502a. type
specimen? (P); Assam: Jenkins (P); Tenasserim and
Andamans: Herb, of the latest E. I. Comp. ex herb. H e 1 fe r
n. 4067 (P,U);?: Gaudichaud n. 289 (P).

Malay Peninsula — Open country in the north, Kedah,
Alor Sta, Perlis, Chupeng (Ridley).

Malayan Archipelago — Cult, in Hort. Bog. sub. XI. 1.
7 (B); Cult. Hort. Bog. sub. n. 6603 (B).

The tree is only endemic on the continent of E. Asia
and does not occur in the Malayan Archipelago. In the
Malay Peninsula it is found only in the northern parts
and so it is not probable the species will be later on
found in Sumatra or elsewhere in the Mai. Archipelago,

Bibl. Univ. Genève 17 (1838): Prodr. 9 (1845) 244:
Bureau Monogr. (1854) 57; Miers Transact. Linn. Soc. 26
(1868) 166; Bâillon Bull. Soc. Linn. Paris. 693; Seemann
Transact. Linn. Soc. 28. 17. ; Benth. amp; Hook. Gen. PI. 2
(1875) 1051; Bâillon Hist. PI. 10 (1891) 55 ; K. Schumann
Engler-Prand Pfl. Fam. IV. 3b (1895) 247.

Parmentiera ccrifcra Seem, in Bot. Voy. Herald. 182.
t. 32; K. Schumann in Engler-Prantl. Pfl. Fam. IV. 3b
(1895) 247. fig. 94; Koorders Exc. Fl. 3 (1912) 185.

Shrub or shrubbish tree ca. 3 m., leaves opposite 3-fo-
liolate, 7—8 cm. 1. with a petiole 2.5—3 cm. a little ca.
0.1 cm. br. alate, cuneate to the base ; folioles sessile ovate
mostly shortly acuminate on the elder branches the leaves
are fasciculated by 2 or 3 on lumplike short branches as
in Crescentia, The flowers are cauliflorous; the calyx ca.
2.5 cm. is closed in bud and splits open afterwards as in
a spatha on one side. The corolla ca. 5 cm. is ± regular
campanulate-funnel-shaped, the 5 fertile stamens reaching the
throat; disk indistinct. The ovary is 1-celled or at base
2-celled with many ovules on each placenta in several
rows. The fruit on the old wood is very remarkable
(name!) 0.5—1 ra., ca. straight, hnear-cylindric.

Geogr. distr. Malayan Archipelago. Java — Res. Batavia —
Meester Cornelis: Weehuizen cult, in garden (B); Cult,
in Hort. Bog.: Hallier f. (L. sub n. 910, 192...346)-
Res. Cheribon — Indramajoe cult, on a churchyard:
B a c k e r n. 16679 (B) — Res. Soerakarta — Solo : H e m k e n
cult, in garden (B).

This native of Panama is not so generally cultivated
in the Malayan Archipelago as for instance Crescentia;
so in Java, according to Koorders, the quot;candle-treequot; is

sometimes found in gardens as an ornamental shrub. I
also saw specimens from Africa, but nothing, nor literature
about the cultivation in the Philippines and in S. E. Asia.

Koorders 1) made a research upon the buds, which
contain water, a secretion of hydathodes.

Perhaps it would be better to refer also Crescentia
alata H, B. K. 2) 3) to this genus, though K. Schumann
arranges it in Crescentia, Urban\'\') states the differences
of the pollen between Crescentia and Parmentiera,

Gen. n. 762; D. C. Prod. 9 (1845) 246; Miers Trans-
act. Linn. Soc. 26 (1868) 167, t. 7-9: Bureau Monogr.
(1864) 55; Benth. amp; Hook. Gen. PI. 2 (1876) 1053:
Bâillon. Hist. PI. 10 (1891) 17, 54; Schumann in Engler-
Prand Pfl. Fam. IV. 3b. (1895) 248 : Urban Ber. Deutsche

This genus is endemic in the neo-tropical region, but
several species are for centuries ago introduced in the
whole tropical region and cultivated there for ornamental
purposes.

K. Schumann arranges in this genus also C. alata
H.B.K. which Miers thinks to.be a Parmentiera, on
account of its 3-foliolate leaves and the alate rhachis.
Urban is of the same opinion because of the pollen,
which seems to be differing; C, Cujete L. has quot;furchen-
losen sehr feinnetzigen Pollenquot; whereas C. alata H.B,K,
possesses quot;3-furchigen Pollen mit schön netziger Exinequot;.
As I did not study many species of both genera. I agree
for the present with K. Schumann in this point and
arrange C. alata H.B.K. in Crescentia.

gt;) S. H. K o o r d c r s. Ubcr die Blüthcnknospcn-Hydnthodcn clnigcr
tropischen Pflanzen. Ann. Jard. Buitenzorg M (1897) 354-469.

1. Crescentia Cujete Linn. Sp. PI. (1753) 626; La-
marck. 111. t. 547; Seemann Journ. Bot. 6 (1854) 269,
275; Seemann Transact. Linn. Soc. 23 (1860) 20; Miers
Op. cit. 26 (1868) 167; Bot. Mag. t. 3430; Bâillon Hist.
Pl. 10 (1891) t. 44, 45; Schumann in Engler-Prantl, Pfl.
Fam. IV. 3b (1895) fig. 93 E; Boerlage Handl. Fl. Ned.
Ind. 2 (1899) 597; Merrill Fl. Manila (1912) 430; Koorders
Exc. Fl. 3 (1912) 185; Merrill Enum. 3 (1923) 447 ; Heyne
Nuttige Pl. Ned. Ind. 4 (1917) 167.

This small tree attaining ca. 8 m. has simple, scattered,
lanceolate-spatula-shaped ca. 10—20 cm. 1. leaves, mostly
obtuse, but often shortly apiculate or acuminate at the
top and always cuneate at the base, in sicco showing
sometimes a metal-luminous surface; they are fasciculated
on very short lumplike side-branches. The flowers ca. 5
cm. 1. are few and placed on these short branches or
or they are cauliflorus on thin or thick other branches.
The calyx ca. 1 cm. 1. remains long closed in bud and
splits irregularly. The. corolla is ventricose-campanulate
with a transversal plait at the front, and is ivory towards
lightgreen with purple inside. There are 4 stamens and a
staminodium, the ovary has many ovules on 2 broad 2-
lobed placentas: the fruit is head-shaped, the seeds are
imbedded in a juicy fleshy hollow.

Geogr. distr. Sumatra. — Gouvt. Lampongsche Distr. —
P. Sebesi: Docters van Leeuwen n. 5427 (B); pr.
Monggala : G u s d o r f n. 27 (B) — Gouvt. Palembang — pr.
Moearodoa: Greshoff n. 431 (B).

Java — Res. Batavia — Buitenzorg pr. Kotoparis : B a k h.
V. d. Brink, f. n. 1621 (B); Kritëk pr. Tji Omas: Bakh.

V. d. Brink, f. n. 522 (B); Weltevreden: Backer (B);
Buitenzorg on the. way Gardoe: Hallier f. n. 245 (B);
Buitenzorg: Boerlage (L sub n. 908, 352.. .616); Cult.
Hort. Bog.: Boerlage? (L sub n. 908, 337,. .868); Cult.
Hort. Bog. XI. H. 25 (L sub n. 922, 66... 449) - Res.
Preanger — pr. Tjitjalenka: W i s s e n. 869 (B) — Res. Soera-
karta — pr. Tlawa: Kds. 209/3 (L); pr. Passangrahan:
Kds. 209/? (B).

V. n.: sikadel (Jav.); sepokal (Jav.); bila halandha (Mad.);
kalebasboom; bernoek (Soend.).

Kangean Arch. Kangean: Backer n. 27022 (B);
P. Kangean pr. Ardjasi: Backer n. 27363 (B).
V. n.: bila rhadja.

Ambon: — Amboina: B. Robinson n. 1780 (L); P.
Kelang pr. Ceram: Kor nasi n. 1328 (L).
V. n,: kalabassa.
Philippines — cf. Merrill Enum.

The calabash-tree is recently introduced in the Malayan
Archipelago and the Philippines and Is now for ornamental
purposes often cultivated; e. g. very often in Java (Koorders).
It is also used as a quot;pagerquot;-plant (Heyne). However, it
seems that nevertheless this species does not run wild. The
woody shell of the fruit, the true quot;calabashquot;, is used by
the natives.

Bur man -) described another species Crescentia ovata
Burm. from Java, which seems to be identical with Cres-
centia cucurbitina Linn.. I have not seen it among the

\') Uber die niüthenknospcn-Hydathoden einiger tropischen Pflanzen.
Ann. lard. Buitenzorg. 14 (1897) 354-469.
2) Burmannus Fl. Ind. 132: D. C. Prod. 9 (1845) 247,

herbarium-materials I studied. It seems that C cucurbitina
is no longer cultivated in the Malayan region.

2. Crescentia alata H.B. K. Nov. Gen. Sp. 3 (1818)
\\52gt;\'. - Crescentia trifolia Blanco Fl. Filip. (1837) 489, ed 2
(1845) 343, ed. 3.2. (1878) 271, t. 327; D. C. Prod. 9 (1845)
247: Miquel Fl. Ned. Ind. 2 (1856) 759; - = Otophora
paradoxa Blume Rumphia. 3 (1847) 146; Seemann Journ.
Bot. 6 (1854) 269, 275; ibid. Transact. Linn.Soc. 23 (1860)
21; F. Vill. Novis. App. (1880) 151; Vidal Sinopsis. Atlas
(1883)\' 35, t. 73. fig. C; Merrill Fl. Manila. (1912) 430;
Merrill Philip Journ. Sc. Bot. 9 (1914) .141 ; ibid. Spec.
Blancoanae (1918) 350; ibid. Enum. 3 (1923) 447.

This small tree has 3-foliolate opposite ca. 10—12 cm.
long leaves, with a long 6—7 cm., broad ca. 0.2-0.3 cm.
winged petiole, single or fasciculated on lumps as in Parmen-
tiera cerifera Seem., The folioles are long-spatula-shaped
to narrow oblanceolate. The flowers are cauhflorous, great
ca. 7 cm., broad campanulate; the coriaceous ca/yx splits
into 2 lobes.

Geogr, distr, Philippines—Luzon—RizaX prov. pr. Malaban:
Merrill, spec. Blancoanae n. 515 (L); Manila (type spec,
of Otophora? paradoxa Bl)-, Herb. Perrottet n. 95
(L, P).

South America — pr. Grenada, Plantae Nicaraguyenses :
P. Levy n. 153 (P), n. 1383 (P); pr. Acapuleo, Am.
equator.: M. A. Bonpland n. 3858 (P).

According to Merrill this West-Mexican native was
introduced into Guam and into the Philippines by the
Spaniards and still persists in cultivation in both, although
now very rare in the Philippines.

Whether C. alata was naturalised in early times in the
Philippines and has later on disappeared, I cannot tell,
but in relation to C. Cujete this is not probable.

According to MerrilP) it is probably of comperatively
recent introduction in Amboina.

The main problem of plant-taxonomy is still the under-
standing of the ways along which evolution takes place,
that is to say, the rational relation between a new origi-
nated species and its immediate ancestor. At least if one
is convinced, as the present author is, that the acceptance of
the evolution-theory is necessary for the modern student
of biology with respect to the state of science nowadays.

The first is that evolution is determined by adaptation to
external factors, whatever qualities these may be. (Inor-
ganic ones as well as organic (biological) ones). This
standpoint does not include the unscientific telcological
hypothesis.

The second opinion is that of a not-tendenced unfolding
in many directions, not influenced by external factors. One
may think of a law, inherent in the organism itself,
eventually caused by the structure of the protoplasm, the
main substance of all organic life. But this law is not
necessary for the acceptance of the opinion.

It is very difficult, perhaps impossible, to say which of
the two opinions is the right one. Indeed, the second
opinion combined with the rational thesis of the quot;survival
of the fittestquot; has the same result as the first opinion,
namely that quot;all organisms are in agreement with their
neighbourhoodquot; or at least are quot;not in contradictionquot; with
it, which seems to be a fact a priori.

It cannot be the intention of the present author to
bring this fundamental problem nearer to its explanation,
though he believes that the second opinion will appear
to be the right one in the future.

He intends with this treatise to throw some light upon
the problem of the origin, geographical distribution and
affinities of the Malagan Bignoniaceae,

Manner of spreading in Bignoniaceae. This manner
is a very characteristic one in the family. The bulk of the
species possess more or less broad membranously winged
seeds, often with a small germ, representing a more or
less perfect structure for wind-spreading. Notwithstanding
this they show the same general characters of geographical
distribution \') of many families which do not possess
seeds which are easily removed by means of the wind.
Thus they possess many genera with limited areas and
many endemic species. The morphological structure of the
seeds (or fruits) thus seems not to be the prime cause,
determining the distribution as one would think at first
sight and this character is not as profitable for the plants
as it ought to be theoretically. J. C. Willis quot;) has
been the first to emphasize this. Indeed we must accept
that the enlarging of the area per year is rather small.
To be sure each individual takes a part in the surroun-
ding associations sens. lat. and it will be very difficult for
its descendants to get place in these authochthonous
vegetation.

The first difficulty met with is the concurrence, with
which I indicate the complex of difficulties to which a
seed is exposed before it has grown up a fruiting indi-
vidual (germination-conditions, quickness of growth, life-
form. etc.). Further its dependence on other external
influences, organic ones. e.g. dependence on pollination

I fully agree with Willis, when be ascribes the
opinion that plant-spreading should be unlimited as to its
rapidity, to the attention that was paid on the fabulous
rapidity with which adventiva enlarge their area through-
out the world.

It is probable that plants in natural conditions, that is
to say individuals of an authochthonous vegetation but
slowly enlarge their area. Further that different plants
possess a different unit of enlarging the area (enlargement
of the radius per year). Willis\'), however, supposes that
allied species on the average possess nearly the same unit
of enlarging the area. The morphological characters
of spreading will be the same on the average, he
says. I do not accept this supposition. The possibilities
for spreading may be variable, even with allied
species of the same genus. This is the case e.g. with
Dolichandrone (Fenzl) Seem,. D, spathacea (L, f,) K, Sch, is
a litoral species, which spreads by means of sea-drifts and
possesses thick corky wings in this respect. The other
Dolichandrone\'s spread by means of wind. Certainly there
are more examples to be found in this respect.

Moreover, allied species often are in different respects
dependent on climatical or edaphical factors; perhaps they
have diiferent germination-conditions or are in other respects
dependent from the surrounding nature.

It seems to me that the hypothesis, allied species should
have — probably the same unit of enlarging their
areas has not been established as yet, and is by no means
probable.

Further one should have an eye upon accident, which
plays an unknown part in our knowledge about the possi-

For the rest there exists that strange phenomenon of
becoming extinct, and on the other hand the sudden
enlargement of the area.

A few examples may illustrate this. It is a pity — but
it seems rational — that we know only few examples
that have taken place in recent times.

How is it that Trapa nutans L. has disappeared in the
course of a few centuries out of Western Europe, except
in some isolated spots?

How is it that the black rat {Mus rattus L.) disappears
in Europe, now nearly ousted by the common or brown
rat {Mus decumanus Pali), which has migrated from Russia
into Europe?

It is the same case with the crested lark {Galerida cri-
stata viarum Br.), which has migrated from Russia into
Europe in a hundred years or less.

These examples of recent date cannot be explained as due to
human influence, because human influence in Europe has been
acting rather intensely already almost during ten centuries.

Endemism. As to this there are two cases: firstly
one may distinguish relic-endemism and secondly there
exist progressive endemics.

Relic-endemics are species, which have possessed a larger
area in earlier times and are found nowadays in rather
isolated small area\'s, (e.g. Taxodium, Sequoia, Ginkgo).

Relic-finding-places we call the spots where a species was
able to maintain itself. {Trapa nutans L. and Betulanana L.
in W. Europe). Relics are mostly characterized by
continuous, sometimes very small areas or by disjunct
areas. The latter is more especially the case with genera.

Versuch einer Entwickelungs-Geschichtc dcr Pflanzcnwclt. Theil
I-II. Leipzig 1879—1882.

the recent flora as composed of the results of earlier flora\'s
and flora-removals. Thus he accepts the main spreading
of plants being of relic-nature, only few plants representing
progressive endemism and he understands the floras basing
on geology and palaeo-climatology.

On the other hand one may consider the recent vegetable
kingdom as the result of recent progressive development,
the recent flora being composed of many progressive
endemics and only rather few relics. In recent times this
is especially paid attention to by J. C. Willis. He espe-
cially studied tropical flora\'s probably because he thinks
to have eliminated the influence if the Glacial-Period in
this manner

Geographical distribution of the Bignoniaceae. In
fossil state the family is only known in a few fragmentary
specimens -), which I shafl not mention here with a view
to the extraordinary difficulties met in determining sterile
specimens or leaf-fragments from recent Bignoniaceae, apart
from fossil ones, with which the difficulties of determining
are the larger. It is certain that in the Tertiary Period
Bignoniaceae existed already. The enormous spreading of
the family throughout the world also points to this.

A remarkable fact is the existence of a high percent of
monotypic or small genera, which cannot be referred to
the interpretation of the generic rank. Indeed the Bigno-
niaceae represent a multiform and highly differentiated
family in many respects. The following table may illustrate
this. The number of genera I obtained with the study of
K. Schumann in Engler-Prantl together with the publi-
cations in Just\'s Botanische Jahresberichte 1893—1923.

Thus in general the comparison will be a right one. though
the Bignoniaceae often alter as to their genus.

Whether this is a proof, however, for the rather large age
of a family or perhaps of its prosperity I do not know.
Which factor is the proof for the prosperity of a family ?
Is it determined by the number of large genera, the number
of genera, the number of species, the total number of
individuals or perhaps by the morphological multiformity?

The recent distribution extends over the palaeo- and
neo-tropics, some genera also occur in sub-tropical or
even in the temperate regions (Japan. North America, Andine
South America).

As to their altitude they occur in general at low and
medium altitudes up to 1000—1500 m. Montane and
alpine species are found indeed in some genera, e.g. in
Pandorea {Fenzl) Spach, Tecomanthe H, Bn. in New Guinea
and in Argylia D. Don and Campsidium Seem, et Reiss
in South America. So e.g. Tecomanthe arfaki vSts. attains
2500 m., T. volubilis Gibbs. 3100 m. and T. nitida-vSts.
3000 m.

The largest number of genera is found in America, as
may be seen in the following table.

It is remarkable that many Tecomeae are found in America,
whereas Asia only possesses 3 genera, viz. Oroxylum,
Millingtonia and Nyctocalos.

The genera which occur on 2 continents are not sepa-
rately inserted in the table. They are few and all Tecomeae:
in Asia and America: Campsis Lour., Catalpa Juss.
in Asia and Australia: Pandorea {Endl.) Spach., Deplanchea

Vieill., Dolichandrone {Fenzl.) Seem.
in Asia and Africa: Stereospermum Cham., Markhamia
Seem., Dolichandrone (Fenzl.) Seem.

It is peculiar that the Hawai Archipelago and New
Zealand are not inhabited by Bignoniaceae. In the Fiji
Archipelago one species occurs, viz. Tecoma filicifolia
Nichols, (Diet. Gard. 4 (1887)13), but I do not know this
species.

Affinities of the genera. 1 will mention several affinities
which seem to be of importance for an explanation as to
the origin of the African, Asiatic and Australian Bignoniaceae.
Though this hypothesis is most plausible, a monographic
study ought to be done, as I do not know many African
and American genera.

a. Affinities of E. African and Madagascarian genera with
S. E. Asiatic and Malayan genera.
Stereospermum Cham. — Africa. S. E. Asia to Borneo.
Markhamia Seem. — E. Africa, India.

Dolichandrone {Fenzl.) Seem. — E. Africa, India,
Malaya, Australia. Closely allied to Markhamia.

Podranea Sprague, Perichlaena H.Bn. and Kigelianthe
H.Bn. (S. and. S. E. Africa and Madagascar) allied
with genera of the Pandorea-group.

Many Crescentieae, Kigelia D.C. (Africa), Paracolea
H.Bn., Rhodocolea H.Bn., Siphocolea H.Bn., Phylloc-
tenium H.Bn. (all of Madagascar), Colea Baj.,
Phyllarthron D. C. (Madagascar, Seychelles and
Mascarenes); for the rest the tribus only occurs in
America.

Perichlaena H.Bn. (Madagascar), according to Sprague
(Hook. Icon. pi. 2749), allied with Tynnanthus Miers
from tropical America.

Neosepicaea Diels (New Guinea) is allied with Pandorea.
Hausmannia F.v.Muell.nbsp;......

Radermachera Zoll, occurs in both countries.
/. Affinities of Malayan and Australian genera with S.
American ones.

Nvctocalos T. et B. (Malaya, Br. India) probably allied
with Tanaecium Sw.; not allied with Rhigozum
Burch,, as Bureau (Monogr. (1864)52) remarks.
Campsidium Seem, et Reiss. (Chili) allied with the

Pandorea-Qtoup.
Argylia D. Don, allied with the Pandorea-group.
g. Affinities of N. E. Asiatic and N. American Bignoniaceae.
Campsis Lour, occurs in both countries.

Genctic phytogcography of the Bignoniaccae.
Though this enumeration may be incomplete, I am
convinced, that these affinities force the acceptance of the
hypothesis that the Bignoniaceae of Australia, Africa and
Asia have found their origin in America and that the
Bignoniaceae have developed in America, most probably
in tropical America.

All affinities point to the east, so that I suppose that
the palaeo-tropical species participated in a flora-migration
along the West-coast of the Pacific, to China and India
and that this migration went further on in 2 directions,
viz. a first group along Malacca and the Malayan Archi-
pelago to New Guinea and Australia, the other group
going to the west to British India, the Mascarenes, Seychelles,
Madagascar and E. Africa.

There are 2 books I have paid especially attention to,
as to the genetic-phytogeographical relations, viz. the ex-
cellent foundation of genetic-plantgeography quot;Versuch
einer Entwickelungsgeschichte der Pflanzenwelt seit der
Tertiär-Periodequot; of A. Engler (1879—82) and the work
of I. H. BurkilP) on the flora of Upper Assam and
its relations.

In general the relations of the Bignoniaceae mentioned
are in agreement with the indications of these two authors.

I used the ingenious theory about modern geology
of A. Wegener-), which seems to be of high interest
for plantgeography.

First we have to deal with the question about the
Behring-Straits Bridge, a necessarity of prime importance.

Since the beginning of the Tertiary Period, N. Amerika
and N. E. Asia were connected with each other. In the
Pliocene Period or probably in the beginning of the
Quaternairy-Period this bridge has definitively broken
down. Engler and Burkill both are in full agreement
with this geologic date. Thus the Bignoniaceae may have
gone in the Tertiary Period from America to Asia; pro-
bably not in the beginning of it, but perhaps in the
Miocene (Burkill), as the Tertiair-arctic-pole must have

1) I. H. Burkill. The botany of the Abor Expedition. Rec. Bot.
Surv. Ind. 10 (1924-5) 1-420.

•) A. Wegener. Die Entstehung der Kontinente und Oceane. Dritte
Auflage. Hamburg 1922.

been then near the Alutian Islands, but has gone to
Greenland. We do not know whether the Tertiary Big-
noniaceae were all tropical plants or also sub-temperate
genera, such as Catalpa and Campsis, but probably
they were sub-tropical or tropical.

On the other hand Engler and Wegener (I.e. 53)
point also to a connection of South America, the Antarctic
Region, New Zealand and Australia. It is probable that
the Bignoniaceae do not have migrated along this route,
in earlier times belonging to one continental quot;Schollequot;.
The Australian Bignoniaceae are allied with Malayan and
Chinese ones, and in New Zealand no Bignonian species
is found.

Thus the connection with the neo-tropical region must
be found in the recent-arctic region.

Burkill (I.e. 187) gives an account of several genera,
now extending as well in America, as well as in India,
which have migrated east- or westwards, but all of them
using the Behring-Straits Bridge in some warm period.
When they crossed the bridge its climate must have been
sub-tropical and fairly humid. He mentions the following
genera of several families, not occurring in Africa, but
extending from Abor-Land to N. America: Magnolia,
Dicentra, Cedrela, Leucaena, Polygonatum, Dispomm, Spi-
ranthes, Microtropis, Meliosma, Gymnocladus, Pachyrrhizus,
Hydrangea, Pentapanax, Pratia, Symplocos, Callicarpa,
Armdinaria, Podocarpus and Lindera; the first seven genera
of which have probably come from America.

Burkill says (Ic. 193): quot;A period when a Rain-forest-
flora could pass north of an Arabian sea seems to have
been Miocene: geologists tell us that the union of Mada-
gascar to Africa become broken during the Miocene age
and it would appear that the Miocene-Rain-forest in
Asia spread northwards over the whole of the then exis-
ting India, so as to pass round the seas that bounded

the Deccan both east and west, i.e., to the north of lati-
tude 20° N. It may also be said that at one period it
went yet much further north; for genera, such as Gordonia
including Haemocharis (now found in the West Indies, in
Venezuela and in contiguous parts of S, America as well
as is Malaysia), and Eurya (just a litde less wide), and
other forest plants appear to have been able to cross
the Behring-Straits Bridge in latitude of 60° N., together
with various forest animals, such as the early apes of the
Prosimudae. and this in the Miocene Period.

It is remembered, that humidity as well as warmth, is
necessary for the Rain-forest, and a drying of the climate
is as fatal to it as a lowering of temperature. It is most
likely, that want of humidity limited the Miocene Rain-
forest before want of warmth and that it was an extension
of an adjoining dry climate that cut the route between
Africa and India. I offer no opinion upon the possibihty
of the spread forthwith into Miocene India, but there seems
to be good reason, as has been already mentioned, to think
that a dry climate ruled in north-western India towards
the end of the Pliocene, and at that time, the dry condi-
tions may have been so wide spread in Asia, on the
north side of the tropics as to extend to the Behring-
Straits Bridge. However, as Engler showed forty years
ago (l.c. 26) woodland species seem to have been among
the last wanderers across the bridge, which broke down
in the Pleistocene Period at a time when it was temperate
to arctic.quot; And on p. 194: quot;The dry conditions, which
I assume to have cut the Miocene Rain-forest of Africa,
apart from the Miocene Rain-forest of Asia, persisted
until the Glacier Period produced a new and intenser
barrier by the descent of the cold to the shores of the
Arabian sea, the cold bringing down temperate genera,
giving them access to aequatorial mountains, and com-
pressing that flora, which needed a dry heat, between its

invading temperate plants and the fringe of tropical Rain-
forest which clung to the Southern End of India and
Ceylon. That this Rain-forest survived, we know; its
survival makes us to assume, that the effect of the glacia-
tion in the north was rapidly lost towards the equator;
for, if it were otherwise, by what means could the Rain-
forest have held on?quot;

Campsis and Catalpa may be of relative young age in
the sub-temperate regions of E. Asia and North America.
The disjunct area may have been originated in the Qua-
ternary Period, when (Wegener) the Behring-Straits
Bridge was broken. Whether some Bignoniaceae have gone
along the Alutian Islands I do not know.

Burkill says (Ic. 180): quot;The Glacial Period seems not to
have had such an influence upon the Indian and Malay-
sian Rain-forests, at least did not act destructively upon the
Malayan flora, though it may be asserted to have com-
pressed it from the north, when it let in the chief part
if not all the northern Spermatophytes found upon the
tropical mountains of both worlds,quot; with which remark I
am in full agreement. Only two genera, viz. Campsis
and Catalpa with sub-temperate species reach ± 40° N.L.
the recent northern frontier of the Bignoniaceae,

The migration southwards to Japan, China and S.E.
Asia must have taken place in the Tertiairy Period. Per-
haps the bulk of them were Tecomeae and only few Big-
nonieae; but also there were Crescentieae, as these plants
have reached the Seychelles, and even Madagascar. Most
of them took part in tropical forest-vegetation. Even in
the Tertiary Period (Miocene) these forests extended over
a great deal of India, in those times united with E. Africa
by Ceylon, the Seychelles, the Mascarenes and Madagascar,
all of them forming a coherent geological unit, „Schollequot;
(Wegener), before the ..Lemurian Zusammenschub.quot; In
this way the Crescentieae and the ancestors of other ende-

mic Madagascarian genera have come in their recent places,
and must have become extinct in the areas between. Bur-
kill (l.c. 190) also mentions the close affinities of S. E.
Asia with the Malagassian Region, Madagascar and E.
Africa. The Lemurs occur in Africa, Ceylon and Malaya.
He says: quot;The Sclaters in their geography of Animals
(1899—104, 108, 151) suggest that between Madagascar
and Ceylon they passed via the north side of the Arabian
Sea, via Africa whence Madagascar obtained its mammal
fauna. We must assume this that at.one date there existed
a Rain-forest climate around the north side of the Arabian
Sea. If the Rain-forest conditions be held as allowing the
Lemuridae to attain their Malagassy-Malaysian distribution,
the occurrence of quite a considerable series of genera,
such as. Nepenthes, Tamtourrissia, Erythrospermum, etc. in
Madagascar as in Malaysia, seems to require no further
explanation,quot; and further on p. 192: quot;The Rain-forest
areas in Africa may at one time have been extensive and
must have been more extensive than now: but from very
far back in the history of the Spermatophytes, it, in gene-
ral, has had a dry climate: wherein apparently it evolved
its numerous genera of terrestrial orchids, now so charac-
teristic, and in the possession of which it is such a contrast
to Malaysia and South America.quot;

Engler (l.c. 289-297) and Burkill (l.c. 191) give
several remarks on the alliance of the Mascarenes, Seychelles
and Madagascar with Malayan, Australian and Pacific forms.
Of this examples I will mention: Weihea, Memecylon,
Woodfordia, Parapsia, Pipturus, Thesperia, Evodia, Polyscias,
Geniostema, Cerbera, Ochrosia, Alyxia, Exocarpus, Tricho-
desma, Medinilla, Oberonia, Dillenia, Pothos, Melostoma,
etc.. Weimannia his especially developed in Madagascar,
the Mascarenes, Java, the Fiji Island and South America;
the same with Nepenthes, Rubtis moluccanus, Nesogenes
(Rodriguez and Pacific Islands) and the allied genus Acha-

ratea (Madagascar), whilst other allied genera of Chloan-
theae are limited to Australia; the same with Acacia
heterophylla (Madagascar) showing affinities wich those of
the Sandwich Islands and of S. America. Pisonia and Ocotea
( § Mespilodaphne), Omphalea, Milla and the Turneraceae
show the same relations.

Madagascar and the whole Malagassian Region show
a high percent of endemism, due to their high age and in
early times already isolated position. Certainly Wegener\'s
theory about the existence of a united Africa. Madagascar,
Malagassy, Ceylon and Deccan makes it easy to under-
stand the phytogeographical relations, which were already

Whether Africa has got American Bignoniaceae from the
west I do not know, as I did not study all African genera,
but I do not think this to be probable, as all those I know,
are closely allied with Indian or Malayan genera and
much less with American ones.

The same is to be found in Europe, which does not
possess any Bignoniaceae though the eastern United States
are inhabited by three genera with 4 species and an
exchange of this sub-temperate species has not taken place
evidently before the Quaternary Period, when the Atlantic
breaking up between Africa and South America, extended
northwards, and separated Europe and North America

As to the relations of India, the Malayan Archipelago
and Australia, the Bignoniaceae must have reached Australia
in the Tertiary Period; they are known in New Caledonia
and the Fiji Archipelago, where occur few. but endemic
species.

The general relations I mentioned already are so striking
that I give up attempts to illustrate these nearer, but I
will pay attention to some peculiar ones within the
Malayan Archipelago. The Philippines got their genera

most probably from the other Malayan Islands and not
along Korea, Japan and Formosa. No Japanese species
is found here. Probably the Philippine Bignoniaceae migrated
along the Malay Peninsula and Borneo to Palawan, etc.,
but perhaps also over Celebes and the Talaud Archipelago.
Radermachera and Nyctocalos show resemblance with each
other in this respect. Nyctocalos shows perhaps the route
along Sumatra •gt; Java and Celebes. The two other Philippine
genera viz. Oroxylum and Dolichandrone possess a very
large area and are common within their areas, but have
certainly migrated along Borneo and Celebes to the Philip-
pines. The only indication for a migration from China
and Formosa to the Philippines may be found in Rader-
machera also occurring in China near Kanton, but it seems
more rational to accept the first route, because the Chinese
species is not immediately allied with Philippine ones.

How it is that Pandorea occurs in Malacca and further
in Ceram. Ambon and the Key Islands I do know, but it
is very peculiar. Deplanchea shows the same, having one
species in the Malay Peninsula, Sumatra and W. Borneo
and the islands between, whilst the others are found in
New Guinea and further in Australia. Tecomanthe occurs
in Ternate, Ambon, Ceram, New Guinea and Australia.

For this we can accept the route Malacca-^Sumatra-^
Javaexternal Sunda-Islands-garland-^New Guinea. In this
respect and according to Wegener\'s hypothesis, this route
may be understood, these islands forming a straight bridge
in earlier times. How it is that no Deplanchea, Pandorea
or Tecomanthe is found in the areas between W. Borneo
and Malacca and Ambon and Ternate on the other side.
I cannot understand. Certainly they became extinct on
their route, but for what reason one cannot say. It is shown
that ± all Malayan Bignoniaceae show rather strongly
pronounced disjunct areas.

letic (polytope origin of species) development, the present
author does not accept this for the following reasons.

Firstly he does not know any well-argued example. The
fact that Oenothera Lamarckiana of H. de Vries gave
several times raise to the mutant O. gigas and other
mutants, does not say anything about this question; where-
as it is established that 0. Lamarckiana shows hybride-
characters.

Moreover the examples given by Briquet and others
are based on indications about former climates and geological
arguments, which are not sufficiently proved. Thereabove
they often studied vicarious species, which may be parallely
developed ones.

The natural system of historic geology and the palaeon-
tology give no indications in this direction either. The fossil
species and genera, which characterize certain geological
strata, on which even the division of strata has been
based in a lot of cases, are certainly of the same origin,
which has been proved by the other strata immediately
above and beneath, showing always the same series.

Large groups may be composed of phylogenetically
heterogeneous elements as for instance may be the case
with the Sympetalae. As to the tribi of Bignoniaceae there
seems to me no reason at hand to consider them as a
result of polyphyletic development. They are truly too
closely allied with each other and rather too sharply
separated from the allied families.

Age and Area. It seems to me that the quot;Age and Areaquot;
hypothesis of J. C. Willis is in contradiction with what
I found when studying the Bignoniaceae.

Striking examples are: Nyctocalos T. et B. (Fig. 1. p. 806),
Dolichandrone {Fenzl.) Seem. (Fig. 11. p. 934), Pandorea
(Endl.) Spach. (Fig. 4. p. 847), Radermachera Zoll. et Mor,

One cannot say that this is due to the fact that these
regions have not been worked out nowadays. Though
Borneo, Celebes and New Guinea are more a less unknown,
considerable collections have been made during the last
few decennies in these countries.

Secondly the Bignoniaceae show a striking endemic
development and many local spread species. So e.g. in
Tecomanthe H.Bn. (Fig. 6. p. 885), Deplanchea VieilL
(Fig. 9. p. 915), Radermachera ZolL et Mor. (Fig. 12.
p. 954-5), Dolichandrone {Fenzl.) Seem. (Fig. 11. p. 934),
Pandorea {Endl.) Spach. (Fig. 4. p. 847) and Nyctocalos
T. et B. (Fig. 1. p. 806).

This cannot be denied even in New Guinea. What
might be the reason that Tecomanthe dendrophila {BL) K.
Sch. is found throughout New Guinea and even in the
Solomon Archipelago, whereas many other species have
a local spreading?

The most striking example is the case with Nyctocalos-
T. et B.; the four well-defined, though closely allied
species known, are found in four different countries and
separated by enormous distances. And Nyctocalos must
have an origin of its own as it represents an absolutely
isolated Bignoniea in Asia. It is neither allied with Oroxylum
nor with Millingtonia, the two other Bignonieae-Qemva
not occurring in America. Oroxylum on the contrary
shows a well-defined continuous area.

As to Radermachera, the Indian species R. xylocarpa
{Roxb.) K. Sch. (Fig. 12 (6). p. 954-5) and the Chinese species
R. pentandra HemsL (Fig. 12(13). p. 954-5) and R. sinica
{Hance) HemsL (Fig. 12(11). p. 954-5) are entirely isolated.
The other 17 species are chiefly found in the Philippines,
except the two common Malayan species viz. R. gigantea
{BL) *Miq. (Fig. 12(17). p. 954-5) and R. glandulosa {BL)
Miq. (Fig. 11 (5). p. 954-5). In the Philippines there are

few species with a greater spreading, the bulk being
endemic in different islands. Only i?. pinnata {Blanco)
Seem, (Fig. 12 (10). p. 954-5), R. fenicis Merr, (Fig. 12 (14).
p. 954-5) and R. acuminata Merr, (Fig. 12(15). p. 954-5)
are more common species in the Phihppines.

With Tecomanthe it is the same case. T. ternatensis vSts,
(Fig. 6(12). p. 885), T, Hillii {F, v, Muell,)- vSts. (Fig.
6(13). p. 885) and T. amboinensis {Bl) Boerl (Fig. 6 (11).
p. 885) are rather isolated, especially the second rare one;
but there are also 2 more common species, viz. T, den-
drophila {Bl) K, Sch, (Fig. 6 (5). p. 885) and T, venusta
S, Moore (Fig. 6(16). p. 885). The others are local
spreaded species—endemics.

Deplanchea Vieill (Fig. 9. p. 915) and Pandorea {Endl)
Spach, (Fig. 4. p. 847) show a striking resemblance. Of
both genera, which are not in the least allied with each
other, one species is found in Malacca {P. Curtisii Ridl,
strongly endemic) or Malacca and surrounding countries
{D, bancana {Scheff,) vSts,, Malacca. Riouw Archipelago,
Banka. Billiton, W. Borneo and S. E. Sumatra). Further
both genera occur in New Guinea (distance ca. 3000 km!)
and surrounding countries and at last some species are
inhabitants of Australia and New Caledonia.

Of Dolichandrone {Fenzl) Seem, (Fig. 11. p. 934) one
species is found in E. Africa, also at a distance of about
3000 km from the allied Brit. Indian species.

The theory of *\'Age and Areaquot; is principally based on
progressive endemism. A species slowly enlarges its area;
it produces new species from time to time (mutation-
periods); the new species do the same, and so on. Thus
the eldest species must show the largest areas, the
youngest will be highly endemic.

It is a well-known fact that several Gymnosperms, which
in earlier periods extended over very large areas (paleon-
tology) are now limited to local countries {Sequoia, Ginkgo,

Taxodium.). Willis does know these facts, but he con-
siders this relic-endemism as an exceptional case. Besides,
he considers especially tropical areas and thus thinks to
avoid the difficulties with glacial relics.

In my opinion there are not only such apparent relics
as mentioned above, but relic-endemism is also to be
found in the tropical primeval forests. Why not? They
also are composed of members of earlier floras and not
only the result of progressive-endemic development.

My main argument against Willis is that generally
one is unable to distinguish at first sight the difference
between a progressive- and a relic-endemic.

This can only be made by a close study of the geo-
graphical distribution of all the species of a genus; of the
affinities of the genera and of all species of a genus.

E.nbsp;g. in Deplanchea Vieill. the New Caledonian species
are more closely allied with those of New Guinea than
with those of Queensland and the Queensland species
D. hirsuta {Bailey) vSts. and D. tetraphylla {R. Br.)

F.nbsp;V. Muell. show most affinities with D. bancana {Scheff.)
vSts. at a distance of about 3500—4000 km! No species
has been found in the area between them, though the
trees are rather conspicuous as to their habit.

In my opinion the geographical distribution of the
Bignoniaceae points to relic-endemism, the only manner in
which the difficulties may be cleared up, though I consider
some species of Radermachera in the Philippines and of
Tecomanthe in New Guinea as the result of a progressive-
endemic development.

Willis treats the age of a flora by means of adding
all species, each with a number as to its rarity. His results
are almost too good; he obtains 90 % and more the same
as the results which could be obtained from a theoretical
point of view. But if it will appear — and I am sure of
that — that more families in the tropics exist with a high

grade of relic-endemism, then it will be clear that his
ingenious theory is unsound. Especially this will be the
case in my opinion with his statistical researches. Statistics
are always a dangerous means for such researches.

At last one does not known any longer the object one
studies, but chiefly gives attention to the curves. And curves
also are very difficult to interprete.

Burkill (I.e. 198) does not accept quot;Age and Areaquot; as
an explanation for the origin of tropical flora\'s. He says:
quot;but the reason which he gives for endemics being found
over narrower areas in Ceylon than species widely spread
beyond Ceylon, namely that the endemics are new species
which have not had time to spread, will not in his simplicity

endemics are figures for Rain-forest, and, though he does
not bring the point out, his figures for wides, are a mixture
of species of all sorts of conditions, dry and wet, agrestal
and forest so that he has contrasted species for which
the area suited is obviously limited with species for which
the area suited is wide, and therefrom comes the obvious
order in his results.quot;

Of course I fully agree with Burkill, who studied the
Indian Rain-forest-flora just as Willis did.

Indeed only the most simple case: a newly originated
species spreads from a centre, gives raise to new species,
etc., the new species being similar as to their manner of
spreading, their resistence against external factors, etc.:
only this can be explained with Willis\' theory, but also
without it.

But if the case is more complicated, quot;Age and Areaquot;
cannot explain anything,\'as this hypothesis does not take
account of the earlier flora\'s which are certainly for a great
deal extinct nowadays, nor of the biology of the species.

Sunda- and Sahul-shclf. (Fig. 18). Upon the importance
of these biological data which are based on geological ones

and concern the origin and explanation of the distributions of
animals and plants in Malaya. Merrill^)-) has recently
written some excellent papers. Also H. J. Lam^) and
others are interested in this problem, which was put
forward by Wallace, and later on by P. Pelseneer^)
and was established by G. A. Molengraaff

First I will deal with the limits of the Sunda-shelf to
its north. MerrilP) has pointed out that Formosa belongs
in plant-geographical respect to the China-shelf. It possesses
many Himalayan and Chinese continental temperate genera
and even families, which do not occur in the Philippines.

The Philippines on the other side possess a lot of espe-
cially tropical-forest-genera, which do not occur in Formosa
(Merrill l.c. p. 600). Though one can see Formosa from
L\'Yami, the most northern island of the Philippine Archi-
pelago, a deep separates the regions since the beginning
of the Tertiary Period. The most striking example of the
extremely sharp separation is that of the Dipterocarpaceae ®),
a relatively newly originated family, the bulk of which
has been found in Borneo, but has certainly gone along
the Sulu- and Palawan Archipelago to the Philippines.
Notwithstanding the fact, that fossil Dipterocarpaceae
have been found in the Pliocene of Luzon, no species

1) E. D. Merrill. Die Pflanzen-gcographische Scheiding von For-
mosa und der PhiUppinen. Engl. Bot. Jahrb. 58 (1923) 599-604.

-) E. D. Merrill. The correlation of biological distribution with
the geologic history of Malaysia. Proc. Pan-Pacific Sci. Congress. Australia.
(1923) 1148—55.

•f) H. J. Lam. Remarks on the genetic-phytogeography Malayan
Archipelago. Ann. Jard. Buitenzorg 37(1927) 33—48.

■») R. Pelseneer. La ligne de Weber, limite zoologique de I\'Asie
et d-Austrahe. Bull. Acad. Belg. (1904) 1001-1022.

5)nbsp;G. A. F. Molengraaf f. Modern deep-sea research in the East-
Indian Achipelago. Geogr. Journ. 57 (1921) 95—121. fig. 1-9, map.

6)nbsp;E. D. Merrill. Distribution of the Dipterocarpaceae. Philip.
Journ. Sci. 23 (1923) 1—33. t. 8.

occurs in Formosa and only few have been found in
Celebes, the Moluccas and New Guinea.

The Bignoniaceae show the same relations in this case;
as I above pointed out the route over China, Malacca
and Borneo to the Philippines and not over Formosa.
So e.g. Oroxylum Vent, (Fig. 2). Dolichondrone {Fenzl) Seem.
(Fig. 11), Radermachera Zoll, (Fig. 12) and Nyctocalos T. et B,
(Fig. 1) the only Philippine Bignoniaceae.

The limits of the northern part of the Sunda-shelf are
geologically and botanically undoubtedly established.

Merrill (I.e. 603) says: quot;Der Philippinen Archipel
zeigt hauptsächlich eine malayische Flora, und diese ma-
layischen Elemente sind abzuleiten im Südwesten von den
Sunda-Inseln, im Süden und Osten von Celebes, der Mo-
lukken und Neu-Guineaquot;, and I give up further botanical
illustration of this fact by means of Merrill\'s most inte-
resting examples.

A more troublesome matter is that of the limits of the
Sunda-shelf to the cast and the Sahul-shelf to the west.

A first line has been drawn by Wallace between the
Philippines, Borneo and Bali on one side and Celebes
and Lombok on the other side, along the Makassar Strait
and the Lombok Passage.

The second line has been called Weber\'s Line by
Pelseneer and has been drawn more eastwards; it extends
east of Timor, west of Buru and Halmaheira and cast of
the Philippines.

The Sahul-shelf is constituted of Australia and New
Guinea: an elevation of 20 metres of the shelf would
connect Australia and New Guinea with each other. Both
shelf-regions have been stable at least since the close of
the Pliocene. During Pleistocene Ice-Ages New Guinea
was alternately connected and disconnected from the Asia-
tic continent.

approximately delimited by the present of the 200 m.
isobath, probably persisted to a greater or less degree in
the preceding geological periods.

The break between Australia and Asia probably came
in the late Cretaceous or in the Epi-Mesozoic-interval, and
since that time it seems to be evident that intermigrations
of plants and animals between Australia and Asia, and
between eastern and western Malaya, have been inter-
rupted or inhibited by the constant archipelagic condition
of this intermediate unstable area quot;Wallaceaquot; (Merrill).

So Wallace\'s Line is the place where the unstable
Wallacea impinges to the west on the stable Asiatic conti-
nental area, W e b e r\'s Line where it impinges to the east on
the Papuan — Australian Pliocene — Pleistocene continent.

Merrill and others suggest that both continental areas
represent both centres of origin and distribution, the inter-
changes of which have taken place over quot;Wallaceaquot;.
Thus in quot;Wallaceaquot; a mixture of Asiatic and Papuan
(and Australian) forms has been found. The bulk of them
can be explained by the two lines mentioned. But I fully
agree with Merrill (I.e. 1155) where he says: quot;In sum-
marizing I would emphasize the conclusion that Wallace\'s
Line cannot be abandoned in favour of Weber\'s Line,
nor vice versa. Both are important, although neither is an
absolute boundary between Asia and Australia, when all
groups of plants and animals arc taken in consideration.
Both are due to fundamental geological causes, one being
the western, the other being the eastern boundary of a
long-continued insular unstable region situated between
the two ancient stable continents. No matter how much
more convenient a simple line of demarcation would be
from the stand-point of delimiting bio-geographical areas. I
believe that we must be forced to abandon the idea of
a simple boundary line, and accept a region of transi-
tion as separating Australia and Asia, and this transition

Though the number of genera is relatively small, they
are to divide in some groups.

The first group, is characterized by a continen tal
Asiatic character, viz. Haplophragma P. Dop. (Fig. 15.
p. 1005) from Burma to Malacca and N. Sumatra and the allied
Heterophragma D. C. (p. 999) not found as yet in the
Malayan Archipelago: Stereospermum Cham. (p. 947) ex-
tending from Africa to India, Burma, Malacca and Borneo
and several S. E. Asiatic genera, such as Tecomella Seem.,
Pajanelia D. C., etc.

Secondly one genus extends throughout the palaeo-
tropics, viz. Dolichandrone (Fenzl.) Seem. (Fig. 11. p. 934) from
E. Africa to New Caledonia. Especially D. spathacea (L.f.)
K. Sch. (Fig. 11 (4). p. 934), a distinctly distinguishable
coastal tree occurs from Malabar to New Caledonia, but
is not known from Australia, a problem in itself in regard
to the relative easily spreading of its corky seeds by sea-
currents. The genus is represented in N. Australia with
a sub-genus Coriaceae vSts., which is limited to that
region. Thus Dolichandrone (Fenzl.) Seem., most probably
of high age in the branch of the Bignoniaceae, is not in-
terrupted by any of the lines of Weber or Wallace.

Thirdly there are 3 genera which are only known from
the regions west of Weber\'s Line, viz. Radermachera
Zoll. (Fig. 12. p. 954-5), Oroxylum Vent.(¥[g. 2. p.817) and
Nyctocalos T. et B. (Fig. 1. p. 806).

The first two have perhaps gone to the Philippines
along Borneo, though R. fenicis Merr. points also to
Celebes as a probable route. Nyctocalos T. et B. shows
the close alliance of Celebes with the Philippines; N. cus-
pidatum (Bl) Miq. has been found throughout the Philip-
pines and is known also from the Minahassa.

The fourth group is that of the genera belonging to
the Papuan-Australian continent east of Weber\'s Line,
viz. Tecomanthe H. Bn. (Ternate, Ceram and Ambon,
N. Guinea and Queensland) (Fig. 6. p. 885) with 16 species,
Neosepicaea Diels (E. N. Guinea) (Fig. 7 (1). p. 899) mono-
typic. and Hausmannia (Queensland) (Fig. 7 (2). p. 899)
monotypic.

All of these three are allied and belong to the Pandorea-
group (p. 837). Perhaps one would believe the 2 mono-
typic genera to be more or less differentiated, younger
ones. But in morphological respect they show the (perhaps)
primitive actinomorphic Tecoma-fiower, whilst Tecomanthe
is mostly distinctly zygomorphic, so that the case may
be justly the contrary.

The fifth group includes still 2 genera, which cannot
be explained by Weber\'s Line, viz. Pandorea {Endl.)
Spach. and Deplanchea Vieill. Both genera belong princi-
pally to the Papuan-Australian continent following
Weber\'s Line, but each of them is represented by one
species in the western part of Malaya. Pandorea Curtisii
Ridl. is a strongly endemic species in Malacca, and De-
planchea bancana {Scheff.) vSts. is a rather common
spread species in Malacca, S.E. Sumatra, the Riouw Ar-
chipelago, Banka, Billiton and W. Borneo. This may be un-
derstood, if we suppose that both genera possessed much
larger areas in earlier times. Pliocene or Miocene perhaps
and that both have partly become extinct.

Perhaps later on the Papuan species have again en-
larged their area and may have produced new species,
extending the area again westwards, but was limited there
by Weber\'s Line which had become a real barrier in
the meantime and thus P. ceramensis may be interpreted
(or the lt;ancestors of this species).

In general we can say that the geographical distribution
of the Malayan Bignoniaceae may be explained by the

unmodified line of Weber, though they existed already
in Malaya and Australia before Weber\'s Line was a
real barrier for distribution.

Though in chapter II I have mentioned already something
about the use made of Malayan Bignoniaceae, I intend to
give an enumeration here.

Generally speaking, the species do not give a good
timber. Oroxylum indicum (L) Vent., Dolichandrone spathacea
(L. f.) K.Sch., Radermachera glandulosa (Bl) Miq. and
Crescentia Cujete L. are small trees, mostly with a crooked
main stem and of a crowded habitus.

Lians. such as Nyctocalos T.et B., Pandorea (Endl.) Spach.
and Tecomanthe H.Bn. cannot be used as timber either.

Deplanchea bancana (Scheff.) vSts. and all other Deplan-
chea\'s possess a very soft wood, not fit for timber.

The only species which produce a good timber are the
large trees (30—50 m.): Haplophragma macroloba (Miq.)
vSts. (= Heterophragma macrolobum Backer in Heyne
Nuttige Planten Ned. Ind. 2. (1927) 1371) and Radermachera

The first species only occurs in N.W.Sumatra (1006):
rather frequent in the quot;Padangsche Bovenlandenquot; (Res.
Sumatra\'s Westkust) and less frequent at lower altitudes
in Priaman and moreover spread to Atjeh. It gives an
excellent timber, solid and delicately fibrous; the natives
use it often for proa\'s. Perhaps it would be interesting to
cultivate the gigantic tree elsewhere in the Malayan
Archipelago. In the Hort. Bogor. the species is cultivated
and seems to prosper pretty well, though it never produces
capsules there, but only flowers.

The second species: Radermachera gigantea (Bl) Miq.,
is also useful for its timber ; it is a tree of 25 m. high on
the average. It occurs in the entire Malayan Archipelago
and is often frequently found up to 1500 m. altitude.
The wood is just as that of Haplophragma macroloba
(Miq). vSts. useful for building-purposes; the structure is
durable and the timber rather large. Nevertheless the use
of it is only local.

It would be of interest to make experiments with the
cultivation of both species.

The wood and bark of Oroxylum indicum (L) Vent.,
Dolichandrone spathacea (L.f.) K.Sch., Radermachera gigantea
(Bl) Miq., Radermachera glandulosa (Bl.) Miq., Haplo-
phragma macroloba (Miq.) vSts. and Cresentia Cujete L.
is further used for several officinal and other purposed.
See for this Heyne: Nuttige Planten Ned. Indië.
2 (1927) 1370-2.

1.nbsp;An oudine of the taxonomy is given, together with
the geographical distribution.

2.nbsp;It has turned out that the monotypic genus Haplo-
phragma P. Dop. (1925) possesses a second species in
Sumatra, viz. H. macroloba (Miq.) vSts.

3.nbsp;To emphasize the fact that the Australian Dolichan-
drone\'s are separated from the other species, it is sug-
gested that Dolichandrone (Fenzl.) Seem is constituted
of 2 sub-genera, viz. Coriaceae vSts. (Australia) and
Membranaceae vSts. (Africa, S. E. Asia, Malaya,
New Guinea, New Caledonia).

4.nbsp;A scheme of the phylogenetic relations within Tecomanthe
H.Bn. is made, based on the structure of the ovary,
together with the habitus and geographical distribution.

©I

1 /unknown\\4-
flnccMor;

^VOLUlilLCy

6.nbsp;The affinities with the neo-tropical Bignoniaceae make
it probable that the Malayan Bignoniaceae have origi-
nated from neo-tropical ones and have taken part in a
Tertiary flora. A hypothesis is made about the routes
along which these tropical genera have migrated. Most
probably they have gone along N. America, to N. E.
Asia, China and Japan along the Behring-Straits Bridge
and perhaps also along the Alutian Islands: the bulk
of them belong to the Tecomeae-stem. One part has
gone to the east: Malaya, Philippines, New Guinea and
Australia. Others have migrated to British India and
E. Africa, perhaps along the north of the Arabian Sea,
but certainly also directly to E. Africa over Ceylon,
the Mascarenes, Seychelles and Madagascar (Wegener).
Thus the Bignoniaceae point to the same genetic-phyto-
geographical relations as many other families do.
(Fig. 17).

7.nbsp;The Bignoniaceae are most probably at least of Tertiary
age. They show many endemic species and genera, but
the latter often also with highly disjunct area\'s. These
genera are partially extinct nowadays, but must have
possessed a greater spreading in earlier times. This
fact cannot be explainecUwith Willis\' quot;Age and Areaquot;
hypothesis, this theory being only of value for very
young flora\'s. Only in Radermachera and Tecomanthe
some rather young species occur, these being progres-
sive endemics, which can also easily be understood
without the quot;Age and Areaquot; hypothesis.

My main argument against Willis is that generally
one is unable to distinguish at first sight the difference
between a progressive- and a relic-endemic. Only a
close monographical study may clear up the phylo-
genetic relations.

Relic-endemism is not rare in the Angiosperms and is
not only due to Ice-ages, but reUc-endemism is also
to be found in the tropical primeval forests (and Rain-
forests), which also are composed of members of earlier
floras and are not only the result of progressive-
endemic development in agreement with the ingenious
outline of genetic-phytogeography of A. Engler.
Moreover quot;Age and Areaquot; is a statistical (mathema-
thical) theory, not keeping account with the biological
peculiarities of the species.

8. The geographical distribution of the Malayan Bigno-
niaceae may be explained by Weber\'s Line, though
some of them existed already in Malaya before this hne
represented a real barrier for biological distribution.
Fig. 18.

The present work has been made in the Botanical
Laboratory of the University at Utrecht.

The author wishes to express especially his sincere
thanks to Professor Dr. A. Pulle. Director of the
„Botanisch Museum en Herbariumquot;, who enabled him to
work out the present publication under his authority and
kindly put his most valuable help in all respects at his
disposal.

New species, varieties, forms, etc. and descriptions are marked with:
nov. spec., nov. var., descr. emend., etc. behind the name.

adenophylla (Wall.) P. Dop... 1006
macroloba (Miq.) vSts.
nom.nov. dcscr. emend.1002, 1039

THE MALAYSIAN REGION
SHOWING THE ASIATIC AND AUSTRALIAN CONTINENTAL SHELVES

AND THE ASSOCIATED DEEPS\'.
CHIEFLY AFTER MOLENGRAAFF.

LEGEND

Asiatic continental shelf and stable area,
delimited by the ZOO meter line.
Australian continental shelf and stable area,
delimited by the ZOO meter line.
Wallace\'s and Weber:5 lines.

---Walldce\'5-and Weber\'s lines modified.

Deep sea basins the figure indicating the dflh in meters.

De „Agc and Arcaquot;-hypothcse van J. C. Willis
is geen biologische, maar een mathematische theorie.

Een „biochronische Gleichungquot; zooals deze
geformuleerd is door H. de Vries mist een
phytographische zoowel als een genetische basis.

Het is niet bewezen en in hooge mate onwaar-
schijnlijk, dat de aanpassing in het evolutie-proces
een rol heeft gespeeld.

Zelfbestuiving en zelf-fertiliteit is een algemeen
voorkomend verschijnsel bij Orchidaceae.

De frequentie-krommen van Raunkiaer en de
konstanten-krommen van du Rietz c.s. vinden hun

Het tot stand komen eener voorloopige Flora
van Nederlandscfi Oost-Indië is van urgent belang
en hierbij mag de daadwerkelijke hulp van \'s Rijks
Herbarium niet gemist worden.

De conclusie van Weber, dat de geotropische
kromming de theorie van Blaauw volgt, is onge-
oorloofd.

De voorwaartsche verplaatsing der Gregarinen
is het resultaat eener slijmuitscheiding.

Het is niet bewezen en hoogst onwaarschijnlijk,
dat virus-ziekten van planten worden veroorzaakt
door levende wezens.

De proeven van Anselmino bewijzen niet,
dat bij Tropidonotus natrix geen terugresorbtie
plaats vindt.

De veenvorming in de tropen is van groote
beteekenis voor de geologie, de palaeo- en phyto-
geographie. Het voortgezet onderzoek behoort te
steunen op de innige samenwerking van agro-
geologen, palaeo-geographen en botanici.

90	100 /n	r	120 130
^ \'i \' • / 10 e /	L S\'	/ D	4\\
	, X Ùgt;gt; (	/	» ^ \\ 4 ✓
h 10	V		1 csgt; \'

i»	130	140	150
1		CD ( S Q	■ N -I--
ió - /	^^—5 lcgt;
20		L ^	i?

100 ^ , s ts •	no ^^		gt;30	140	150	160
¥ 6 V 10	f i.	1 Cigt; •	f* --	• 1 c		1 1 !
10				w\\		••
		1		\\

				Genera with
	Number of	Genera with Genera with	6—50 or
	Genera	1—2 speciesjl—5 species	more species
Bignonieae . . .	56	1 36	47	9
Tecomeae....	47	1 28	38	9
Tourretieae . . .	1	! 1	1	1
Crescentieae. . •	14	1 ^	13	I 1
Eccremocarpeae .	1 1	j —	1
Bignoniaceae . .	119	quot; quot;72quot;quot;	100	r 19



		\' i .Vß^.-
		■ ......
ecf	......... \\ ,..... :-T.a\'i.		........
	........jm-\'i; .■gt;• tr.T-V
	______ ;		..........
		T .... .• -
Tc*®	v\'.^\' .... .liiiicrj:» -Trj^vh		.„^cilh\'Tî;^ ■■ -^iri\'-\'
	;_________.-iZv	. -. V.	.rn-ïvi ,v\'ia.
	.-■iZ\'f C.iG) .»Jan\'.j\'^k • îrrtfc	• : Ti-? .....	........r!\' ,Ci
r	.... ......... .von .mvn-	: i ; ......	. ràii\'-j ii..:-V.iîfcV/
			. . ...
	.............quot;i;	. ■ , V . . .	......■
	........\'-A -.v-;		, .... . - , v.-H V .
	............ . ■ \'s. \'i.:	quot; • J \' \'	... , ^ •. . •
			- ■ , \' . quot;■■gt;
	.... . ______ .u\'-. ^		. .. • ■ ■. . \'..\'ÏBin
	. . - . •■tv.\'.; ; ■ ■ ■■■K;\'
■			. ■ .■ . gt; . .. \' \'-lt;\'\'gt;
	/\'rtcr. . ■quot; ■ .{. ..i ,v .\'i;		, ...... ■ \' ,
	. ., . .• . ...... :r \'
	... • ; ■e\'. y I		. ■ .\'
	..... jöiU\' äfc-rt.\'î		• gt; r- quot; •
; »	.........		. - : .„: \'
.. .		N	;; .\' : :r-
	.... ■ ■ - =		... : ■
	. .. : .-gt;	lt;:t ,,,	. \'. }
	. . . . ^-\'.i:- . \'. N\'-r .
	i

	America	Asia	Australia 1	Africa
Bignonieae • • •	52	3	__
Tecomeae....	22	17	4	12
Tourretieae ■ • •	1	—		—
Crescentieae • • •	7	—	—	6
Eccremocarpeae . \|	1	—	—	—
Total . . . 1	83	20	4	18