tenet »
Pipe eT eT J

arate rte tL po fia ha thet
shy : ere

ee | 6 3

res ecatin : rot pn oie patirl ye adios pone nae & reqge(in tm
wists mre grernanie tion cpu ree + 2, Ser ptt rt puakten sat rans mt ore aR Pe sun. Nokehmhebat
cree : ~ i 5 ST ora enka ~g-aiel TD abate ee PUR PLE Ne ee vals Lh ote ag Ne te at daieiad ib
‘ ei ge : ‘ 4greeue jello tts TT en te beieall phat ve POT ele al Aeho aruree wt ‘. = afty> . Fo a) ath hate dae .
+. s é ; Aree sal bet ty et + cv ° ' enfin dt. od Liebe a PEPO te PY Ee og is al mele 1h a!
: poet {paper . n J . unt eer os : pe AA LETRLS BENT S
perry te a saw pope pate oer aciadaiht
. Sigel ruekerer
pater rire

Punarsienry ee
LEEPER
Perret! a
PER OR APT sper
Ned
ae

hae Pe ;
pipereeees 4 i $e \" : " .
i ‘ peer eens
Mey a a, abe heer
irdaih-dafepeterency

2 pecan ;
4 alge eer” Fe ie r eee rae *
Ne raskete’ rarer ie Z s Saenenean inne ns ree
part ees . . ed ‘i me: % pou ‘ $ ~“ oF
"1 seh? ? Lt Arpeds 4.2 . : eres eee é ‘ Nad pa a oe eet
faire Fair® wrk im iy in s . se She bheive he ph ty} oP Lary es Ht e
war : ween ™ + ryote toes = a ee hee
* ort ‘ A EERRE CREO
‘ * Bic muae

pn peeree

a os asdoleier
Z : % See eieteih ee ae ells
Lard fe PLP reve -
rarer Ge EM A pn eoties aig yt et.
4 eee nate

esene sy e papacy
FATTY

La tartahot B 6 y Rite i’
Lng ; mega aes
ee

ot eked

purur-r ort .
mijn payed ; siege *
itt < r ; tts rapier ug eet sign

baste tt ask acto! price nef Discs or reat
ety ¢ POT Tee ee aint :

acpadageseceqeneyee

peers

q heoepentrtts
‘ ie

pepe ar zece
Re ee het ‘
fis gos irr niall - : i
Pears ; eietnt sire iberanenryents re foi y pbb bg dare ‘
sadly alah 10 ee neha BAP Revcts PRS) ofieimsal f y M an reget me d hehe ‘ : i ahune
“ . 7 afenr te! Liebe peti eh ter Rf ip P ars sere f 4 age fags that en fates = fevaie S2ncete ei
: a : ; r ee eo mee 5 pe arr arie

are

Pree yel rarest pa rte Hf ; barca? gate $ nEbtatyeatesdsint=p-peer> - TESA STR potest
ob rs aha Copel? ry v" ‘ r- “ iS - : - ~
a : ress eit gehabs oe : wibusst A> : : ve . aay tatty re
ater “ he hy ; m4 : d rite nigeria pats? - jenee nite aie guia am dient : Sate Lepmiiecerapade le 0ei
‘ ¢ rir eee ’ ; yaya get 4 7 f 4 spe persia at i
: y # : ; \! ot ieee sejrtnts ri
: é : c ts Fagen yee iw ® Rayegueatsaeoeiote:
Tue) ie ate : - reer:
so Lohan ; waged -7eeet

ngebriciegeiagmameas
wie feouee marr eee
| site

*
j wiaonedys t
+ peetassme tit eins
1 lied elit
ofin bef reeks

me ie
shapnsepe? =)?

jayoneany
ake inte B
gehen ®
gta it ¢
sage eto epainnt
ge a jes tthe
yapu segatin joner = pene

Net
peperrt® imagers’
nee

vera Ae
she petal e = 1s pre

Library

Arnold Arboretum

Harvard University

2
~~ hee
a3 ba |

ee

THE

GARDENS’ BULLETIN

STRAITS SETTLEMENTS (Nos. 1-3)
SINGAPORE (No. 4)

Volume XI
May, 1939—September, 1947

To be purchased at the Botanic Gardens, Singapore

Published by Authority.

da AT THE Covesnuene 3 PRINTING OFFICE, paren
y V. C. G. GATRELL, GOVERNMENT PRINTE

1949

Y:

Ves

2

ol
t

+

WY ki:

Sule

:
*
-

THE GARDENS’

BULLETIN

STRAITS SETTLEMENTS (Parts 1—3)
SINGAPORE (part 4)

INDEX TO

VOLUME XI

New Binomials are printed in Bold-faced type and synonyms

in italics.

Abdominea macrantha CaRR, 284.

Acriopsis Carrii Hour, sp. nov., 275.

Adenanthera pavonina L., _leaf-
change in, 130.
Aerides decumbens GRIFF., 286
Alocasia:
indica (LouR.) ScHort., 252.
var. diversifolia ENGL., 255.
var. heterophylla ENGu., 255.
var. metallica ScnuotrT 254.
var. rubra (HAssK,) ENGL., 255.

var. variegata (Kocu & BoucHue)

ScHoTtT., 253.
var. violacea ENGL., 254, 255.
macrorrhiza (L.) SWEET., 252.
and its Varieties, 244.
confusion with Alocasia odora,
248.
confusion with Colocasia and
Xanthosoma species., 250.
Typification of the Species, 245.
Varietates Dubize vel Exclu-
dendz, 255.
Varieties of, 252.
Portei (ScHottT) ENGL., 255.
talihan ELMER, 255.
variegata Kocu & Boucue, 253.
Amendments Proposed to the Inter-
national Rules of Botanical Nomen-
clature (1935), 1.
Anisoptera megistocarpa V. S1i.,
leaf-change in, 130.
Anoectochilus:
duplex Hotrt., sp. nov., 275.
macranthus Ripu., 278.
Andropogon nemoralis BALANSA, 297
Arum:
indicum LouR., 252.
macrorrhizon L., 252.
maximum macrorrhizon zeylanicum
HERoM., 252.
Ascocentrum micranthum
Hotrt. comb, nov., 275.
Ascochilus:
capricornis Ripu., 287.
minutiflorus RIDu., 289.

(LINDL.)

Vol. XI (1947).

-Aspidium Prescottianum WaALL., 94.

Asplenium:
heterophlebium MEtv., 87.
sorzogonense PRESL, 102.
subserratum Bu., 104.
xiphophyllum BAKER, 106.
Athyrium angustisquamatum Hotrr.
sp. nov., 273.
Axonopus:
affinis CHASE, 109.
compressus BEAvv., 109. °

Beccaria Furrapo sect. nov., 37.

Bolbitis singapurensis Hourr. sp.
nov., 271.

Bonia Furrapo sect. nov., 36.

Bulbophyllum (§Micromonanthe)
nigromaculatum Hotrt. sp. nov.,
276.

Czesalpinia ferrea Marrt., leaf-change
in; E31.

Calanthe:
cleistogama Hotrt. sp. nov., 277.
gracilis LINDL., 286.
johorensis Hort. sp. nov., 277.
longipes Hx. F., 286.

Callipteris sylvatica Bory, 99.

Calophyllum calaba L., Typification
of., 258, .

Camarotis adnata
comb. nov., 277.

Canarium rufum BENw., leaf-change
in, 132,

Carpet Grass, Axonopus spp., 109.

Cassia:
fistula L., leaf-change in., 132.
nodosa Bucw. Ham., leaf-change

in., 133)

Cedrela Glaziovii C.DC,, leaf-change
in., 134. .

Ceiba pentandra (L.) var caribea
(DC.) Baxu., leaf-change in, 135.

Chrysopogon nemoralis (BaLANSsA)
Hott. comb, nov., 297,

(Ripu.) Hott.

540

‘

Cleisostoma:
zonosmum RIpDuL., 288.
lanatum LINDL., 288.
micranthum Kine & PANTL., 276.

Ceelogyne (§Longifoliz) stipitibul- -
bum Ho trv. sp. nov., 278.
Colocasia:

indica (LouR.) Kunta, 252.
indica var. atroviridis HAssK., 255.
indica var. pallida Hassk., 256.
indica var. rubra HaAssx., 255.
indica var. viridis HASSsK., 256.
odorata var. rubra Hassk., 255.

CorNER, E. J. H.: Notes on the
Systematy and Distribution of
Malayan Phanerogams, IV: Izora,
aT i.

Corybas caudatus Hotrv. sp. nov.,

Couroupita guianensis AUBL., leaf-
change in, 135.

Cratoxylon formosum (JAcK) DYER,
leaf-change in, 136. ;

Ctenopteris sparsipilosa Hotrr. sp.
nov., 268.

Cyclosorus ecallosus Hour. sp. nov.,
269.

Cystopus macranthus (Hx. F.) Hotrr.
comb. nov., 278.

Cystorchis gracilis (Hx. F.) Horr.
comb. nov., 279.

Dammera Brcc., 37.

Dammera (BEcc.)
nov., 37.

Dammera K. ScHum. & LAUT., 37.

Delonix regia (BoJ:) Rar., leaf-
change in, 137.

Dendrobium:

Furtapo Stat.

(§Calcarifera) brinchangensis Horr.

sp. nov., 279.
(§Conostalix) melanochlamys
Ho.ttr. nom. nov., 280.
Dendrobium villosulwm WaALL.apud
Hx. F., 280.
Dendrocalamus:
dumosus_ (RIpDL.)
nov., 296.
elegans (RIDt.)
nov., 286.
sinuatus (GAMBLE) Hotrr. comb.
nov., 296.
Dendrochilum:
carnosum (Ripu.) Hortr. camb.
nov., 280.
gramineum (Ripu.) Horr. comb.
nov., 280.
lineare (RipL.) Horr. comb. nov.,
280.
Diplacrum reticulatum Horr. sp.
nov., 295,

Horr. comb.

Hottr. comb.

Diplazium:
in the Malay Peninsula, 74.
accedens Bu., 81.
asperum Bu., 93.
bantamense Bt., 82.
cordifolium Bt., 83.
crenatoserratum (Bu.) Moore, 84.
dilatatum Bu. 85.
esculentum (RrEtz.) Sw. 86.
heterophlebium (MEtTT.) DIELs, 87.
insigne HOLTT., 87.
Key to Species of Malaya, 78.
Kunstleri Horr. sp. nov., 88.
latisquamatum Ho trr., 90.
malaccense PRESL, 91.
montanum v.A.vV.R., 92.
polypodioides BL., 93.
Prescottianum (WALL.) Moors, 94.
procumbens Ho trv. sp. nov., 95.
riparium Hotrt. sp. nov., 97.
simplicivenium Hotrv. sp. nov., 100
sorzogonense PRESL, 102.
sorzogonense var. major BEDD., 103.
speciosum Bt., 103.
var. major Hotrt. stat. nov., 103.
subintegrum Ho.urt., 104.
subserratum (Bui.) Moore, 104.
sylvaticum (Bory) Sw., 99.
tomentosum Bt., 105.
velutinosum Hottrt., 106.
xiphophyllum (BAKER) C. CHr.,106
Doritis Wightit Bentu., 286.
Dryopteris (§Polystichopsis) Haniffii
HOLrTrT. sp. nov., 272.
Dyera costulata Hx.F., leaf-change
in, 138.

Elaphoglossum:
brevifolium Hotrv. sp. nov., 270.
peninsulare Hotrv. sp. nov., 270.

Eria (§Hymeneria) clavata Hort.
sp, nov., 280.

Erythrina lithosperma Mriq., leaf-
change in, 139.

Eugenia (§Syzygium):
atronervia HENDERSON Sp. nov.,299
Brantiana HENDERSON §p, nov.,313
camptophylla HENDERSON Sp. NOov.,

= ie he

cerina HENDERSON Sp. nov., 322.
Eugenia §Fissicalyx Henderson sect.
nov., 333 eae
Eugenia (§Syzygium):
flosculifera HENDERSON Sp. HOvV.,
329.
Haniffii HENDERSON sp. nov., 309.
kemamanensis HENDERSON Sp. NOV.
303.

Gardens Bulletin, S.

o41

Eugenia (§Syzygium):
Kiahii Henperson ©. nov., 307.
Millsi Henpexson o3. noyv., 301.
nemestrina HenprErson sp. nov.,
roots
Eugenia, New Species of, 299.
Eugenia (§Syzygium):
Ngadimaniana HENDERSON Sp. nov.
305.
prestigiosa HENDERSON Sp. NnOvV.,
318.
pseudeclaviflora HENDERSON sp.
33.

pseudosyzygioides HENDERSON Sp.
nov., 315.

punctulata KiInG, 322.

quadribracteata) HeNDERSON sp.
nov.,

(§Fissicalyx) Symingtoniana
HENDERSON §)p. fdv., 333.

(§Syzygium) taipingensis HEnDER-
SON Sp. NOV., 327.

(§Fissicalyx) Watsoniana HENDER-
SON, Sp. Nov., 336.

Eu-Licuala Drupg, 36.
(§SWurmbia) FurtTapo, 44.

Fagrea fragrans Roxs., leaf-change

an, 139.

Ficus:
caulocarpa Miq., leaf-change in, 140
variegata Bu., leaf-change in, 141.
Flamma®*silvarum Rumpu., 194.
peregrina RuMPH., 183.
FurRTADO, C. X.: Alocasia macrorrhiza

and its Varieties, 244.

Amendments Proposed to the
International Rules of Nomen-
clature (1935), 1.

Palme Malesice VIIi: The Genus
Licuala in the Malay Peninsula,
ot.

Palme Malesice, IX: Two New
Corpyhaceous Genera in Malaya,
236.

The Typification of Calophyllum
calaba L., 258.

Goodyera gracilis Hx. F., 279.
Grammitis crispatula Hour. sp. nov.,
~ 268.

Habenaria sumatrana SCHLTR., var.
major Hotrt. var. nov., 281.

’ Hemionitis esculenta REetz., 86.

HENDERSON, M.R.: New Species of
Hugenia Linn. (Myrtacez), 299.

Vol. XI (1947).

Heritiera:
elata RiDt., leaf-change in, 142.
macrophylla Wat.., leaf-change

in, 143.
Hevea brasiliensis M.A.,
in, 143.

Hourrum, R.E.: New Species of Vas-
cular Plants from the Malay
Peninsula, 267.

Periodic Leaf-Change and Flower-
ing of Trees in Singapore II, 119.

The Fern Genus Diplazium in the
Malay Peninsula, 74.

Homalium grandiflorum BENTH.,
leaf-change in, 145.

Homonymas, 4.

Hybrids, status of names of, 10

Hymenza courbaril L., leaf-change
in, 146.

leaf-change

Illegitimate names, 2.
Impriorable names, Il.
Index:
Botanical names of Licuala, 73.
Collectors numbers of Diplaziwm,
108.
Collectors numbers of Licuala, 72.
International Rules of Nomenclature
(1935) Amendments Proposed to 1.
Invalid names, 1.
Ixora:
affinis Don, 223.
var. arguta CRAIB, 223.
var. plumea CRAIB, 223.
amcena Don, 182.
amoena Don, 206.
arborescens HASSK., 197.
arguta KInG, 223.
aurorea RiptL., 206.
var. major. R1ipu., 216.
bandhuca Roxs., 185.
Blumei Zouu. & Mor., 197.
Brunonis Don, 183.
calycina THw., 232. _
candida Ript., 226.
chinensis LAm., 183.
clerodendron Ripu., 185.
coccinea L., 185.
Key to Forms and Varieties of
188.
var. auree-rosea
nov., 186.
var. decolorans CoRNER Var. nov.

_ var. bandhuca, 185.
var. lutea, 186.
- var. rosea CoRNER Var. ine 186.
concinna Hk. F., 188.

ee

CORNER Var.

342

Ixora:
congesta Roxs., 189.
Key to forms of, 190.
coriacea Ripu., 197.
crassifolia Ripu., 197.
elliptica Ripu., 198.
erubescens Don, 223.
finlaysoniana Don, 193.
fluminalis Ripu., 197.
fulgens AucttT., non Roxs., 206.
fulgens Roxs., 194.
fulgens Roxs., sensu HK. F., 189,

216.
gigantea Rivt., 198.
grandiflora BrR., 185.
grandifolia ZoLu. & Mor., 197.

Key to Forms and Variety of, 203

var. acutifolia Hocur., 198.

var. arborescens Hx. F., 198.

var. coriacea HK. F., 198.

var. gigantea K. & G., 198.

var. Kurzeana HK. F., 198.

var. lancifolia CoRNER var. nov.,

198, 205.

var. rosella Hk. F., 198.
Griffithii Hoox., 189.
humilis K. & G., 230.

var. Scortechinii Ripu., 230.
javanica.(Bu.) DC., 206.

Key to the Varieties of, 206.

var. multinervia CorNER var.

nov., 206.
Key to the Forms of, 208.
var. paucinervia CorNER var.
nov., 206.
Key to the Forms of, 209.
var. retinervia CoRNER var. nov.,
206, 207.
Key to Species in Malaya, 179.
Kingiana Crars, 230
Kingstonii Hx. F., 215
Korthalsiana Kurz, 232.
Korthalsiana (Miq.) BoERL., 232.
lacuum BREMEKAMP, 207.
lancifolia Ripu., 198.
Lobbii K. & G., 216.

Key to the Forms of, 218.

var. angustifolia K. & G., 216

var. Stenophylla CoRNER,- var.

nov., 216.
Key to the Forms of, 218.
lucida HK. F., 224.

var. densipila CRAIB, 224.
lutea Hutcu., 186.
macrosiphon Kurz, 197.
micrantha Ripu., 222.

Key to the Forms of, 222.
montana Ripu., 226.
multibracteata K. & G., 234.
nigricans W. & A., 223.

var. arguta Hx.F., 223.

var. erubescens Kurz, 223.

.

Ixora:
var. ovalis Pirarp, 224.
Notes on the Systematy and Distri-
bution of, 177.
odorata (Bu.) K. & V., 197.
opaca Don, 226.
ovalifolia BREMEKAMP, 207.
Parkinsaniana Crats, 226.
patens RIDL., 222.
pendula Jack, 226.
var. opaca Ripu., 226.
plumea RIDL., 223.
pseudojavanica BREMEKAMP, 207.
pumila Ripu., 228.
Ridleyt MERR., 198.
Robinsonii Ripu., 229.
rosella Kurz, 197.
salicina Ripu., 216.
Scortechinii K. & G., 230.
Key to the Varieties and Forms
of, 231. . :
var. Caulescens CORNER Var. NOV.,
230.
stenophylla (KORTH.) KUNTZE sensu
Ript., 216.
stricta AucTT. non Roxs., 206.
stricta RoxB., 183.
var. amoena RIDL., 182, 206.
Sunkist CorNER, nom. hort. nov.,
235.
umbellata K. & V. 232.
var. multibracteata CoRNER var.
nov., 234.
undulata Roxs., 234.
Valetonit Hocur., 198. =

J AGOE, R.B.: Carpet Grass, Aronopus
spp., 109.

Kigelia pinnata DC., leaf-change in,
146.

Kingiella decumbens RoLFE, 286.

Koompassia malaccensis, leaf-change
in, 147.

Kuhlhasseltia Carrii HoLrr. sp. nov.,
281.

Lagerstremia flos-regine RETz.,
leaf-change in, 148. “
Lecythis sp., leaf-change in, 149.
Legitimate names, 2.
Lepironia:
cuspidata Miq., 293.
humilis Miq., 293.
Liberbaileya Furrapo gen. nov., 238.
lankawiensis FuRTADO sp. nov., 238
Libericula Furrapo subgen. nov.,
36, 44.
Licuaceila Brcc., 36.

Gardens Bulletin, S.

343

Licuala:
acutifida MART., 44.
acutifida RIDuL., 66.
acutifida Brcc. var. peninsularis
Brcc., 44.
Analytical Key to Species, 38.
Beccariana FurTADO nom. nov., 37.
confusa FurrApDo §p. nov., 44.
Corneri FuRrTADO sp. nov., 47.
fatua Brecc., 37.
ferruginea Brcc., 47.
glabra GRIFF., 49. |
var. selangorensis Brcc., 50.
kemamanensis FurTADoO sp. nov.,

Kiahii Furrapo sp. nov., 52.
Kingiana Becc., 54.
Kunstleri Brcc., 54.
lanuginosa RIp.L., 54.
longepedunculata Ripu., 49.
longicalycata FURTADO Sp. nov., 54.
longipes GRIFF., 57.
lengipes Ripu., 54.
malajana Brcc., 58.
malajana Rivu., 63, 66.
mirabilis FURTADO sp. nov., 58.
modesta Brcc., 60.
Moyseyi FurtTapo sp. nov., 61.
pahangensis FurtTADO sp. nov., 63.
paludosa GrirF., 65.
paniculata Ripu., 65.
patens RivDu., 65.
peltata Roxs., 36, 44.
penduliflora (Bu.) M1q., 38.
pusilla BEcc., 66.
pygmea MERR., 68.
ramosa Bu., 37.
ramosa (K.Scuum. & LAutT.) Brcc.,
5
reptans BEcc., 37.
Ridleyana Bercc., 66.
Scortechinii Brcc., 67.
spinosa RIDL., 63, 65.
spinosa WURMB., 36, 67.
Subgenera and Sections of, 34.
tiomanensis FuRTADO sp. nov., 68.
triphylla GRIFF., 68.
var. integrifolia Ript., 70.
Wrayi Brcc., 60.
Licualina BEcc., 36.
Licualopsis Becc., 36.
Limatodis pallidus Ripu., 286.
Liparis (§Platystylis) hirundo Hourr.
sp. nov., 281.
Livistona rupicola Ripu., 240.

Malaxis:
macrochila (RoLFE) Hourr. comb.
nov., 282.
nemoralis (Ripu.) Hotrr. comb.
nov., 283

Vol. XI (1947).

Sy

Malaxis:
perakensis (Ripu.) Hourr. comb.
nov., 283.
reniloba (Carr)
nov., 283.
stenophylla HoLrv. sp. nov., 283.
Malleola:
altocarinata Hour. sp. nov., 283.
macranthera (Ripu.) Hotrr. comb.
nov., 284.
Mangifera:
foetida Lour., leaf-change in, 130.
indica L., leaf-change in, 151.
Mapania:
cuspidata (Miq.,) Howrr.
nov., 293
insignis HoLrr. sp. nov., 293.
lucida N.E.Br., 293.
micropandanus Ho trv. sp. nov.,294
petiolata var. cuspidata Uirtren,
293.
triquetra Ripu., 293.
Maxburretia Furrapo gen. nov., 240.
rupicola (Ripu.) Furrapo comb.
nov., 240.
Microstylis:
macrochila Rou¥rE, 282.
nemoralis Ripu., 283.
perakensis Ripu., 283.
reniloba CARR, 283.
Mr. J. C. NAUEN, Obituary, 266.

Hourr. comb.

comb.

New Species of Eugenia LINN.,
(Myrtacez), 299.

New Species of Vascular Plants from
the Malay Peninsula, 267.

Nomina Ambigua, 7.

Nomina Conservanda aut Rejicienda,
re

Notes on the Systematy and Distri-
bution of Malayan Phanerogams,
IV: Ixora, 177.

Oberonia:
calcicola Horr. sp. nov., 284.
condensata Ripu., 285.
flabellifera Horr. sp. nov., 284.
transversiloba Hotrr. sp. nov., 285.
Obituary: Mr. J. C. NAvEN, 266.
Ochlandra Ridleyt GAMBLE, 296.
Octarrhenacondensata(Ript.) Horr.
comb. nov., 285.
Odontochilus macranthus HkK.F., 278.
Orthography, 4.

Oxytenanthera sinuata GAMBLE, 296.

344

Palme Malesice, VIII: The Genus
Incuala in the Malay Peninsula, 31.
Palme Malesice, IX: Two New
Coryphaceous Genera in Malaya,

Parishia Maingayi H«.F., leaf-change
hi Wag $597 4s ;

Parkia javanica (LAM.) MERR., leaf-
change in, 153.

Paspalum:
compressum Sw., 109.
platycaule Porr., 109.

Pavetta:

acutiflora KortH., 223.

calycina Kortu., 232.

fulgens Miq., 194.

incarnata Bu., 186.

javanica Bu., 206.

Korthalsiana Miq., 232.

macrocoma Miq., 197.

macrophylla Bu., 197.

odorata Bu., 197.

subulata TEysmM. & BINN., 223.

Teysmanniana Miq., 197.
Peltophorum pterocarpum (DC.)

Backer, leaf-change in, 153.
‘(Penang Lawyer’’, 44.
Pennilabium acuminatum

Ho.urt. comb. nov., 285.
Pericycla Bu., 37.
Pericycla (Bu.) DrupE, 37.
Pericycla (Bu.) R1ipu., 37.
Pericycla penduliflora Bu., 38.
Periodic Leaf-Change and Flowering

of Trees in Singapore (II), 119.

Phaius:
longipes (Hx.F.)
nov., 286.
pallidus Ripu., 286.
pauciflorus Bu.var. pallidus (R1pt.)
Hour. Stat. nov., 286.

Phaleznopsis:
decumbens (Grirr.) HoLtrr. comb.
nov., 286.
hebe Rows. F., 286.
wight Rows. F., 286.

(RIDL.)

Hourt. comb.

Pholidota longibulba HoLtt, sp. nov.,

286.
Phymatodes subfasciata Horr. sp.
nov., 267.

Platyclinis:
carnosa Ripu., 280.
graminea RIDL., 280.
lineare Ripu., 280.
Pomatocalpa setulense(Ripu.) Horr.
comb. nov., 287.
Priorable Names, 1
Pterocarpus indicus WILLD.,
change in, 154.

leaf-

Representative Botanical
tions, 13.

Rhynchostylis gigantea (LinDt.)
Ripu. var. Harrisoniana (HK.)
Hourtt., 287.

gE a foveolata Horr. sp. nov.,
297.

Rainfall in inches at the Botanic
Gardens Singapore:

Institu-

1927 <2 Gs
1928. So oe
1929 .. 165
1930... 166
O2E en OUDGT
$932. 37-168
1933...%2>.468
1934 0.) FO
1935 71
1936 (eo eZ
193% %: ses
1986. co - 174
1939. ea SS
Saccolabium:

adnatum Ripu., 277.
fissum Rivu., 276.
Harrisenianum Hx., 287.
tonosmum Ript.., 288.
macrantherum Ripu., 284.
micranthum LInDL., 275.
miserum Ripu., 292.
rugulosum RipDu., 288.
setulense RIDL., 287.
tenuicaule HxK.F., 292.
Salmalia malabarica (DC.) Scuorr &
ENDL., leaf-change in, 155.
Saraca taipingensis CANTLEY, leaf-
change in, 156.
Sarcanthus:
bracteatus RiDu., 288.
capricornis (RipL.) Hoxrr. comb.
nov., 287
inflexilobus Horr. sp. nov., 288.
ionosmus (Ripu.) Hoxrrr. comb.
nov., 288. ;
lanatus (LINDL.)
nov., 288.
rugulosus (Ripu.) Horr. comb.
nov., 288.
Sarchochilus:
acuminatus Ripu., 285.
adnatus Ripu., 277.
anceps RIDL., 292.
Carrii Hoxrr. sp. nov., 288.
johorensis Hotrr. sp. nov., 289.
minutiflorus (RipL.) HoLrr. comb.
nov., 289.
scopa Hx.F. 289.
stamensis sensu CARR, 288.
Schizea paucijuga Horr. sp. nov.,
267. :

Hoirr. comb.

Gardens Bulletin, S.

Schizostachyum
dumesum Ripu., 296.
elegans Ripu., 296.
Ridleyi (GamsLe) Hoxrtrr. comb.
nov., 296.
Scleria cyathophora Hourr. sp. nov.,
294.

Sindora Wallichii
change in, 157.
Spodiopogon velutinus Horr. sp.

nov., 297.
Sterculia:
carthaginensis CAV.,
m,. 15ST.
macrophylla
in, 159.
sp., leaf-change in, 160.
Syzygium incarnatum MERR. & PERRY,
322.

BEnTH., leaf-

leaf-change

VENT.,

Tenitis dimorpha Hott. sp. nov.,
274.

Tamarindus indica L., leaf-change
in, 160.

Tautonyms, 5.

Terminalia:
catappa L., leaf-change in, 160.
subspathulata Kine, leaf-change

in, 161.

The Fern Genus Diplazium in the
Malay Peninsula, 74.
The Singapore Botanic
during 1941-46, 263.
The Typification of Calophyllum

calaba L., 258.

Gardens

Vol. XI (1947).

leaf-change

345

Thelypteris herbacea Hotrr. sp. n ov.
268.
Thrixspermum «,
arachnites qaoad RIpt., 289.
(§Dendrocolla) brevicapsularis
Ho.rr. sp. nov.,
290.
Corneri Hotrv. sp.
nov., 292.
duplocallosum
HOLTT. sp. nov.,
290.
platycaule Hotrr. nom. nov., 292.
aio (Hx.F.) Hotrr. comb. nov.,
89.
Trichoglottis misera (Ripu.) Hourr.
comb. nov., 292.
Types, Nomenclature of, 13.

.

Uncifera tenuicaulis (Hx.F.) Hourr.
comb. nov., 292.

Valid Names, 1.

Vetiveria nemoralis A. Camus, 297.

Vittaria angustissima Horr. sp. nov.,
274.

Wurmbia Furrapo cect. nov., 36, 44.

Xanthosoma violaceum, 255.

*

"

, a

‘a
ee

CONTENTS

Part 1: 31st May, 1939 (pp. 1—30).

Furtapo, C. X.: Amendments Proposed to the Inter-
national Rules of Botanical Nomenclature (1935)

Part 2: 29th November, 1940 (pp. 31—176).
Furtapo, C. X.: Palme Malesice, VIJI—The Genus
Licuala in the Malay Peninsula

Hoxutrum, R. E.: The Fern Genus Diplazium in the
Malay Peninsula ..

JAGOE, R. B.: Carpet Grass, Asis spp.

Ho.trtum, R. E.: Periodic Leaf-Change and d Flowering
of Trees in Singapore

Part 3: 30th August, 1941 (pp. 177—260).
Corner, E. J. H.: Notes on the Systematy and Distri-
bution of Malayan Phanerogams, [V—Iyrora

Furtapo, C. X.: Palme Malesice, IX—Two New
Coryphaceous Genera in Malaya

Furtapbo, C. X.: Alocasia macrorrhiza and its Varieties

Furtapo, C. X.: The sa alia of alee i a
calaba Ais

Part 4: 30th September, 1947 (pp. 263—338).

The Singapore Botanic Gardens during 1941—46

Mr. J. C. Naven, Obituary :

Ho.rrttum, R. E.: New Species of Vascular Plants oe
the Malay Peninsula

Henverson, M. R.: New Species of Eugen tie:
(Myrtacez)

PAGE

31

74
109

119

Lie

236
244

258

_. GARDENS’ BULLETIN

STRAITS SETTLEMENTS

| Vol. XI... 31st t May, 1939 ——s~Part 1
CONTENTS
PAGE

Amendments Proposed to the International Rules. of
Botanical Nomenclature (1935) by C. X. Furtado—

1. Introductory Remarks - - Ee 1
2; Amendments. - : - : 13 i

To be purchased at the Botanic Gardens, Singapore |
: Price $1.00 or 2s. 4d. "

is . Published by Authority |
aa -2 |
; SINGAPORE: |

; Printep at tHe GoveRNMENT PRINTING Office, SINGAPORE,
sy W: Ty CHerry, rea BN PRINTER,
RBORr> .

QLO-ASSORE,,

1989 RECEIVED uu |
te . ’ H
sone ; is Sets |

; LigrRar®,

THE

GARDENS’ BULLETIN

STRAITS SETTLEMENTS

Vol. XI _ 31st May, 1939 Part 1

AMENDMENTS PROPOSED TO THE INTERNATIONAL
RULES OF BOTANICAL NOMENCLATURE (1935)

by C. X. FURTADO,
Botanic Gardens, Singapore.

1. Introductory Remarks

1. In proposing amendments to the Rules, I make
the following distinctions in names: valid and invalid,
_ priorable and impriorable, legitimate and illegitimate.

2. VALID AND INVALID. Names that are invalid
have no status under the Rules and therefore no claim to
recognition by botanists (Art. 19A of present amend-
ments*). Such names are easily recognized by their defects
in one or more of the points enumerated in Art. 2B (for
exception see para. 3 below). The names that satisfy the
Rules on all the points concerning validity are called valid
(Arts. 2B and 19a).

3. The main and really important difficulty in
connection with judging the validity of names concerns the
quality of their description, which, while it may be consider-
ed by some as sufficient to validate a name, by others may
be judged insufficient (see FURTADO in Blumea, Suppl. I,
1937 pp. 128-132). But provision has been made in these
amendments for empowering the Congress to declare, by
means of a decision, the validity or invalidity of such
questionable names (Art. 21B-—b), so that they may not
“throw science into confusion” (Art. 4).

4. PRIORABLE AND IMPRIORABLE. It is admit-
ted in the Rules that the use of certain valid names “may
cause error or ambiguity” in nomenclature (Art. 4). Such

} * Where letters A, B, etc. follow the number of Articles quoted
in this discussion, the reference is to the present proposed amendments.

1

2

names are rejected simultaneous homonyms, later homonyms
of equal rank and later non-typonymous (formal) homo-
nyms of unequal rank (Arts. 614 & B). These are the only
valid names (besides those on which a special decision has
been made by the Congress) that I consider should be
impriorable, that is, they must not be included in priority
considerations (Arts. 56A, 614A & B).

All other kinds of valid names, including those
published in violation of the priority rule (Art. 56A), can
be used without causing “error or ambiguity” in nomen-
clature, provided they are well typifiable. These should
in my opinion be included in priority considerations
(priorable names).

If not certainly typifiable, or if their inclusion in
priority considerations would cause either confusion or
many undesirable changes in the current nomenclature of
plants, the Congress is empowered to suspend in such cases
the rule of priority and prevent the changes. (Arts. 21
A—D and 62A).

No doubt the existence of well typified or well-typifiable
synonyms causes much inconvenience, but under the new
Rules they must be tolerated as a necessary though burden-
some legacy from the past. As it will be seen later, this
burden is made even greater by giving a valid status to
such synonyms while denying them any claim to priority.
Moreover the priority rule decides the legitimacy of names
(vide infra), and its violation can render a name only
illegitimate (vide infra). The priorability of names should
therefore be decided without any reference to the priority
rule, for this is applied only to such names as are priorable;
hence it is fundamentally unsound to make priorability of
names dependent in certain cases on the application of the
priority rule.

5. LEGITIMATE AND ILLEGITIMATE. Names
that are priorable may, under the present ruling, be either
correct (legitimate) or incorrect (illegitimate), according
to the circumscription, position and rank of the taxonomic
group. A taxonomic group in given circumstances can bear
only one legitimate name, the one that satisfies the priority
rule (Arts. 164 and 56A). All other synonyms of the group
under the circumstances are illegitimate names for the
group.

6. A name published in contravention to the principles
embodied in Arts. 16A, 17A—B, and 56A cannot be the oldest
priorable name if the circumscription, position or rank given
by its author is correct. Consequently occasions when it
would be possible to legitimize such a name, or its epithet,
will not often arise; but they may occur for instance when

Gardens Bulletin, SS.

3

older priorable names or epithets have later become un-
usable either in the required position or according to
subsequent correction of typification or circumscription.
But when such an occasion arises it would be better to allow .-
the name or its epithet (if unambiguous and not to be
abandoned because of a special decision of the Congress)
to be legitimized on the principles of economy and seniority,
as embodied in the priority rule (Art. 56A), than, by
denying priorability to such a name, to complicate the
application of the rules both of priority and of priorability.
For the greatest objection against outlawing unambiguous
valid names published in violation of the priority rules
arises from nomenclatural jurisprudence itself, which, as
shown in Section 4 above, cannot logically allow the
priorability of names to become dependent on the application
of a rule (the priority rule) which is applied only to
priorable names or epithets; the violation of the rule of
priority should only render a name illegitimate in the sense
defined in these amendments.

Moreover no advantage is gained by denying prior-
ability to such unambiguous specific names as are capable of
rendering their later homonyms impriorable: on the
contrary such a denial not only does not maintain the
principle of economy in epithets but also creates a class of
unusable names which are both valid and unambiguous.
No doubt such names were unusable also according to the
1905-1910 Rules; but under those Rules such names were
also invalid and so incapable of rendering their later
homonyms impriorable. The 1935 Rules give valid status
(Art. 87) to names published in contravention to the
priority rule; and it is absolutely essential that the rules
concerning validity, priorability and priority (or legi-
timacy) should not be confused so as to render their
application difficult.

Moreover, as the example discussed below will show,
there is no proper reason why certain offences against the
priority rule should be singled out for penalization and
others escape it.

Example:

In a hypothetical case DoE describes a new monotypic genus
called Codactyla DoE (1920) and publishes a non-validable
specific name C. floribus plenis longipedunculatus DoE (1920).
In 1921 RoE redescribes the genus as Dactycola Rok and renames
the type species of Codactyla as D. multiandra Ror. Later in
1922 Roe describes another species D. quinquefolia RoE. In 1923,
however, Dor claims priority to Codactyla DoE (1920) over
Dactycola RoE (1921) and manifests his predilection to Greek
by renaming the species C. polyandra DOE nom. nov. (C. floribus
plenis longipendunculatus DoE = D. multiandra Roe) and C.
pentaphylla DOE nom, nov. (= D. quinquefolia Roe).

V.ob. XT> (1939).

4

Now here ROE was wrong because the taxonomic group
Dactycola as circumscribed by him included the type of Codactyla
which name he ought to have adopted under one or more of
the Rules not only as the correct name for the genus but also
as the first part of the specific names. Consequently both
D. multiandra and D. quinquefolia are “illegitimate” names under
Art. 2. and the names or the combinations not being “strictly in
accordance with the Rules” cannot be taken into consideration
for the purposes of priority (Art. 45); and so C. polyandra and
C. pentaphylla should be legitimized. But the definition of
“illegitimate” names or combinations in Art. 60 of the 1935
Rules does not cover offences of this category; and, consequently,
under Arts. 60 and 69, D. multiandra and D. quinquefolia are
good priorable names and C. pentaphylla and C. polyandra must
be rejected as “illegitimate” impriorable names.

Now a specific name consists of two parts and in my opinion
there is no justifiable reason why an offence concerning the
second part should be penalized and not the offence concerning
the first part. Under the present Rules all the four specific
names are valid; it is therefore but just that the names should
be made priorable also. Normally the specific epithets created
by DoE (1923) could not be legitimized because they are younger
than the ones created by ROE; but circumstances may arise when
ROE’s epithets become non-legitimable under Codactyla (e.g. when
they have been employed under Codactyla before transference
to it of ROE’s epithets). Under such circumstances rather than
create new specific epithets and complicate the rule of priority,
DoE’s specific epithets polyandra and pentaphylla should be
legitimized.

Dr. T. A. SPRAGUE’S definition (approved by the Amsterdam
Congress) of superfluous names reads thus: “A name is
illegitimate ‘if it was nomenclaturally superfluous when published,
i.e. if the group to which it was applied, as circumscribed by
its author, included the type of a name [or epithet?] which the
author ought to have adopted under one or more of the Rules’
(Art. 60, 1).” Under this definition D. multiandra Ror is a
superfluous name because it includes the type of Codactyla DoE
which ROE ought to have adopted “under one or more of the
Rules” as the first part of the binomial; but D. quinquefolia RoE
cannot (7?) be called superfluous. Whether it was the intention
of the legislators to include D. multiandra ROE as a superfluous
name it is not clear.

7. HOMONYMS AND ORTHOGRAPHY. Owing to
the varying taxonomic opinions of botanists, it is impossible
to apply uniformly the old precept “‘once a synonym always a
synonym.” But the maxim “once a later homonym always
a synonym” has been upheld in all names of equal rank and
in all non-typonymous names of unequal rank where formal
homonymy can occur under these amendments. It is there-
fore now no longer possible to invoke the plea of typonymy
and illegitimacy of an earlier valid name in order to
legitimize a later name having the same spelling and position
and rank as the earlier name (Art. 61A & B).

8. I also submit that the Committee of Nomenclature
might be empowered to investigate into the full consequences
of the rules concerning homonymy and orthography of

Gardens Bulletin, S.S.

5

names and to make some general rules by which botanists
might be able to decide in the majority of cases whether or |
not a certain name is a homonym or a orthographic variant.
I further submit that until this problem is fully reported
upon, proposals for the rejection or conservation of
homonymous names should not be approved. (see FURTADO
in Gard. Bull. Straits Settl. IX, 1937 pp. 249-255). It is
also necessary to decide whether or not the groups of
epithets of the following category are homonymous:
javensis, javanensis, javana, and javanica; annamensis and
annamitica; philippina, philippinensis, philippensis and
philippica; ceylonensis, ceylanica, zeylanica, singalana,
singalensis, singhalensis, sinhalensis; celebica, selebica and
celebesiana; lisbonensis and ulyssiponensis; malayana,
malaiana, malajana, malayensis, malaiensis and malesica;
burmensis, birmanensis, birmensis, burmanica and
birmanica; ete.

9. In the proposed amendments which follow, I have
gathered together all the provisions pertaining a particular
aspect of the names of each kind of taxonomic groups in
order to effect a condensation in the Rules. This condensa-
tion is desirable to prevent the invocation of one rule for
defending the validity of a name which is invalid under
another rule. (In some recent papers Art. 37 has been
invoked to maintain as valid certain generic names which
are invalid under Art. 42). The proposed re-arrangement
would also enable anyone to detect more readily any flaws
that later proposals might introduce into the Rules.

10. TAUTONYMS: I propose the elimination of the
rule of tautonyms because the present definition, or its
interpretation, is ambiguous. It is true that, in 1930 at
Cambridge, tautonyms were rejected as invalid (“‘illegiti-
mate” according to the terminology adopted in 1935 Rules)
and the rule was later confirmed at Amsterdam (1935) ; but
this rejection was apparently made without considering the
status of many names that can be called tautonyms. It
also appears that the intention of 1930 legislation was only
to reject absolute tautonyms like Linaria Linaria, and
Radiola Radiola, but not names like Linaria Linariana,
Radiola Radiolaris and Nasturtium Nasturtium-aquaticum,
though the last name is mentioned in the 1935 Rules as an
example of invalid tautonyms. Moreover the rejection of
absolute tautonyms was not considered in relation to the
rule of the orthography of names (Art. 70); for though
Bradlea Braddleya, like Cuminum Cyminum, may not be
regarded as an absolute tautonym, yet when the generic
names is spelt as Braddleya, which according to Art. 70 is
a mere orthographic variant of Bradlea, then the correction

Vol. XI. (1939).

6

of Bradlea Braddleya would produce the absolute tautonym
Braddleya Braddleya. And vice versa, the absolute
tautonym Braddleya Braddleya, when corrected as to the
generic name, would produce a non-tautonymous name
Bradlea Braddleya.

An absolute tautonym must therefore be defined as a
specific name in which the specific epithet is homonymous
with the generic name. If this definition were admitted
and if it were further admitted that Asplenium Trichomanes
and A. Trichomanes-dentatum are not pairs of homonyms,
then Nasturtium Nasturtium-aquaticum and Radiola
Radiolaris cannot be regarded as absolute tautonyms; and
vice versa, if Nasturtium Nasturtium-aquaticum were
admitted as an absolute tautonym under this definition, then
A. Trichomanes and A. Trichomanes-dentatum would have
to be regarded as a pair of homonyms. It is therefore not
possible to legislate on tautonyms without considering the
rules of orthography and homonymy of names.

The objection that the admission of absolute tautonyms
would involve many changes in the names that have become
long current also holds good in the case of quasi-tautonyms;
for formerly absolute tautonyms and quasi-tautonyms were
treated on the same footing. And perhaps more name
changes are involved by admitting quasi-tautonyms like
Cajanus Cajan, Kandelia Candel, Normanbya Normanbyi,
Sullivantea Sullivantii and Timonius Timon than by
admitting absolute tautonyms.

Furthermore it should not be overlooked that, in the
majority of cases, the existence of absolute tautonyms is a
legacy of a nomenclatural system which did not regard the
economy of epithets in naming taxonomic groups as a
fundamental principle and which therefore permitted, in
naming a new genus, the use of the oldest priorable epithet
of the type species on which the new genus was based.

Now that economy in epithets has become a funda-
mental principle in the new Rules, the correct procedure
should have been to reject the generic names published in
contravention to this principle. But since such a rejection
is now not possible because it would involve many name
changes, the remedy does not lie in outlawing the specific
epithets which, by applying the principles of seniority and
economy to the epithets of unambiguous (priorable)
names, have produced, or would produce, absolute
tautonyms. It would be indeed advantageous to admit the
tautonyms in such cases, for they would help to indicate
the generic types. However, the advisability of restricting
their employment in future only to name the generic types

Gardens Bulletin, S.S

7

might be considered. It may be remarked here that zoolo-
gists have found it useful to admit all tautonyms as valid
legitimable names. |

In view of these considerations it appears to me that
the decision to outlaw indiscriminately all absolute
tautonyms is not only arbitrary but also impracticable and
in many cases disadvantageous to the present nomenclatural
system. The decision is also against the principle which
directs that “the rules of nomenclature should be ........
founded on considerations sufficiently clear and forcible for
everyone to comprehend and be disposed to accept”
(Art. 3).

11. NOMINA CONSERVANDA AUT REJICIENDA.
An amendment to the Rules is here proposed for eliminating
the possibility of securing the conservation of a name under
false pretexts and of then claiming for the conserved name
a force far in excess to that given to it by the Congress.
(Art. 21A4-B). Under the 1935 Rules the conservation of
a name secured merely on orthographical grounds could
have veen used to render non-legitimable or impriorable its
earlier synonyms and homonyms. Under the present pro-
posed Amendments this contingency would not be possible.

12. Further, the generic names rejected vis-d-vis a
nomen conservandum are termed at present nomina
rejicienda; but the names thus rejected can be legitimized
when they are neither homonymous nor synonymous with
any of the conserved names. Yet nomina ambigua and
nomina confusa are also regarded as nomina rejicienda,
though they can never be legitimized. This equivocation
or ambiguity is removed in the present proposals by
adopting a classification which shows the precise effects of
a Congress decision to conserve or to reject a name
(Art. 21B).

- 13. I also submit that the Committee may be
empowered to explore into the problem with a view
to establishing some definite principles by which
to appreciate the proposals concerning the conservation
of names under Art. 21B-c. Until this is done I
submit that no name should be permanently added to the
lists of nomina conservanda made in 1905-1912 except to
protect a use of a name current for a hundred years or
more (see FURTADO in Gard. Bull. Straits Settl. X, 1939
pp. 180-181). The list of 1905-1912 and the names which
have more recently been conserved on orthographic grounds
should be re-arranged so as to conform with the proposed
amendments in Art. 21B.

14. NOMINA AMBIGUA. The rule of nomina
ambigua (Art. 62) is here revised so as to conform with its

Vol. XI. (1939).

8

original meaning. There has been of late a tendency to
attach little or no importance to the word “permanent” in
the rule. This, coupled with the abbreviation of the dis-
cussion given under Rosa villosa, has enabled the newer
interpretations of the rule to pass as equivalent to the correct
and traditional one. But from the discussion given of Rosa
villosa under the 1905-1910 Rules and from the remarks
elicited from eminent nomenclaturists who attended the 1930
as well as the 1935 Botanical Congress, it is evident that
the legislators intended to consider as nomina ambigua only
those names which have not only owing to their use in
different senses, become a source of confusion, but also are
admittedly incapable of any certain interpretation, so that
the impossibility of removing the existing confusion is
generally conceded to be permanent. On the other hand the
Rules provide for disregarding all interpretations (irres-
pective of their number and of the length of their currency)
if they are contrary to the type (Arts. 18 & 50-55). The
existence of certainly identifiable types is also recognized
as the chief means of correcting misinterpretations or
misapplications of names.

15. TWO NEW APPENDICES: In _ 1935 at
Amsterdam a proposal to invalidate certain works was
put before the Botanical Congress by Dr. A. J. WILMOTT;
but the Congress, regarding a statistical investigation into
the effects of the proposal upon the current nomenclature
as an essential preliminary to its consideration, referred the
proposal to a special Committee for the necessary explora-
tion. The task assigned to the Committee is not a light one,
and it is doubtful whether all the necessary data will become
available at the next meeting of the Congress. Hence in
Arts. 21C—D means are proposed by which the principal aim
of Dr. WILMOTT’S proposal might be secured without the
statistical data.

The principal object of the proposal evidently is to
prevent botanists from delving into the works proposed for
invalidation in order to resuscitate generic and specific
names, or their interpretations, such as might lead to many
changes in the long established nomenclature and thus
greatly impede botanical research. An examination of the
opinions expressed on Dr. WILMOTT’S proposal and of
similar ones submitted at previous Congresses by several
botanists (cf. also SPRAGUE, Preliminary Opinions, 1935
p. 5) reveals no objections either to the principle, or to the
main purpose, of the proposal. Consequently both its
principle and its purpose may be taken to have the approval
of botanists in general.

Gardens Bulletin, S.S.

9

The importance of the principle underlying the proposal
lies in the fact that it recognizes implicitly the existence of
some strong reasons why certain works were either
generally ignored, or allowed only a partial validity, by con-
temporary and subsequent botanists, and consequently, that
it is now inconvenient, if not unreasonable, to force the
entire nomenclature from such works upon the modern
botanist, even if he has pledged himself to the rules of
priority and retroactivity.

Just as a provision has been made in Art. 21 of the
Rules to suspend in certain cases, the undesirable effects of
the rules of priority and retroactivity, so also a provision
appears to be necessary against the undesirable con-
sequences of having to accept as valid the works, or some
names in them, which had been generally disregarded by
previous botanists. A suitable provision in the Rules, to
guard against such a contigency, would also induce botanists
to consider the advisability of denying validity to obscure
tracts such as they may have unearthed from a long
botanical oblivion and to prevent them from basing on
ene tracts changes in the long established nomenclature of
plants.

The two proposed new Articles, each with its own
Appendix, would secure this result. Art 21-c practically
embodies the whole of Dr. T. A. SPRAGUE’S suggestion made
in the Preliminary Opinions (1935 p. 8); and Art. 21D is
aimed to secure the invalidation of the long disregarded
names in works proposed for rejection by Dr. WILMOTT.
But it may be remarked that the binomial combinations
adopted in GANDOGER’S Flora E'uropae for the subdivisions
of species must be rejected as invalid, because they are not
formed in accordance with the Rules which prescribe
trinomials or multinomials for the taxonomic groups below
the rank of species (Art. 28). Hence the binomials cannot
be included in consideration of homonymy. Even if the
question of their valid formation were overlooked, their
existence could not be invoked to render impriorable specific
names which are later orthographical homonyms of the
“ternary” binomials; for the rule of homonymy operates
only when the names considered are of the same rank
(Art. 61) or formally of the same rank (Art. 614-8). The
principle involved here is the same which prevents the activa-
tion of the rule of homonymy against generic names when
confronted with homonymous earlier univerbal designations
for species. The fear that GANDOGER’S binominals for the
subdivisions of species, if not invalidated by a special
decision of the Congress, “are likely to necessitate the
rejection of a large and increasing number of names by

Vol. XI. (1989).

10

Art. 61 (ater homonyms)’’ (SPRAGUE l.c.) is, therefore,
unfounded. Hence the necessity of including GANDOGER’S
work in this Appendix X should be further explored.

16. HYBRIDS. The Rules seem to be rather ambi-
guous on the status of the names of hybrids, and current
practice is too contradictory to be of any use in expounding
the correct application of the Rules.

One school of botanists would associate the name of a
hybrid with the taxonomic group represented by the type
and would disregard all proofs and speculations concerning
the hybridity or origin of the group. “The purpose of
giving a name to a taxonomic group,” they would say, “is not
to indicate the characters or the history of the group, but to
supply a means of referring to it” (Art. 15). This conten-
tion accords fully with another fundamental principle in
nomencature which requires every name to be associated
with the type of its description, the ultimate type of every
description being a specimen (Arts. 18 and 50-55).

But another school of botanists would disagree with this
view and would change the name of a hybrid every time a
different ancestry were attributed to the hybrid, even when
the ancestry were different nomenclaturally only and not
taxonomically. This procedure allows not only the associa-
tion of names of hybrids with their parents or ancestors in
place of the taxonomic group represented by the type
specimen, but also the creation of names to cover hypothe-
tical groups which have not been produced at present and
which might never be produced in the future. Thus,
according to this view, the name Cistus Skanbergi LOJAC,
published on the assumption that it represented a good
species, ought to have been changed when TURRILL con-
sidered it to represent a hybrid between C. monspeliensis
and C. parviflorus (Bot. Mag., 1938, t. 9514) ; similarly the
name Amelasorbus Jackii REHDER, published on the assump-
tion that it represented a hybrid between Amelanchier
fiorida LINDL. and Sorbus sitchensis ROEM., ought to have
been changed when JONES considered it to be a hybrid
between Amelanchier florida and S. scopulina GREENE
(Journ. Arnold Arb. XX, 1939 p. 22), even though it may
not be possible to produce any cross between the parents
originally indicated for this hybrid. The reason why the
botanists of this school ignore the fundamental principles
respected by the botanists of the first-mentioned school, is
that hybrids, according to their view, are pseudo-taxonomic
groups and not real taxonomic groups, though for the
purposes of homonymy they claim for the names of hybrids
equal footing with the names of non hybrid groups of the

Gardens Bulletin, S.S.

11

corresponding rank. They defend their contention on the
following provisions:

“All plants raised by crossing the same two species receive -
the same‘specific’ name, variations between the seedlings being
indicated where necessary by varietal names............. me
(Appendix VII-g).

Pe Te> RAY TAGES a in 'olees «Adz ak > between the same two genera
bear the same‘generic’ name” (Art. 32).

A curious anomaly in this procedure is that, while a
proper description based on a type is considered as essential
for the validation of a “specific” name of a hybrid (though
both the type and the description are to be ignored under
this procedure should there be a change in the speculations
concerning the parentage), a similar description is con-
sidered unnecessary, if not misleading, to validate a
“generic” name of an intergeneric hybrid (the formula
indicating the parents being considered as absolutely
essential for its validation).

A third school of botanists would invalidate all names of
hybrids of uncertain origin and quote the following in
support of this their contention:

“If a Latin name has been given to a hybrid form of
uncertain origin which cannot be referred to a Latin binomial,
it must be treated like a vernacular (fancy) name; e.g.
Rhododendron ‘Atrosanguineum’................ ” (Appendix
VIII-f).

Here the rule is also ambiguous: Is the name to be
invalidated only when the hybrid is of uncertain origin, or
when the name cannot be referred to a Latin binomial? If
the former, then the procedure defended by this school of
botanists is justified; if the latter, then the rule is super-
fluous because all non-binomial names of species (real or
hybrid) are invalid. Or does it mean that names are
invalid only when they satisfy both these conditions, so that
a name has to be regarded as valid if it is either a Latin non-
binomial denoting a hybrid of certain origin, or a Latin
binomial denoting a hybrid of uncertain origin?

Obviously such a state of affairs is not conducive to the
stability in nomenclature. And unless new fundamentals
are clearly propounded to defend a contrary procedure, I
submit that the procedure expounded by the first named
school should be admitted as the correct one, because it
complies with the fundamentals of the present nomencla-
ture. Moreover it recognizes the well-known facts that it
is sometimes possible either to produce the same kinds of
hybrids from different pedigrees, or to obtain, from a
progeny of a cross, plants which are exactly identical with
either of the parents; consequently it is unsystematic either
to distinguish the same kinds of plants by different names
merely because they are of different ancestry, or to lump

Vol. XI. (1939).

12

together different kinds of plants under the same name
merely because they have the same ancestry. The proce-
dure recommended here, therefore, disallows a formula
indicating the parents (real or putative) to take the place
of the description required to validate a name under
Art. 37.

The rejection of a formula as insufficient to validate
the technical names of hybrids would also contribute to
restore some order in the nomenclature of hybrids. Some
gardeners, in publishing names to the hybrids raised by
them, have deliberately assigned a wrong parentage in order
to delay the production of the same hybrids by their rivals.
Yet such names become current long before the public come
to know of the true parentage. If the ordinary procedure
of a description in Latin were required in the case of all
valid names of hybrids of all degrees, the names created by
gardeners would be technically invalid, unless they were also
accompanied by appropriate descriptions; and so botanists
would be able to straighten the nomenclature of the hybrids.

However in the case of “generic” names of intergeneric
hybrids, this procedure, if approved, might provoke resent-
ment among horticulturists: firstly, because they have come
to associate such a “generic”? name as a convenient abbre-
viation of a formula denoting the generic status of the
hybrid; and secondly, because many such “generic” names,
although established by reference to their parents only,
have become long current and their rejection now would
cause many name changes. In order, therefore, to avoid
such name changes, it would be desirable to appoint a
Committee to investigate the status of “‘generic’”’ names of
intergeneric hybrids and to recommend their conservation
and typification where necessary.

On the other hand, an asexual “hybrid”, if not a
monstrosity falling within the meaning of the provision in
Art. 65, is at least a compound or composite individual
formed by the fusion or union of somatic parts of two or
more individuals, each often retaining its specific character.
In some respects an asexual “hybrid” is comparable to the
compound individual resulting from a symbiotic or parasitic
association of one organism with another: e.g. the associa-
tion of the parasite Cuscuta with the host Myrica, or of the
scion Achras sapota with the stock of Mimusops hexandra;
and so such an individual cannot be regarded as representing
a taxonomic group and should not have a valid name under
the proposed Art. 22-c (cf also Art. 64). The names of
asexual “hybrids” must therefore be given the same status
as the horticultural names, which have no power to render
their later homonyms impriorable.

| Gardens Bulletin, S.S.

13

However such chimaeras and the “monstrous” hybrids
as are reproduceable by seed should become the subjects of
a taxonomic inquiry even though their reproduction be by
apomitic seed only; and so their names, if duly published by
a description, should be nomenclaturally valid.

If these principles were admitted, appropriate changes
in the wording of Arts. 31-34 would be necessary.

17. TYPES: I also submit that the nomenclature of
types and the regulations to determine them as worked out
by me in Gard. Bull. Straits Settl. IX 1937, pp. 285-309,
may be considered by a special Committee with a view to
investigating the feasibility of their being incorporated in
Appendix I.

18. REPRESENTATIVE BOTANICAL INSTITU-
TIONS. I also submit that the Committee might consider
the utility of working out a list of the Representative
Botanical Institutions under Art. 36 on the principles
elaborated by me in Gard. Bull. Str. Settl. IX 1937
pp. 273-279.

2. Amendments

(Where the letter A follows the number of an Article,
the amendment is to the text of that Article in the existing
Rules. Where other letters are used, the amendments are
additions to or transpositions of the existing Rules).

Art. 2A: DELETE: “Names or forms of nomenclature
contrary to a rule (illegitimate names or forms)
cannot be maintained.”

[This forms a rule by itself; moreover, the definition of
an illegitimate name or form given here does not accord with
its definition given elsewhere].

‘ADD the following to this Article: ‘“‘Where necessary
notes are added in order to make the meaning of the
rules, or of the words used in them, unequivocal and
precise.”

Art. 2B: The rules are divisible into two main classes:
(a) those dealing with fundamentals or validity, and
(b) those dealing with applications or legitimacy.

(a) The rules concerning validity of names are again
divisible into the following groups: (1) the admis-
sible order of the different taxonomic groups;
(2) the formation of botanical names to denote
these categories and their order; (3) the nature
of a description or citation upon which the name
may be based; (4) decisions of the Congress to
meet special cases; (5) the nature of publications
or literature, wherein the descriptions and names
are published; and (6) the dates, the typification
and the interpretation of names.

Vol. XT. (1939).

Art.

Art.

Art.

Art.

Art.

Art.

Art.

14

(b) The rules of legitimacy are divisible into the
following groups: (1) the correct name for a given
taxonomic group under given circumstances; (2)
the correct orthography of names; (3) the correct
gender of names; (4) priorability of names.

134: DELETE the last sentence which is by itself
a rule. (It is embodied in Art. 19A.).

164A: REVISE as follows:

Each taxonomic group with a given circumscrip-
tion, position, and rank can bear only one legitimate
name, the one that is in accordance with the type and
priority principles (cf. Arts. 18A, 19B, 21B, 53A, and
56A).

17B: No one may change, modify, or refuse to accept

a name, or epithet, merely because it is badly chosen, or
disagreeable, or because another is preferable or better
known.

[A name published in contravention to Art. 59 may still
be valid and if so would be included in considerations of
homonymy. Art. 59 is thus 2 commentary on Art. 17 and
should be placed with it].

18A: In the first paragraph instead of “‘see Art. 66”
in brackets READ: “see Arts. 50-52, 53B and
Appendix I.”

19A: REVISE the rule as follows:

Names, descriptions, publications, interpretations,
typifications, or procedures contrary to a rule of vali-
dity are invalid, that is, they have no status under the
Rules, and no claim to recognition by botanists; while
those names, descriptions, etc., that satisfy every one of
the validity rules are valid, that is, they have a status
under the Rules and a claim to recognition by botanists.
Names and descriptions denoting a classification con-
trary to Arts. 10-14 or _ published in _ invalid
literature are not valid.

19B: Rules of legitimacy regulate the use of names
which are valid under Art. 194A. The use of a valid
name is correct or legitimate only if it satisfies every-
one one of the legitimacy rules: otherwise the use is
illegitimate.

20A: SUBSTITUTE for the first sentence the
following:

_. No literature can be valid unless it is validly
published (cf. Art. 364A). Valid botanical literature
begins for the different groups of plants at the dates
and with the books specified below:

Gardens. Bulletin, S.S.

15

Art. 21A: DELETE Notes 3 and 4 and Examples.

[It is desirable that every proposal for the conservation
or rejection of names should be accompanied with an indication
of its full force, so that decisions of the Congress may not
be given a force far in excess to that given by the Congress.
The Congress act on the premises submitted, so that if the
premises are essentially invalid, the decision given under those
premises should also be invalid. Thus, if the Congress were
to conserve Endlicheria NEES (1833) (Lauracee) against
Endlichera PRESL. (1832) (Rubiacez) on the assumption that
these two are homonymous names, this decision should stand
as long as the Rules permit botanists to consider these two
names as homonymous, but not otherwise. If the two names
were not homonymous and if it were still desired to abandon
Endlichera PRESL. (1832) altogether, then a different decision
of the Congress would be necessary and under the text of
Art. 21 motives for outlawing the names would have to be
explained. Further, even if Endlicheria and Endlichera were
admittedly a pair of homonymous names, Endlicheria NEES
(1833), if conserved against Endlichera PRESL. (1832), should
not take precedence over all synonyms and homonyms which
are older than Endlichera PRESL. (1832). Proposals for
conservation or rejection of a name should be accompanied
with the fullest possible details concerning the necessity for
conservation or rejection especially now when there is a
tendency among some botanists to split older genera so as to
render a synonym of one into a legitimate name of another.
It is also undesirable that the conservation of a name made
only on grounds of orthography should be the means to invest
that name with all the power embodied in Notes 3 and 4 which
would render all its earlier synonyms and homonyms impriora-
ble; if the latter was also desired, then the required data
should be presented at the time of making the proposal. The
amendments proposed below (Art. 21B) eliminate these
undesirable elements in the original Art. 21].

Art. 21B: The conserved and rejected names must be
classified in the following groups: (a) Orthographia
nominum approbata; (b) Nomina rejecta aut invali-
data; and (c) Nomina precedenda.

(a) Orthographia nominum approbata. The decisions
given here affect only the orthography of the names
placed in the list, but the dates of their publica-
tion (under the rejected orthography) are retained.
Examples: Thus when Humiria JAUME ST. HIL.
(1805) is conserved against Houmiri AUBL.
(1775) on orthographic grounds, Howmiri AUBL.
(1775) must be spelt as Humiria and its full cita-
tion is Humiria AUBL. (1775) emend. orth.
JAUME ST. HIL. (1805) vice Howmiri.
Similarly, Heleocharis R. BR. (1814) emend. orth.
vice Eleocharis.

(b) Nomina rejecta aut invalidata. Names lose their
valid status when placed here: if used again for

Vol. XI. (1939).

16

the same or another taxonomic group with ade-

quate description, these have the status of new

names.

[Certain names of a doubtful status (e.g. nomina confusa)
and those that are generally ignored as insufficiently described
(nomina semi-nuda, e.g. Giganthemum WELW. 1859) should be
placed here. Placing such names among nomina invalidata
would save a good deal of trouble to botanists and avoid
unnecessary changes in nomenclature on the grounds of priority
or homonymy. Perhaps here could also be included the nomina
ambigua as defined below in Art. 62A].

(c) Nomina precedenda: A name placed by the
Congress in this category acquires a right of
precedence in matters of priority or homonymy
over certain other names specified in the decision
and their later synonyms and homonyms, and over
no others, provided the specified names over
which precedence is given are either synonymous
or homonymous with the nomen precedendum.

Examples: (1) In the Rules Spergularia J. &
C. PRESL. (1819) has been conserved against
Buda ADANS. (1763) and Tissa ADANS. (1763).
This means that either of these names are to
yield precedence to Spergularia whenever they are
synonymous with the latter, but not otherwise.
It also means that all later synonyms of Buda and
Tissa cannot claim priority over Spergularia, even
though they be earlier than Spergularia itself.
But should there be a priorable synonym to
Spergularia which is earlier than Buda or Tissa,
it would be able to claim right of priority over
Spergularia until special decision of the Congress
makes it yield that right to Spergularia.

(2) If it is desired to conserve Hndlicheria
NEES (1833) (Lauracee) against Hndlichera
PRESL. (1832) on the grounds of homonymy, then
the latter must be mentioned vis-d-vis the former.
Such a conservation would make Endlicheria
NEES (1833) take precedence over Endlichera
PRESL. (1832) and its later homonyms, but would

not give EHndlicheria NEES (1833) precedence
over the synonyms or homonyms that are earlier
than Endlichera PRESL. (1832). But this de-
cision would lose its validity as soon as an
amendment were introduced in the Rules to make

Endlichera PRESL. and Endlicheria NEES as non

homonymous names.

Art. 21c: Works listed in Appendix IX are treated as
invalid, because their acceptance would constitute a

Gardens Bulletin, S.S.

17

serious impediment to botanical research. Only works
of doubtful validity, or published posthumously thirty
or more years after the deaths of the authors, or
ignored generally by the contemporaries of the authors,
may be listed in this Appendix.

(1) RAFINESQUE, Autikon Botanikon, 1840;
(2) MOCINO et SESSE, Flora Mexicana, 1888 et 1894;
(3) ibid, Plantae Novae Hispaniae, 1886 et 1893; (4)
TEYSMANN et BINNENDIJK, (Plant Catalogue of
the Buitenzorg Gardens), 1855; (5) Sir WILLIAM
HuNTER, Plants of Prince of Wales Island, (ed. H. N.
RIDLEY), 1909; (6) VOIGT, Hortus Suburbanus
Calcuttensis, 1845; (7) Numerous Theses including
those of some of the pupils of LINNAEUS, THUN-
BERG, etc., Catalogues, Nomenclators and Indexes
which have been ignored in larger works up to very
recent times: e.g. ROXBURGH, Hortus Bengalensis,
1814; MOON, Catalogue of Indigenous and Exotic
Plants of Ceylon, 1824; ZINN, Catalogus Plantarum
hackehrales rats 2 lis Gottingensis, 1757; BURMAN, Index to
RUMPHIUS’S Herbarium Amboinense, 1755 et 1769;
STICKMAN, dissertation on RUMPHIUS’S Herbarium
Amboinense, 1754 et 1759; BERZELIUS, dissertation
on Nomenclator Botanicus, 1759; ete.

[Most of the names from the works listed in (7) should

be invalid according to my interpretation of the Rules—see
Gard. Bull. Straits Settl. X, 1939, pp. 162-172].

Art. 21D: Both generic and specific names published in
the works listed in Appendix X are treated as invalid
if they have consistently been either ignored, or reduced
to synonymy, in monographs and important floristic
works published between 1798 (the date of WILL-
DENOW’S Species Plantarum, Vol. 1) and 1890 (the year
preceding the publication of KUNTZE’S Revisio) both
inclusive; but such names from books listed in
Appendix X as have been either conserved in the Rules
or used in one or more monographs or important
floristic works (1.e. floristic works running to more
than one volume) issued between 1798 and 1890 are
treated as valid. Mere registration of names in
nomenclators and indexes does not constitute their
subsequent use under this rule. Only works published
between 1753 and 1800 both inclusive may be included
in this Appendix X.

[All works proposed for invalidation by Dr .WILMOTT. in
Appendix VIIbis of 1935 are to be included in this Appendix X
provided the reasons given be satisfactory. See Preliminary
Remarks, Sect. 15].

Vol. XI. (19389).

Art.

Art.

Art.

18

228: The long accepted interpretation of a name must

not be disregarded without proofs or adequate botanical
reasons. When retypification is necessary, under
equality of circumstances, the lectotype or neotype
selected must conform with the long accepted inter-
pretation of the name.

Note.—By PROOFS or ADEQUATE BOTANICAL
REASONS are meant reasons why the name must be
attached to a particular taxonomic group and not to
any others, or proofs that the description given under
the name agrees with the taxonomic group newly
identified with it and conflicts with the older
identifications.

254: DELETE: “and written with an initial capital’.
(This belongs to the orthography of names. cf. Art.
70C).

ADD: But no generic name is valid unless: (1)
it is intended definitely as a botanical name, and not
merely a non-botanical designation of the genus; (2)
when coinciding with a technical term currently used
in morphology, it was both published before 1912 and
accompanied simultaneously by one or more duly
validated specific epithets under it; and (3) it is
a univerbal consisting either of a single word or of two
or more words united or hyphened when originally
published.

ADD: Examples (1), (2) & (4) from Art. 67.

[This addition is from Art. 67 which would be better united

with Art. 25. Reference to “unitary designation of species”

is unnecessary, because specific names cannot be generic names,
ef. also Art. i9A and Art. 27A].

26A: READ: “Epithets denoting subgenera and
section” instead of ‘““Names of subgenera and sections”
in the first line; and “Epithets denoting subsections”
instead of ‘‘Names of subsections” in the second line.

ADD: It is permissible to reduce more com-
plicated names to biverbal combinations provided an
appropriate sign or word precedes the epithet to denote
its category.

OMIT: “Agreeing in gender with the generic
name and written with an initial capital”.

[Since the old phrase “specific name” has given way to
“specific epithet”, it is necessary also to abolish the phrases

“subgeneric name’, “sectional name”, etc., because the epithet
denoting a subgenus, section, etc. is not the full name and
cannot stand by itself without referring to the appropriate
genus. The portions referring to gender and orthography are

transferred to their appropriate sections].

Gardens Bulletin, S.S.

19

Art. 27A: REVISE thus:

Names of species are binary combinations con-
sisting of a valid name of the genus to which the species
is referred followed by a single specific epithet.
Symbols forming a part of specific epithets proposed by
LINNAEUS must be transcribed. But no_ specific
epithet is valid unless: (1) it is a univerbal, consisting
either of one word or of two or more words united
or hyphened at the time of publication (an epithet
of two disunited or unhyphened words being allowed
as exceptions in books and papers in which univerbal
epithets have been generally employed); (2) it is
intended definitely to be a botanical specific epithet and
not merely a non-botanical designation; (3) it is not
merely an ordinal adjective being used for enumeration ;
and (4) it is published in a work in which the biverbal
binomial system of nomenclature for species as defined
in the first sentence and the first alinea of this rule has
been consistently employed (an occasional exception
only in any work shall not render such work invalid).

ADD: Examples: (1), (2) and (4) in Art 68,
but delete the last sentence in (4) because Apocynum
folus Androsaemi L. is a good binary binomial like
Adiantum capillus veneris L. and Atropa bella dona L.
discussed in Art. 27).

Art. 28A. OMIT: ‘When adjectival in form and not used
as substantives, they agree in gender with the generic
names.”

[This has been transferred to the section on gender].

Art. 28B: No varietal or subvarietal epithet may be given
to the variety of a species which includes the type.

Different ecologic variations in plants and in parts
of plants belonging to species, subspecies, variety or
sub-variety may be indicated with appropriate epithets
preceded by the words forma and subforma; and
the names thus formed are not reducible to simpler
combinations without intercalating the epithet of the
species, subspecies, variety or subvariety to which the
forms belong.

Note.—When a species is interpreted sensu lato and
it is desirable to indicate the division to which the type
belongs, the type variety of the species may be indicated
by repeating the specific name preceded by the prefix
eu or by the epithet typicus or genuinus but none of
these epithets shall have a status under the Rules.

Examples: The name Nelosuma polynesicum var.
typicum H. J. LAM (Bern. Bishop Mus. Hawaii, Occ.
Pap. XIV, 1938 p. 148) is invalid because it is given

Vol. XI. (1939).

20

to the type of N. polynesicum (HILLEBR.) BAILL.
(1891), but the forme genuinum, longipetiolatum, and
longipetiolatum subforma originarium, published by
Dr. LAM, under N. polynesicum sensu stricto are valid
and are not formal homonyms (Art. 61B) of forme
genuinum, longipetiolatum and genuinum subforma
originarium published respectively under N. polyne-
sicum var. glabrum H. J. LAM (1938).

[The Rules do not admit a ternary system of names for
species; and giving varietal and subvarietal names to the type
variety amounts to giving a ternary name to it. Strictly
speaking forme and subforme are not taxonomic groups,
because the variations may be found even in the same
individual, e.g. the bathyphyll and the acrophyll stages of figs,
ferns and aroids. The names of forme and subforme should
therefore be taken on the same footing as horticultural names
of plants. Under Art. 30 some of Dr. Lam’s forme and
subforme would have to be renamed, while his var. typicum
would be valid. The nomenclatural complications that would
arise by admitting as valid varietal names to the type of a
species have been discussed by me in Gard. Bull. Straits Settl.
IX, 1937 pp. 242-244. That the use of such epithets as typicus,
genuinus, originarius, verus, veridicus, etc. to form varietal
names to the holotype of species produce undesirable complica-
tions have been shown by Dr. F. BoLLe (Notizbl. Bot. Gart.
u. Mus. Berlin-Dahlem, XIII, 1937 pp. 524-530); and in the
appendix to the same paper two divergent opinions, the one of
Dr. T. A. SPRAGUE, and the other of Drs. H. Harms,
J. MATTFELD, and R. PILGER, have been recorded concerning the
status of the epithets of the above mentioned category.
Recently Drs. H. ST. JOHN and E. Y. HOSAKA (Bern. Bishop
Mus. Hawaii, Occ. Pap., XIV, 1938 pp. 118-119) have shown
that the typification of species becomes difficult, if epithets
other than typicus (and perhaps also genuinus, originarius,
verus, and viridicus) were accepted in giving ternary names
to the holotypes of species. They have also tried to prove that,
under Art. 6C (1) Lobelia Gaudichaudii var. coccinea ROCK
(1917) is a superfluous name because it is an exact synonym
of L. Gaudichaudii A. DC. pro parte typica, but they have
not shown how their naming the type of the same species as
L. Gaudichaudu var. typica ST. JOHN et Hos. (1938) with
Rock’s trinomial as a synonym, is justified under the existing
Rules].

Art. 30: DELETE. Incorporated in Art 61B. See
remarks under Art. 28B.

Rec. XVIII: DELETE: Contradicts Art. 28B.
Art. 31-34: SEE Introductory Remarks, Section 16.
Section 5A: READ “valid” instead of “effective’’.

Art. 364A: READ “validated” for ‘effected’, “‘valid’’ for
““effective’’. .

REVISE the 1935 amendment concerning the

separates thus: “The issue of advance separates is not

Gardens Bulletin, S.S.

“ =

21

valid unless their distribution satisfies the conditions
required in Art. 36A concerning the sale or the private
distribution of independent works.”

ADD: “From 1942 no work published indepen-
dently and offered for sale can be valid unless the
author (or its publisher) either conforms with the
requirements for books not placed on sale, or announces
the work previous to, or simultaneously with, its issue

_ in three botanical periodicals of international circula-
tion to be specified under this rule, and unless he
makes not less than 100 copies available to the botanical
public.

“From 1942 new nomenclatural entities and new
taxonomic descriptions will not be valid unless
published in monographs or botanical periodicals and
appropriately indicated as new.

[There are cases where new species have been published
in school text-books, seed-lists, and _ political periodicals.
Elisabethia miniata Trev. nov. gen and nov spec. was for
instance published in a political daily, Gazzeta Ufficiale di
Venezia, n. 53 (1885, 6th March) p. 3. Hryngium Grosii
F. Q. spec nov. was published in Index Seminum que Hortus
Botanicus Barcinonensis mutua commutatione offert, 1938, p. 12.
It is desirable to prohibit the publication of new nomenclatural
entities in such books or periodicals].

Art. 837A: REVISE as follows:

A name of a taxonomic group is not valid unless it
is (1) correctly formed, (Sect. 4), (2) published in
valid literature (Arts. 20A & 36A), and (3) accom-
panied by (a) a valid description of the group (Notes
1-3), or (b) by a reference to a previous valid descrip-
tion of it published under either a valid or invalid but
different name (Note 4), (see also Arts. 40A-B and
and 414).

Note 1. Names in LINNAEUS’S Species Plantarum
eds. 1 & 2, and in other books adopted as the starting
points of valid botanical literature for different groups
of plants cannot be rejected as invalid merely on the
grounds of absence, or invalidity, of either description
or reference. (see Art. 20A).

Note 2. The citation of the type locality or the
peculiar habitat of a species is not sufficient to establish
a name under this rule. If specific characters are
given in addition to the type locality or the habitat, the
type locality or the habitat becomes a part of the des-
cription and is to be considered as an important element
in determining the identity of species. However
economic uses and vernacular names do not become a

Vol. XI. (1939). .

22

part of the description and so they cannot be used in
determining the identity of species.

[Particulars concerning the locality and the nature of a
habitat can, like the morphological characters of plants, be
obtained or observed by an ordinary plant-collector in the
field without any fear of a serious mistake, and can also be
verified by subsequent collectors; but the particulars concerning
the vernacular names and economic uses cannot be gathered
in the field and the correctness of the information depends
either on the collector’s special ability, or that of his guides, to
recognise the plants in the field and to associate them with
economic uses].

Note 3. In this Article by the word description is
meant a botanical description published in valid litera-
ture printed in any European language written in
Roman characters if before 1935, or in Latin only if
published after 1934, the two exceptions to this being
the following :—

(A) In the case of bacteria and fossil plants, a
description in any above-mentioned European languages
is admitted. But from January Ist, 1912, no descrip-
tion of fossil plants can be valid unless it is accompanied
(a) by illustrations or figures showing essential
characters, or (b) by a reference to such illustrations or
figures published previously in valid literature; if the
required illustrations or figures are published after the
description, then the later date alone must be taken into
consideration for the purposes of the validity of the
description or of its simultaneously published name.

(B) In the case of other (recent) plants, a plate
or figure with analyses showing essential characters can
take the place of a description provided it was published
before 1908.

Note 4: (a) The reference must be made in the
form of a conventional formula or expression intelligible
to systematists in general, a full reference to the text
and page being necessary when the citation is to a valid
description published under an invalid name.

(b) In the case of a generic name, the reference
must be to a previous valid description either published
under another generic name (valid or invalid),
or under any name (valid or invalid) of a generic sub-
division of any category higher than a species.

(c) In the case of a specific (or subspecific) name,
the reference must be either to a valid description
published of the group under a subspecific (or specific)
or another specific (or subspecific) name, or to a valid
description of a monotypic genus provided the descrip-
tion (in the last case) is new and the name to be

Gardens Bulletin, S.S.

23

validated is of the type species (descriptio generico-
specifica). It is permitted to give the description of a
new monotypic genus under the name of the type
species.

ADD examples from Arts. 42, 48 & 44.

[At present the regulation which should form a part of
Art. 37 is distributed to Arts. 38, 39, 42, 44 and 45].

Arts. 38 & 39: DELETE. Incorporated in Art. 374A.

Art. 40A: ADD: But it is permissible to validate alter-
native names (simultaneous isonyms) provided the
alternative name is a combinatio nova equal in rank to
its simultaneous basinym.

[The alternative names were declared valid at the
Amsterdam Congress. But no provision was made against
publishing alternative names of unequal rank or that are not
new combinations, e.g. Minima gen nov. vel Mesembryanthemum
sect. Minima sect. nov.; Cymbopogon viparium. spec. nov. vel
C. fluminense var. riparium var. nov. vel C. riveriense spec.
nov].

Art. 40B: A name proposed provisionally (nomen proviso-
rium) to be adopted by future investigators in case
certain possible circumscription, position or rank be
accepted, or merely mentioned incidentally, is not valid.

[This gives a more precise definition to the nomen
provisorium than the one adopted at the Amsterdam Congress].

Art. 41A: ADD: An exception is made for the generic
names published in LINNAEvs’s Species Plantarum ed.
1 (1753) and ed. 2 (1762-1763) and in other books
which have been adopted as the starting points of valid
botanical literature for the different kinds of plants,
so that the taxonomic groups in these books are treated
as having been validly characterized. (see Art. 37A:
Note 1). ;

Art. 42-44: DELETE. Incorporated in Arts. 37A & 414A.

Art. 45A: REVISE as follows:

“The date of a name or a combination is that of
its valid publication (see Arts. 194 & 37A). In the
absence of proof to the contrary, the date given in the
work containing the name must be regarded as correct.”

[DELETE the other parts of the rule as they are irrelevant
here. cf. Arts. 837A & 61A-B].

Art. 50B: If the name of an order, suborder, family,
subfamily, tribe or subtribe is taken from the name of
a genus, the latter must be taken as the nomenclatural]
type of the former.
Examples as in Art. 66.
[This is Art. 66 revised. It forms a part of this Section].

Vol. XI. (1939).

24

Section 9A: REVISE the wording: Naming and inter-

Art.

Art.

preting taxonomic groups on transference to another
position of the same rank.

[This Section should also deal with the names or epithets

denoting .the subdivisions of order or families].
538A: REVISE as follows:

When a taxonomic group is transferred to another
nomenclatural position without change of rank, the
epithet denoting the group must be retained or, if it
has not been retained, must be re-established unless one
of the following obstacles exists: (1) that there is
available an epithet having a better claim under the
priority rule (Art. 56A); (2) that the original name
‘was not priorable (Art. 61B); (3) that the resulting
combination is an impriorable homonym (Art. 61A-B).

If any of these obstacles occur, then the name or
epithet must be legitimized which satisfies the rule of
priority (Art. 56A).

[This combines the rules in Arts. 53, 54-55 partly, and

69 partly, from which the examples referring to this should
be transferred here. The rule of tautonyms has been ignored
(See Introductory Remarks, Sect. 10). In dividing this rule
into many parts or Articles, not only no advantage is gained
but many contradictory procedures are also made possible.
Thus when in 1804 PorrReT transferred Pinus taxifolia LAMB.
(1808), non Pinus taxifolia SALISB. (1796), to Abies, Art. 54
obliged POIRET (retroactively) to make the combination A.
taxifolia, and, if he had not made that combination, a subse-
quent author would have been justified in making the combina-
tion; but under Art. 69 POIRET was not obliged to make the
combination A. taxifolia, nor would a subsequent botanist have
been justified in making it if PorRET had adopted another
combination for the species].

Note.—The oldest valid name on which the new
combination is based is the basinym of the combination,
and the new combination is the isonym of the basinym.
All isonyms must be typified on the types of their
respective basinyms. (see Art. 53B).

53B: When the epithet, on transference to another
position, has been applied erroneously in its new
position to a different type, the new combination must
be retained for the type on which the epithet was
originally based.

[The example of Pinus Mertensiana discussed at the
Amsterdam Congress, 1935, should form a part of this Article].

Arts. 54 &55: DELETE. Incorporated in Arts. 53A & 53B.

Art.

56A: REVISE as follows:

When two or more groups of the same rank are
united, the oldest priorable name, or epithet, (Art. 61B)
must be retained. When the epithet is not in the
required position, it must be instated in that position

Gardens Bulletin, S.S.

25

provided it does not produce an impriorable homonym;
but if, when so placed, it would produce an impriorable
homonym, then the next oldest priorable name or
epithet that would not produce an impriorable homonym
in the required position must be legitimized. If no
such name, or epithet is available, then the author is
at liberty to legitimize any epithet that becomes
priorable in the new position, even an epithet from an
invalid, or impriorable, name. For the nomenclature
of Fungi with a pleomorphic life-cycle, see Art. 57.

If two or more priorable names, or epithets, have
equal claim under this rule, precedence must be given
to the name, or epithet, that is earlier in the correct
position, or to a new combination over a new name
(see also Art. 614A). In other cases the author who
unites the groups has the right of choosing one of
them as the legitimate name for the group; the author
who first exercises the right under this, definitely
treating one name as a synonym or a subordinate group
of the other, must be followed, so long as the priority
of the names is considered under the same position,
but not otherwise.

ADD the following four examples:

(a) Monospora grandifolia HOCHST. and M. rotun-
difolia HOCHST. were published simultaneously in the
same book (Flora, XXIV, 1841, p. 661). The isonyms
to these two species under Trimeria Harv. (1838) are
T. grandifolia (HOCHST.) WARB. (1893) and T. rotundi-
folia (HocuHstT.) GILG (1921) respectively. When
these two species are treated as synonymous, 7. grandi-
folia (HOCHST.) WARB. (1893) is the correct name for
the united group under Trimeria.

[Mr. E. MILNE-REDHEAD discussed the above case in Kew
Bull. 1939, pp. 34-85. He appears, however, to have over-
looked the fact that WARBURG, in uniting Monospora to
Trimeria, by implication regarded M. rotundifolia as an
unusable synonym. It may therefore fairly be argued that
he was the first to select the name M. grandifolia for the
species. There is also another decision earlier than GILG’s,
by DURAND and SCHINZ (Conspectus FI]. Afr. I, 2, p. 225).

An enquiry into the treatment of the specific epithets
grandifolia and rotundifolia under the genus Monospora itself
might lead to still further complications. The amendment
above proposed, therefore, would furnish botanists with an
easy means by which precise decisions could be given in such
complicated cases, both without undertaking a lengthy investi-
gation into the history of all rival names or epithets of equal
age, and without violating the principle of priority].

(b) When Rhizophora conjugata L. (1753) and
R. gymnorrhiza L. (1753) are united together under

Bruguiera, precedence must be given to B. gymnorrhiza

Vol. XI. (1939).

26

(L.) LAM. (1797) over B. conjungata (L.) MERR.
(1914) and any decisions given concerning the pre-
cedence of these two epithets under Rhizophora must
be ignored.

(c) The combination Talinum polyandrum Hook.
(in Bot. Mag., 1855, t. 4833), being a later homonym of

T. polyandrum Ruiz. et PAv. (1798), is impriorable:
when BENTHAM transferred T. polyandrum Hook. to
Calandrinia, he called it C. polyandra (FI. Austral. I,
1863 p. 172). Now C. polyandra (HookK.) BENTH.
itself is not a later homonym, and so it is a priorable
name. Since there is no earlier priorable name to the
species, C. polyandra (HooK.) BENTH. becomes also
its legitimate name.

(b) Nicotiana? minima PHIL. (1864) is impriorable
because it is a later homonym of N. minima MOLINA
(1782). But the name Petunia minima [PHIL.] REICHE
(1910) is priorable although it is based on the
impriorable N. minima PHIL. Hence Combera minima
SANDW. (Hook. Ic. Pl., 1939, t. 3400) must be not only
cited as C. minima (REICHE) SANDW. (or as C. minima
({PHIL.] REICHE) SANDW.), but aiso typified on the
holotype of N. minima PHIL. in the Museo Nacional
in Santiago, Chile.

[Mr. N. Y. SANDWITH has argued (cf. Arts. 54 and 69),
that he was not bound to adopt the epithet minima when he
transferred N. minima PHIL.— Petunia minima [PHIL.] REICHE
to the genus Combera, that C. minima SANDW. should be called
a nomen novum and not combinatio nova, and that a specimen

in the Kew Herbarium should be taken as the holotype of
C. minima SANDW.].

Section 12A: REVISE: Priorability of Names.
Art. 59: DELETE: Incorporated in Art. 17B.
Art. 60: DELETE:

[The important part of this is incorporated in Arts.
2B, 19A & 19B, 386A, 837A & 61A & B. The rest is unnecessary
and equivocal. In a code of rules one cannot enumerate the
various ways in which an offence may be committed against
the code, without adding to its bulk or creating ambiguity;
but one can classify offences, according to their nature (cf.
Arts. 19A, 19B). Besides by putting invalid names on the
same footing as those valid names which have been published
in violation of the priority rule much confusion has been
created in the Rules].

Art. 614: When there are two or more simultaneous

homonyms (that is, when the same name, or its formal
equivalent, is validly published simultaneously for more
than one taxonomic group of equal rank), the first
reviser who adopts one of them, or substitutes another
name for one of them, must be followed, provided that

Gardens Bulletin, S.S.

27

a new priorable combination having an older, or the
oldest, priorable basinym of equal rank is given
precedence. (see Art. 61B).

‘Note 1—A homonym is a valid name identical

- with another valid name, or an orthographic variant

Art.

Vol.

which is regarded as identical. A homonym that has
been validated earlier is an earlier homonym; while
one validated later is a later homonym. When the
same name or its orthographic variant has been
simultaneously applied to more than one taxonomic
group, each valid publication of the name constitutes a
simultaneous homonym.

Note 2.—In the case of epithets denoting sub-
divisions of a family, a genus, or of a species, formal
homonymy occurs if two identical combinations are
obtained by connecting the epithets concerned directly
with the names of their respective major groups (7.e.
family, genus, or species).

61B: Later homonyms, and such simultaneous ho-
monyms as have been correctly discarded under Art.
61A are impriorable, typonymous formal homonyms
representing subdivisions of unequal rank of a family,
genus or a species being excepted (Art. 28B):
typonymous formal homonyms of unequal rank and
other valid names that are not covered by the first
part of this rule are priorable.

In special cases, the Congress is empowered to
make names impriorable under certain circumstances
(see Arts. 21A—D, 624A, & 634A).

Note 1.—When a real (not formal) homonym is
typonymous with an earlier homonym, it is permissible
to consider the former as either distinct from, or
identical with, the latter (see also Arts. 46 & 47) ; but
neither treatment makes the later homonym priorable.

Note 2.—A name is called priorable when it must
be included in priority considerations; or impriorable,
when it must not be included in priority considerations.
Impriorable names and combinations cannot be legi-
timized, but they can be used as basinyms for making
new priorable isonyms. (cf. Art. 53A Note 1 and 564A).

[The reason why later homonyms and such simultaneous
homonyms as are rejectable under Art. 61A should be made
impriorable is that their use “may cause error or ambiguity”
(Art. 4). Since it is permissible to simplify more complicated
names of the subdivisions of a family, genus or species (see
also Arts. 26A & 28A-B),-the use of the same epithet for

ATI. (1939).

28

two or more subdivisions of different rank having different
types would, when simplified, produce formal homonymy and
cause error or ambiguity: Hence the necessity of making
impriorable all non-typonymous formal homonyms. But no
ambiguity or error would result in the case of typonymous
formal homonyms, for they would all represent a taxonomic
group having the same type: hence the priorability of
typonymous formal homonyms. See also Introductory Remarks
Sect. 4 & 6].

ADD: Examples from Arts. 30 & 61.

Art. 62A: REVISE as follows:

Art.

A name of a taxonomic group whose type is not
extant, or if extant is inadequate for any decisive
interpretation, must not be used in priority considera-
tions or legitimized if, owing to its use with different
meanings, it has become a permanent source of con-
fusion or error. A list of names to be abandoned for
this reason (Nomina ambigua) will form Appendix IV.

[In view of the wording of the 1905-1910 Rules and the
opinions expressed by competent nomenclaturists at the
Amsterdam Botanical Congress (1935) it appears that this
rule was originally intended by the legislators to cover only
those names which, owing to the absence or the imperfection
of types, did not admit of a definite interpretation. The
example in the old Rules reads thus:

“Linné a décrit sous le nom de Rosa villosa une plante
qui a été rapportée a plusieurs espéces différentes
et dont l’interprétation certaine parait impossible;
pour éviter la confusion qui résulte de l’emploi du
nom Rosa villosa, il est préférable dans ce cas,
comme dans d’autres analogues, d’abandonner com-
pletement ce nom.”

In the 1935 Rules, the discussion of this example has been
abbreviated. This abbreviation coupled with the lack of a
definition of the word “permanent” as used in the rule itself,
has been responsible for interpretations which were not foreseen
when Art. 62 was revised in 1930. The new interpretations
contradict the type principle, which obliges botanists to dis-
regard all interpretations of a taxonomic group that do not
include the type. See also the Introductory Remarks, Sect. 14]-

624: DELETE the example concerning Alsine L.

[In view of the official typification of the genus on A. media

L.—the correct lectotype for the genus (see also SPRAGUE in
Kew Bull. 1920 p. 308)—the misinterpretations cannot be said
to be permanent. Under Art. 56A, Alsine L. (1753) becomes
a synonym of Stellaria L. (1753), ViLuarRs (Hist. Pl. Dauph.,
1789) being the first person to choose between these two
synonymous names of equal age].

ADD: Note 1.—A nomen confusum is a special
instance of a nomen ambiguum whose description, being
based on a type composed of two or more discordant
elements pertaining to different species, genera or
orders, was supposed to be of the same species or even
of the same individual, and is moreover incapable of

Gardens Bulletin, S.S.

Art.

Art.

Art.
Art.
Art.

Vol.

29

a certain typification. A list of names to be abandoned
for this reason will form Appendix V. (see also Art.
21B-6).

Examples as in Art. 64.

Note 2.—A nomen monstrositatis is another special
case of a nomen ambiguum whose description, being
based on a monstrosity, cannot give any certain clue
to the identity of the taxonomic group to which the
monstrous type specimen belongs. A list of names to
be abandoned for this reason will form Appendix V bis.

Examples as in Art. 65.

[It is difficult to give a definition of a monstrosity that
would be acceptable to all systematists. Opinions would differ
whether a specific name based on a specimen which had
produced either hexamerous flowers instead of the normally
pentamerous ones, or entire leaves instead of the usually
divided ones, should be included under the category of nomina
monstrositatium. It is therefore desirable to enumerate in a
list all the names to be rejected under this part of the
provision].

Note 3.—If a nomen confusum or nomen monstro-
sitatis has, subsequently to its publication, been typified
on any one part of the original specimens and the error
excluded, and moreover, if the new typification is
accompanied by an amended description and by a
citation of one or more new specimens agreeing with
the new typification, then the name becomes priorable
from the latter date.

[N.B.—It may be useful “to deny validity to nomina
ambigua, confusa, vel monstrositatium. No useful purpose is
served by allowing the names validity and, at the same time,
by denying priorability. On the contrary, by denying the
names a valid status, changes would be avoided in such
homonyms as were published in accordance with the old Rules

which regarded all kinds of nomina ambigua as _ invalid.
(Cf. Art. 21B-b) ].

638A: REVISE as follows:

A name of uncertain application (nomen dubium)
must neither be included in priority considerations nor
be legitimized until its application has been made quite
precise on botanical grounds. (see also Art. 22B).

[The word reject has been used in different senses in

Arts. 61-62, 64, 65, etc., so that to reject sometimes means
also “to render impriorable”. (cf. Art. 62, and also the current
phrase: nominua specifica rejicienda). Hence the necessity of
the paraphrase to make the rule quite unambiguous].

64 & 65: DELETE: Incorporated in Art. 624
Notes 1 & 2.

66: DELETE: Incorporated in Art. 50s.
67: DELETE: Incorporated in Art. 25a.
68: DELETE: Incorporated in Art. 274.

Xf. (1939).

30

Art. 69: DELETE: It is a part of Arts. 534-B & 56<.

Art. 70B: Epithets of species and of lower groups must
be written with a small initial.

[The Recommendation for capitalizing specific epithets in

certain cases causes a good deal of trouble and is based on

the false assumption that pre-Linnean generic names have a

status under the Rules. In many parts of the world, foresters

and economic botanists decapitalize specific epithets; so do also

the majority of American systematists and all zoologists. No

useful purpose is served by maintaining this Recommendation].

Art. 70c: Names or epithets of taxonomic groups higher

than species are written with an initial capital letter.
[Parts of Arts. 25 and 26].

tec. XLII: DELETE: Annulled by Art. 7OB.

Section 14: READ: “Gender of names” instead of “‘gender
of generic names’’.

Art. 72B: Enpithets denoting either species or the sub-
divisions of a genus or species, when adjectival in
form and not used as substantives agree in gender
with the generic name.

[Parts of Arts. 26, 27 and 28].

Gardens Bulletin, S.S.

, y ;

Megs

iy
i.

i
Mad

oa GARDENS’ BULLETIN

"STRAITS. SETTLEMENTS

PAGE

= “Palmae: Ui Mateeose tie The Genus Licuala' in the
it Malay Peninsula; by C. X. Furtado .. - 3l

“The Fern Genus Diplazium in the Dieta Peninsula, bp
wR. EB. Holtturm - - 74

Carseh Grass, Axonopus Spp., by R B. Jagoe - 109

Utyners Leaf-Change and. Flowering of Trees in x net
het by R EB. Holttum: - 119

soe ;
*

i “To be purchased a at the Botanic Gardens, Singapore
ee ion nes . Price, ee 50 or 8s, 2d.

Cs tes Published by Authority

SINGAPORE:

“Penn ar tHe Government PRintinc QOrrice, SINGAPORE,
Ww. : “CHERRY, GOVERNMENT PRINTER:

1940

wi
¥ Le
i}
A
PY

————— —— ~

ES GET TOS SER SSE FL RC BO ee OH

"

THE

GARDENS’ BULLETIN

ssalick Stabe ae

Vol. XI 29th November, 1940 U Bact 2

PALMA MALESIC

VIWI—THE GENUS LICUALA IN THE
MALAY PENINSULA

By C. X. FURTADO
Botanic Gardens, Singapore

Page
1. INTRODUCTION ba a ai. if ol
2. HINTS TO COLLECTORS ue £ 1a 3
3. SUBGENERA AND SECTIONS .. ite fag 34
4. ANALYTICAL KEY TO THE SPECIES oe wy 38
5. SYSTEMATIC NOTES 4 an a 44
6. SUMMARY AND chbecosawe Ye Ke 70
7. INDEX TO COLLECTORS’ NUMBERS at i} iva
8. INDEX TO NAMES KE ny sy 73

1. Introduction

The results embodied in this paper are of a study
undertaken with the view of arranging systematically the
Singapore herbarium specimens of the Licuala species
indigenous in the Malay Peninsula. The present world
conditions, entailing the cessation of intercourse with certain
botanical institutions, the dislocation of the staff at some
others, and the risks to parcels during transit across oceans
have precluded my obtaining either critical notes of the
type specimens preserved outside the Malay Peninsula, or
a loan of the specimens themselves. Under such limitations
the results of my undertaking would have been of little
systematic value but for the valuable assistance derived
from BECCARI’S monographic work on the genus published
first in an abbreviated form in Webbia V, 1921, pp. 22-55,.
as a part of Recensione delle Palme del Vecchio Mondo
Appartenenti alla Tribu delle Corypheae and then in an
extended form in the Annals of the Royal Botanic Gardens,
Calcutta, XIII, printed in 1931 and published in 1933 (cited

1

(ow)

32

below as Calcutta Annals or Calc. Annals XIII, 1933). In
this latter work BECCARI incorporated the results of the
opportunities he had not only of consulting the type material
of most of the species he studied and of comparing newer
material with it, but also of making lengthy critical descrip-
tions, analytical drawings and photographic plates of such
specimens.

2. Hints to Collectors

Incuala is a genus of small palms the largest of which
attain about 15-20 feet in height and a few (+ 3) inches
in stem diameter. However, most of the species are almost
stemless or only a few feet in height. Compared therefore
with the long, thorny rattans or tall giants like Corypha,
Borassus and Oncosperma, the Licuala species present very
few obstacles to the collector making good herbarium
specimens. The tallest ones can be cut down with an
ordinary jungle knife within a few minutes, and the
specimens of leaves and of spadices with flowers and fruits
do not form very weighty or bulky specimens. Though the
petioles of most species are provided with thorns, yet the
specimens are easily handled even without leather gloves,
which are required in handling specimens of Rattans or
Oncosperma. Hence one would have expected this genus
to be well represented in most herbaria, and also to be
specifically well known in areas botanically fairly well
explored. From the study of the species in the Singapore
herbarium and from the notes given by BECCARI, I conclude
that this is far from the case. This country has been
explored botanically for the last fifty years or more by
different botanical collectors and I find the Licuala species
are very badly represented in the herbarium. Even in
regions botanically well known, species formerly not
recorded are being discovered. The principal reason for
this is that in the field most of the species look alike and
are not easily distinguished one from another except when
the collector has made a special point to study them in
the field. The number of segments present on a leaf often
depends on the age of the plant and to a certain extent on
the conditions under which it grows. Many times really
acaulescent species may be mistaken for others which flower
and fruit when quite smal] or stemless. In many species,
moreover, the flowers and fruits remain hidden among the
leaves so that they may be passed as the sterile stages of
others.

If the collector is therefore to see that he does not
miss any uncommon species he meets with in the jungle,
he must learn to distinguish the species in the field. To

Gardens Bulletin, S.S.

Do

do this satisfactorily the following pointers may be
useful :—

1. The average number of main nerves (costae)
radiating from the apex of the petiole is a
more stable character than the average
number of leaflets into which a leaf is divided,
though the latter is sometimes useful in the
identification of a species.

2. Some species have always undivided leaves (not
noticed in any species wild in the Peninsula).

3. In some species the median segment of the leaf
is three or more times as broad as the other
lobes, in others the segment is nearly as broad
as the other segments.

4. In some species the broad median segment is
entire, in others it is divided half-way.

5. In some species the middle segment becomes
distinctly petiolulate, in others it is always
sessile.

6. In some species the lateral margins of the leaf
segments are arcuately cuneate, in others the
margins are straight.

fe aa species are always solitary, others form
tufts.

8. Some are always stemless, others produce stems
though they may begin to flower when
stemless.

9. In some species the spadices reach high above
the height of the leaves so as to become visible
from a distance; in others the spadices remain
hidden among the leaves.

10. In some species the spadices are simple, ter-
minating with one or more (floriferous
branchlets (spikelets) ; in others the spadices

| are compound.

11. In some species the lower branches of the com-
pound spadices are subdivided into 10 or
more spreading spikelets; in some others the
branches contain only 2-5, usually digitate,
branchlets; in still others the branches are
simple, undivided.

12. Different species are characterised by different
colour of spadices, flowers and fruits.

13. In some the fruits are 3—4 times as long as they
are thick; in others the fruits are globose.

In a given field, all Licualas may have many of the
above-mentioned characters in common, but they will differ
in others, which, though a few, will suffice to distinguish

Vol. XI. (1940).

34

between the species, even disregarding such variable
characters as the height of the plant, and the nervation and
the number of the leaf-segments. But it is desirable to
record in the field notes as many particulars as possible
concerning the habit, stem height (without leaves), etc.,
of the plants, size of the petioles, leaf-lamina and spadix,
the average number of segments to a leaf, the average
number of main branches to a spadix, the colour of fruits,
flowers, etc.

For herbarium purposes the leaf specimens should be
taken from flower-bearing crowns; and if there is some
variation in size, etc., in the leaves of non-flowering crowns,
some few specimens may be made to indicate this variation.
Care, however, should be taken to use appropriate tags so
that the leaves may be sorted and mounted in the herbarium
according to their status without any fear of confusion (vide
my remarks on numbering the specimens in the field in
Gard. Bull., Straits Settl. TX, 19387 p. 155, 286-287, and
308 in example).

If it is desired to reduce the bulk of the specimens
one may split the leaves longitudinally, taking care to leave
the middle segment intact; the half without median segment
becomes practically useless, unless the species is rare and
a sufficient number of leaves is not available for making
up the sets.

3. Subgenera and Sections

Apart from the collector’s difficulties, the genus Licuala
presents many difficulties to a systematist. There is a good
deal of variation in the vegetative parts of the plants.
Though the branching of the spadices frequently occurs
in a definite manner and forms a useful character to
differentiate between species, it is not one that can be
employed to subdivide the genus into subgenera or sections;
such a classification not only places very closely related
species in different sections or subgenera, but also divides
subgenerically or sectionally polymorphic specific units.

BECCARI, who in 1886 had relied on the branching
of spadices to divide the genus Licuala subgenerically into
Eu-LICUALA (implicit), LICUALOPSIS, LICUACELLA and
LICUALINA (Malesia III, 1886, pp. 69-90), was obliged in
1921 to reduce these subdivisions to two, namely, Ev-
LICUALA and LICUACELLA, the latter to include LICUALOPSIS
and LICUALINA (Webbia. V, 1921, pp. 22-55). Here BECCARI
made also a new subgenus DAMMERA, but the main character
used to define this subgenus is derived from the flower.
This disposition has also been adopted in BECCARI’S
monograph published in the Calcutta Annals, XIII (1933).
Leaving the subgenus DAMMERA out of consideration for

Gardens Bulletin, S.S.

30

the present, this modified subgeneric division of Licuala
is still unsatisfactory, because the basis used to distinguish
these two subgenera is still the character of the branching
of the spadices. Under such a classification, for instance,
L. lanuginosa and L. Kingiana fall into two subgenera, when
the species are so very closely related that they may be
regarded as two varieties of the same species. Further,
this classification obliges one to split such variable species
as L. modesta Becc. sensu lato into L. modesta Becc. sensu
stricto and L. Wrayi Becc. according to the development
of the spadix, and then to place these two species thus
defined into two different subgenera.

The other grouping of the Licuala species is one
published by DRUDE (ENGL. u. PRANTL, Pflanzenf. II,
3, 1887, p. 35), subsequently modified by RIDLEY. DRUDE
was apparently unaware of the subgenera published by
BECCARI, for, without making any reference to the latter’s
subgenera, DRUDE proposed independently two subgenera,
namely: EU-LICUALA to include the species with a dorsal
embryo, and PERICYCLA (Bl.) Drude (spelt by error
Pericyla) to include the species with a basilar embryo.
The type of this second subgenus was L. penduliflora (Bl.)
Migq., published previously as Pericycla penduliflora Bl. It is
apparent that, in reducing Pericycla Bl. to a subgenus of
Inicuala, DRUDE relied on the characters mentioned by
BLUME, who had examined the position of the embryo in
the flower only, where the position is often misleading. I
point this out because I have not seen any species of Licuala
having a basilar embryo, and BECCARI, who had opportuni-
ties to examine a very large collection of fruiting specimens,
does not mention any such species.

In 1903 RIDLEY reduced Pericycla Bl. to a section of
Incuala, defining it to include all species having a “‘panicled
inflorescence” (Journ. Roy. Asiat. Soc., Straits Settl., 41,
1903, p. 42). The section was naturally typified on L. pen-
duliflora, of New Guinea, for which RIDLEY adopted the
name L. pericycla Zipp. Mss.; but the only Malayan species
that RIDLEY included in the section was L. paniculata Ridl.,
though L. longipes, L. paludosa, etc. have inflorescences
similarly branched. RIDLEY adopted this section also in his
later works, viz.: the Materials for a Flora of the Malay
Peninsula, Monocotyledons (cited below as Materials or
Mat.), II, 1907, pp. 159-165, and the Flora of the Malay
Peninsula (cited below as Flora), V, 1925, pp. 24-30. But
if, following RIDLEY’s definition, one were to place all the
species having ‘panicled inflorescences” in the section
PERICYCLA, the group would consist of utterly unrelated
species, several closely related species remaining outside.

Vol. XI. (1940).

56

In view of these drawbacks of previous definitions, an
inquiry was made into the problem of subdividing the genus.
The most stable characters were found to lie in the flowers,
the nature of the divisions of the androecium affording
characters to group the species into well defined subgenera.
The utility of these characters was recognized by BECCARI
(op. cit 1921 & 1933); but he employed them only in the
analytical keys provided for the identification of the species,
and not to define his subgenera. On the basis of these
characters, I divide the genus Licuala into three subgenera,
namely: LIBERICULA (new), EU-LICUALA and PERICYCLA.
The second is further divided into four sections, but
LICUACELLA is not retained.

I. LIBERICULA. Furtado subgen. nov.

Staminum filamenta 6 erecta aequalia, basi lata superne
subulata, ad basin libera vel fere, nec in annulum pro-
minentem ad corolle faucem connata. Flores inter majores,
circa 15 mm. longi, 5 mm. crassi.

Species unica, L. peltata Roxb., in regionibus humidis
inter septentriones et orientem Indiarum spectantibus ex
Assam, Khasia, Sikkim, etc. ad meridiem in Insulis
srt Tenasserim, et Thailandia (Siam) meridionali
habitat.

II. EU-LICUALA Drude in Engl. u. Prantl, Pflanzenf. II,
3 (1887) 35; Ridl., Flora V (1925) 24 loco sectionis,
nom. nud.

Licualopsis Becc., Malesia III (1886) 85.
Licualina Bece. op cit p. 88.
Licuacella Bece. op cit p. 86, in Webbia V (1921), 24 et
37, in Cale. Annals XIII (1933), 116 et 128. Syn nov.
Staminum filamenta 6, aequalia, erecta vel inflexa, ad
corolle faucem in annulum prominentem connata. Flores
3-8 mm. longi, 2 mm. crassi.
Sectio 1: WURMBIA Furtado sect. nov.

Annulus staminalis membranaceus, aut truncatus,
apice in filamenta parva, erecta, filiformia productus,
aut in filamenta erecta, lobiformia, abrupte subulata
divisus. Anthere haud aristate.

Inter divisiones Licualae hec sectio species
plurimas includens, qui in regionibus torridis et semi-
torridis ex Birmania, Thailandia (Siam) et China
meridionale ad Celebesiam et Novam Guineam habitant,
maxime species in regionibus malayanis (viz. in
Peninsula Malayana, Sumatra et Borneo).

Species typica: L. spinosa Wurmb.

Nomen hujus sectionis in honorem cl. F. VON
WURMB, auctoris speciei typice.

Sectio 2: BONIA Furtado sect. nov.

Gardens Bulletin, SS.

o7

Annulus staminalis brevis, crassus. Filamenta
erectiuscula, lobiformia, elongata, subbulbosa, apicem
in connectivum discoideum antherarum dorso adnatum,
expansa. Anthere haud aristate.

Species hujus sectionis adhuc unica (L. fatua
Bece.) cognita, habitat in Indo-China. Nomen hujus
sectionis in honorem Rev. Presb. BON, qui specimen
typicum speciei in Tonkinia Occidentale legit.

Sectio 3: DAMMERA (Becc.) Furtado stat. nov.

Dammera Bece. in Webbia V (1921), 24 et 38 et
in Cale. Annals XIII (1933) 116 et 130 (loco
subgeneris).

Dammera K. Schum. et Laut, Fl. deutsch
Schutzgeb. Sudsee (1900) 201 t. A-F et 2 (loco
generis).

Annulus staminalis elevatus, crassus. Filamenta
e basi valida superne attenuata et antherarum dorso
prolixe adnata. Anthere haud aristate. Species
hujus sectionis due tantum cognite, ambe in Nova
Guinea habitant.

Species typica: L. Beccariana Furtado nom. nov.
Basinym: L. ramosa (K. Schum. et Laut)
Bece. (1921), nec L. ramosa Bl.
Sectio 4: BECCARIA Furtado sect. nov.

Annulus staminalis brevis, crassus, conspicuus.
Filamenta lobiformia, perlonga, subulata, apice bis
inflexa (i.e. apice introflexa et rursus erecta), antheras
apiculato-aristatos ferentia.

Species hujus sectionis adhuc unica (L. reptans
Becc.) cognita in Borneo habitat.

Nomen hujus sectionis in honorem Cl. O. BECCARI,
palmographi magni.

II. PERICYCLA (Bl) Drude in Engl. u. Prantl, Pflanzenf.

II, 3 (1887) 35 (Sphalmate Pericyla).

Pericycla Bl., Rumphia II (1844) 47 t. 94 (loco

generis).

Pericycla (Bl.) Ridl. in Journ. Roy. Asiat. Soc., Straits

Settl. 41 (1903) 42, Mat. II (1907) 159 et Flora V
(1925) 25— (loco sectionis ; sphalmate
Pericyclus).

Staminum filamenta 6, erecta, valde inequalia, 3 ad
loborum apicum majorum inserta, et 3 in sinubus vel ad
loborum apicum minorum sita, ad corolle faucem in
annulum prominentem connata. Anthere haud aristate.
Flores inter minores.

Species hujus subgeneris pauce, omnes in regionibus
oceanlis habitant; unica in Australia, alterz in insulis

Vol. XI. (1940).

38

papuanesianis usque ad insulas Solomonenses, maxime in
Nova Guinea.
Species typica: L. penduliflora (Bl.) Mig. (=Pericycla
penduliflora Bl.).
4. Analytical Key to the Species
A. Flowers about 15 mm. or more
long and 5 mm. or more
across. Stamens 6, equal,
free at the throat of the
corolla, the staminal ring
being almost obsolete
(LIBERICULA). . LL. peltata Roxb.
AA. Flowers about 3-8 mm. long
and 2 mm. across. Stamens
united at base into a
conspicuous ring at the
throat of the corolla ; (B).
B. Staminal ring 3-lobed, lobes
emarginate; 1 filament in
the notch at the apex of
each lobe and 1 in each
sinus between the _ lobes.

(PERICYCLA).
(a) Leaf-blade entire. |
Partial inflore- f

scences panicu-

lately divided into

8-10 branchlets
(spikelets) .. L. grandis Wendl}. ;
(cultivated).
(aa) Leaf-blade multipar-
tite. Partial in-

florescences digi- ;
tately divided into :
3—4 branchlets .. L. Rumphii BI. i

(cultivated). .

BB. Staminal annulus ends in 6
almost equal lobes’ or
filaments (EU-LICUALA) .. (1).

1A. Spadix simple, terminated by
one or more floriferous
spikelets i (2).
1B. Spadix branched into two or
more partial inflorescences,
each internode sheathed by

a separate spathe (4).
2A. Spadix terminated by one or
two spikelets BANS tiie ps

Gardens Bulletin, S.S.

2B.

oA.

3B.

4A,

6B.

TA.

7B.

39

Spadix terminated by many
floriferous spikelets (some-
times a_ short liguliform
spathe intervening)

Spadix axis and flower-buds
covered with long coarse
hairs. Calyx striate, split-
ting into many longitudinal
fibres. Leaf divided into
5-8 segments

Spadix axis and flower-buds
covered with minute hairs,
Calyx membranous, not
dissolving into fibres. Leaf
segments 12-15

Partial inflorescences un-
branched (1.e. consisting of
solitary spikelets)

Partial inflorescences
branched

Flowers distinctly pedicelled.
Calyx turbinate or nar-
rowed into a long pedicelli-
form base ay Zi

Flowers not distinctly
pedicelled. Calyx cylin-
drical or campanulate

Spikelets distinctly hairy.
Calyx hairy, or, if
deciduously hairy, the
corolla is distinctly hairy

Spikelets and calyx not hairy,
but covered with fugaceous
rusty-brown scales. Corolla
glabrous

Hairs on spikelets and calyx
long, coarse. Calyx nearly
as broad as long, membran-
ous, striate, turbinato-
campanulate, suddenly
ending in a _ short solid
cylindrical base, more or
less lobed at apex. Flowers
usually solitary

Hairs on spikelets and calyx
fine. Calyx coriaceous, not
striate, cup-shaped, nar-

Vol. XI. (1940).

L. modesta Bece.
(partly).

L. Kingiana Bece.

L. Seortechini Becce.

(5).
(14).

(6).

(8).

(7),

L. Moyseyi Furtado

L. Kunstleri Bece.

SA.
SB.
9A.

OB.

10A.

1OB.
11A.

11B.

12A.

12B.

135A.

40)

rowed into a long (longer
than the cup) pedicelli-
form base, obscurely denti-
culate at apex but not lobed
or lacerate. Flowers in
groups of 2—3 in some parts
Calyx cyathiform-campanu-
late, deeply lobed
Calyx cylindrical,
or lobed at apex
Calyx covered with
coarse hairs, usually one
carpel fertile. The lateral
margins of leaflets nearly
straight t
Calyx covered with fine hairs
which later fall off partly.
Frequently two or three
carpels fertile. Lateral
margins of leaflets arched,
so that leaflets are arcuately
cuneate

truncate

long

Hairs tawny
Hairs whitish

Calyx +4 mm. long, densely
tomentose, curvately cune-
ate at base, little lobed at
apex

Calyx smaller, puberulous or
glabrous, not curvately
cuneate at base; truncate or
deeply lobed at apex

Calyx conspicuously striate,
glabrous, asymetrically
lobed at apex, truncate,
caudiculate at base

Calyx puberulous or glabrous,
but neither conspicuously
striate nor caudiculate

Calyx apex almost truncate
at first, then irregularly

L. Corneri Furtado
(9).
(11).

(10).

L. kemamanensis
Furtado.

L. acutifida Mart.
(partly).

L. pusilla Becc.
(partly).

L. tiomanensis
Furtado.

(12).

L. pahangensis

Furtado.

(15).

split
Calyx deeply lobed

L. Ridleyana Becc.

i152. L. confusa Furtado.

Gardens Bulletin, SS.

14A.

14B.

15A.

15B.

16B.

17A.

17B.

18A.

18B.

19A.
19B.

Al

Lower partial inflorescences
paniculiform, composed of
10 or more _— scattered
spikelets (=floriferous
branchlets)

Lower partial inflorescences
composed of 2-5, usually
digitate, spikelets

Spikelets slender, somewhat
flexuose. Flowers _ fre-
quently in groups of 2,
spirally, or in terminal
parts subalternately,
arranged. Calyx 5 mm.
long, about twice as long as
it is broad 3%

Spikelets stouter, not flex-
uose. Flowers pluriseriate,
usually solitary. Calyx
shorter, almost as long as
broad

The spathes much inflated.
Spadix flexuose. Calyx
cylindrical, thick from the
very beginning. Ovary
villous in the upper part ..

Spathes not’ so_ inflated.
Spadix not flexuose. Calyx
campanulate, membranous
at first. Ovary glabrous ..

Calyx glabrous, not split or
lobate, but denticulate at
apex

Calyx otherwise .¥

Leaf-lamina about 12/’—16”
in radius, divided into 12 or
more, triangular, subequal
segments

Leaf-lamina 812" j in radius,
divided into 5-8, arcuately
cuneate, unequal segments,
median segment divided
half-way

Flowers pedicellate
Flowers not pedicellate

Vol. XI. (1940).

(15).

LLY).

L. longicalycata
Furtado.

(16).

L. longipes Griff.

L. paludosa Griff.

(18).
(19).

L. glabra

L. glabra var.

selangorensis.
L. mirabilis.
(20).

Z0A.

20B.

21A.

21B.

22A,

22B.

235A.

23B.

42

Spathes and spadix densely
covered with dark fuscous
ferruginous scurf, and the
spikelets and calyces
densely covered with
similarly coloured hairs .

Scurf and hairs when present
not similarly coloured and
frequently not so thick ..

Calyx cylindric, truncate at
first, slightly split later.
Spadix and spikelets not
covered with a thick coat of
deciduous tomentum

Calyx campanulate, lobed, or,
if apparently truncate in
early phases, the spadix
and spikelets covered with a
thick coat of deciduous
tomentum

Plants stemless, about
12”-18” long, smallest in
the genus. Leaves’. the
smallest in the _ genus,
divided into 3-8, rarely
more, segments. Spadix
about 12”-15” long, with
spikelets 114”-214” long ..

Plants and leaves larger,
with more leaf segments.
Spadix and spikelets usually
longer ay i

Flowers frequently in groups
of 2-3, often more than one
fertile carpel. Fruit nar-
row elongate, about 3—4
times as long as it is thick.
Ovary glabrous. Median
leaf-segment never petiolu-
late Ai :

Flowers solitary, with only
one fertile carpel. Fruit
globose. Ovary hairy.
Median segment frequently
petiolulate uF

—" ao

4
!
4
A
‘
;
;
;

L. ferruginea.

(21).

L. malajana Bece.

(22).

(23).

(24).

L. Kiahu Furtado.

L. triphylla Griff.

Gardens Bulletin, SS.

45

24a. Lower partial inflorescences
usually bifid, the upper ones
unbranched. Most of the
leaflets nearly linear, about
34”-1l4 inches in width (25).

248. Lower partial inflorescences
usually 3—5—branched.
Leaflets in most’ cases
broader and conspicuously

triangular an Gees
25A. Spikelets and calyces rusty-
tomentose. Flowers pluri-
seriate on a_ cylindrical
spikelet as .. LL. acutifida Mart.
(partly).
25B. Spikelets and calyces whitish-
tomentose. Flowers §ar-
ranged spirally on flexuose
spikelets aa .. LL. pusilla Bece.
(partly).

26A. Spadix, spathes and spike-
lets covered in early stages
with a thick coat of cream-
ish deciduous woolly to-
mentum. Corolla densely

sericeous. Flowers solitary JL. lanuginosa Ridi.
26B. 'Tomentum, when present on
spadix, spathes or spikelets,
not thick, and minute.
Flowers frequently in
groups of 2-3, Corolla

glabrous or minutely hairy (27).

27A. Leaflets about 12” long.
Spadix at the most 214 ft.
long. Spikelets usually less
than 6” long and the lower
- spadix branches about
6” or less apart. Fruit
elliptic, about 10-13 mm.

long, 6-8 mm. through .. UL. modesta Becc.

(partly).

278. Leaflets 20” or more long.
Spadix 6 ft. or more long,
with lower branches 9-18
inches apart, each spikelet
in lower branches being
9-15” long. Fruit globose,

6-8 mm. in diameter .. LL. spinosa Wurmb.

Vol. XI. (1940).

44

5. Systematic Notes

A. LIBERICULA Furtado.
Licuala peltata Roxb., Fl. Ind. II (1824) 179; Griff. in
Calc. Journ. Nat. Hist. V- (1845) 324, et Palms
Brit. Ind. (1850), 12012222) Heok.L,, Fil ame
Ind. VI (1892) 480; Bece. in Webbia V (1921)
24 et 39, et in Cale. Annals XIII (1933) 131 t. 74.

MALAY PENINSULA: Lower Siam, Kantang (Haniff &
Nur, 4720).

Distribution: In the monsoon forests of North East

India, Burma, Andamans and Nicobars.

This species was not recorded previously in the
Peninsula, where it occurs only in the northern-most parts
which form the southern-most range of the distribution of
the species. In cultivation it is grown in many gardens
throughout the Peninsula. The species sometimes begins
to flower before its leaves have started to divide into
segments.

B. Eu-LICUALA § Wurmbia Furtado.

1. Licuala acutifida Mart., Hist. Nat. Palm. III, ed. 1
(1842), 237; t.. 135. TH ‘et 1V-s ‘Grit’ in) Cale
Journ. Nat. Hist. V (1845) 237 et Palms Brit.
Ind. (1850) 122 t. 222 A et B; Mart. op. cit. III,
ed. 2 (1849) 236 et (1850) 318; Hook. f., FI.
Brit. Ind. VI (1892) 483; Ridl., Mat. II (1907)
163 pp.; Bece. in Webbia V (1921) 30 et 44;
Ridl., Flora’ V (1925): ‘27 pp.; Beecs in’ Cale:
Annals XIII (1933) 168. ¥

MALAY PENINSULA: Penang, Tulloh Bahang (Curtis,
1010) ; Government Hill (Ridley and Curtis, 7906) ; Water-
fall (Curtis in June 1890); Penang Hill (Ridley in July
1898).

This species is very near to L. Kunstleri and so far not
known from outside Penang. RIDLEY referred here some
specimens of L. pusilla Becc. LOBB 280 cited by BECCARI as
from Singapore may have come from Penang, for LOBE
frequently gave incorrect localities to his plants, and there
are reasons to believe that the majority of the Malayan
plants collected by LOBB were from Penang (vide, BURKILL,
in Gard. Bull. Straits Settl. IV, 1927 p. 127).

CURTIS notes on his specimen numbered 1010 that this
is not the palm that supplies the walking sticks known as
the “Penang Lawyer’’.

2. lLicuala confusa Furtado sp. nov. Fig. 1.

? L. acutifida Becce. var. peninsularis Bece. in Webbia
V (1921) 30 et 44, et in Cale. Annals XIII (1933)
169 t. 10-II. Syn. nov.?

Gardens Bulletin, S.S.

Fig. 1. Licuala confusa (Holotypus: FURTADO no. A).

a. Frons. b. Spadix.

resecta, ut aestivatio staminum appareat.
Giscissum. f. Semen verticaliter discissum.

Vol. XT. (1940).

45

c. Alabastrum. d. Alabastrum, petala
e. Alabastrum verticaliter

46

A L. Ridley ana, cui valde affinis, differt calyce conspicue
lobato, apice nonnihil membranaceo.

Palma _ solitaria, acaulescens. Petiolus ultra 1 m.
longus, interdum minor, ad medium aculeatus. Segmenta
foliorum circa 17, subaequalia vel inaequalia, 2—4 costata,
34-45 cm. longa; medianum indivisum, interdum alteris
multo latius, 6-12 costatum. Spadizx petiolo brevior, in 4—5
partiales inflorescentias simplices, axi tomentosas divisus.
Flores pluriseriatim dispositi, solitarii, plerumque super
pulvinos prominulos interdum in depressionibus inserti.
Calyx puberulus, cylindricus, basi truncatus, apice 3-
lobatus, Iobis rotundatis, saepe in 2 lobulos fissis, parum
striatis, nonnihil membranaceis. Corolla calyce duplo
longior, lobis acutis, puberulis vel non. Annulus staminalis
in 6 filamenta antherifera abrupte subulata divisus.
Ovarvm glabrum. Fructus vivo rubrus, in sicco brunnes-
cens, rugosus, circa 8 mm. in diam.; semen orbiculare 5 mm.
in diam., gramineum, fossa integumentale cylindrica, ad
centrum seminis cursa, apice parum curvata.

MALAY PENINSULA: Perak, Tapah Hill (Furtado sub
htt. A, 11 June 19387—Holotypus) ; Tapah (Ridley, 14112) ;
Bujong Malacca (Curtis in Dec. 1895 et in Aug. 1898).
Dindings, Lumut (Ridley in Feb. 1892). Selangor,
Semenyih (Hume, 7962).

RIDLEY 14112 is the haptoholotype of L. Ridleyana and
in the capsules mounted on the sheet there are a few flowers
which agree with the description and figure given by
BECCARI for that species. But the vast majority of the
flowers in the capsule and the flowers still attached to the
spikelets are like the ones described here. Since I have
typified L. Ridleyana on the characters of the flowers, I
could not retain under that species RIDLEY 14112 from the
Singapore herbarium. There are some minor variations
between RIDLEY 14112 and the holotype of L. confusa. The
Dindings specimens has no flowers and has been cited here
because of its general resemblance to this species.

I have reduced here L. acutifida var peninsularis with
some doubt. RIDLEY 9806 is cited as a paratype of the
variety, but its duplicate in Singapore is L. Ridleyana as
typified here. I have not been able to find in the Singapore
herbarium any material which agrees exactly with the
figures by BECCARI. RIDLEY 10329 from Lumut (the type
collection is cited as 10239) bears calyces somewhat similar
to the ones figured by BECCARI, but the partial inflorescences
are branched in the Singapore specimen and the corolla is
hairy. It seems to represent a variety of L. spinosa, typical
material of which species is also mounted on the same sheet.

Gardens Bulletin, SS.

a

AT

L. confusa has manifest affinities with L. Ridleyana but
this has a truncate calyx which later splits irregularly. L.
nahangensis, also an ally of L. confusa, has a glabrous,
conspicuously striate calyx which bears a somewhat caudicu- ~
-late base and less symmetrical lobes at the apex.

3. Licuala Corneri Furtado sp. nov. Fig. 2.

L. Kunstleri proxima, a qua calycibus basin versus
pedicelliformiter longe productis truncatis, obscure denti-
culatis, sat distincta.

Palma solitaria ut videtur, cum caule 0-60—2 m. longo,
2-4 m. alta. Petiolus 35-100 cm. longus, prope basin
tantum aculeatus, in medio circa 6-8 mm. et apice 4 mm.
latus. Segmenta frondium subaequalia, circa 12-14,
cuneatissima, 2—3, raro 1-4, costata, 25-35 cm. longa, apice
3-45 em. lata; mediana alteris latiora, apice dentibus
brevissimis. Spadices compositi, 35-80 cm. longi, in
dimidia parte basali haud ramosi, cum 2-3 spathis involuti,
in altera parte terminali in ramos simplices, florigeros,
plerumque 3-4, ad 15 cm. usque longos, dense pubescentes,
divisi. Spathae basales bicarinatae, alterae tubulares,
punctis fugaceo fusco-leprosis tectae, apice marcescentes.
Flores geminati vel interdum solitarii, pedicellati, in
alabastro fusiformes, 5-8 mm. longi. Calyx dense pilosus,
4 mm. longus, apice truncatus, obscure 3-denticulatus, in
basin pedicelliformen contractus. Covrolia calyce cupulari
duplo longior, glabra. Annulus staminalis in filamenta 6
apice abrupte contracta, divisus. Fructus immaturus
tantum visus, ellipticus, utrinque acutus, circa 6 mm. longus,
4-5 mm. in diam., fossa albuminali indivisa, cylindrica.

MALAY PENINSULA: Kemaman, Ulu Bendong in
Kajang, alt. 500 ped. (Corner, 30072—Holotypus) ; Sungai
Nipa (Corner, s.n.).

CORNER 30072 was found mixed with some specimens
of L. malajana. The collector notes: palm is slender
solitary; leaflets dark green above, pale beneath, not
glaucous; flowers greenish white; fruit fall orange-red when
ripe; perianth green; staminal tube white; ovary pale
orange; inflorescence hanging”’.

Though this species falls into the group having simple
partial inflorescences, it has no close ally in the Peninsula
except the one described here as L. Moyseyi. In the long
pedicelliform base of the calyx and its obscurely toothed
apex the species may seem to appear very near to L. |
Beccariana of New Guinea but that species has been
described to have very much longer flowers borne on simple
and shorter spadices and belongs to the section DAMMERA.

4. Licuala ferruginea Becc. in Hook. f., Fl. Brit. Ind.
VI (1892) 482; Ridl., Mat. II (1907) 162; Becc.

Vol. XI. (1940).

48

Fig. 2. Licuala Corneri (Holotypus: CORNER 30072).

a. Petioli pars inferior. b. Frondis lamina cum petiolo. ec. Spadix
fructiferus minimus. d. Alabastrum. e. Flos ad anthesin. f. Flos,
corolla resecta, ut aestivatio staminum appareat. g. Fructus cum
perianthio et filamentis staminum. h. Fructus verticaliter discissus,
semen cum substantia integumentale oblique percurrente et enibrae
dorsale exhibens.

Gardens Bulletin,

S.S

49

in Webbia V (1921) 32 et 46; Ridl., Flora V
(1925) 26; Becc. in Cale. Annals XIII (1933)
180 tt. 9-X, 89 et 89 bis.

MALAY PENINSULA: Kemaman Bukit Kajang, alt.
1,000 ft. (Corner, 30398). Pahang, Pulau Tawar (Ridley
in 1891) ; Bukit Sagu (Nur, 25167). Negri Sembilan, Ayer
Kuning near Bahau (Symington, 24380). Johore, Sednah
(Ridley, 18519) ; Sungai Kayu Ara (Corner, 28688) ; Mount
Austen (Ridley in Jan. 1904); Kluang (Holttum, 9260) ;
Gunong Belumut (Holttum, 10605); Sungai Endau
(Holttum, 24944). Singapore, Reservoir Woods (Ridley in
1893) ; Bukit Timah (Ridley, 3512) ; Sungai Jurong (Mat
in 1894); Ang Mo Kio (Ridley, 6676); Bukit Mandai
(Ridley s.n.) ; Bukit Arang (Goodenough on 16 Nov. 1889) ;
Botanic Gardens’ Jungle (Ridley, 3168).

Distribution: Sumatra and Riow Archipelago.

From all the peninsular species having branched partial
inflorescences this species is easily distinguished by the
presence of the ferrugineous tomentum on spikelets and
calyx and by the sessile flowers. It is distinguished from
the Sumatran L. ferruginoides (which frequently produces
branched partial inflorescences) by longer hairs on the calyx
and smaller flowers.

5. Licuala glabra Griff. in Cale. Journ. Nat. Hist. V
(1845) 329 et Palms Brit. Ind. (1850) 124 t.
Zoo> took, f., Fl; Brit.:Ind. VI. (1892) 432;
Ridl., Mat. II (1907) 161; Bece. in Webbia V
(1921) 34 et 48; Ridl., Flora V (1925) 26;
Bece. in Cale. Annals XIII (1933) 192 tt. 14—I
fig. 1-3, et t. 82.
L. longepedunculata Ridi. in Journ. Roy.
Asiat. Soc. Straits Br. XLI (1903) 42, Mat.
II (1907) 161 et Flora V (1925) 26.

MALAY PENINSULA: Peninsular Siam, Kampong Bukit
(Kiah, 24256). Kelantan, Kuala Betis (Henderson, 29725).
Perak, Gunong Batu Puteh (Wray, 254—-Syntype of L.
longepedunculata). Pahang, Karak (Best, 13884) ; Gunong
Tahan (Haniff & Nur, 8111); Fraser Hill (Burkill &
Holttum, 7842) ; Tahan Woods (Ridley, in 1891). Malacca,
Gunong Ledang (Ridley, 3473) ; Mount Ophir (Feilding in
1892; Derry, 633; Hullett, 852; Ridley in Dec. 1899).
Selangor, Bukit Hitam (Kelsall in 1890; Ridley in May
1896) ; Bukit Kutu (Ridley 7894).

In this species there is a good deal of variation in size
and robustness of the leaves, spadices, and flowers, and I
agree with BECCARI in reducing L. longepedunculata to L.
glabra, though the former has somewhat larger flowers than

Vol. XI. (1940).

50

the type of the latter. The specimen from Peninsular Siam
has a somewhat pedicelliform base to the calyx, but
transitions from this to the typical form are also found.

6. Licuala glabra Griff. var selangorensis Becc. in
Webbia V (1921) 35 et 48; Ridl., Flora V (1925)
26; Bece. in Cale. Annals XIII (19383) 194 t.
14—]I.

MALAY PENINSULA: Trengganu, Gunong Padang alt.
circ. 2,500 ft. (Moysey & Kiah, 33399). Pahang, Fraser
Hill, alt. 4,000-4,300 ft. (Burkill & MHolttum, 8426).
Selangor, Semangkok Pass, up to alt. circ. 3,000 ft. (Ridley
in Aug. 1904; 15881, et 12117: Haptoholotype) ; Sempang
(Ridley in April 1911). Johore, Kluang (Holttum, 10601).

The Trengganu specimen cited above is a more robust
form than the others.

7. Licuala kemamanensis Furtado sp. nov. Fig. 3.

A L. ferruginoidea, cui peraffinis, palma acaulescente,
frondibus minoribus, spadicibus haud flexuosis, ramis supra
spathae apicem remote orientibus, floribus minoribus,
bracteolis inconspicuis recedit. Facie L. Kunstleri similis,
sed floribus haud conspicue pedicellati, calyce basin versus
haud valde angustato, ovaris pilosis, foliis minoribus
dissimilis.

Palma humilis, acaulescens. Petiolus circa 30—65 cm.
longus in specimina visa, tertia parte basali aculeis remotis
brevibus, 1-2-5 mm. longis armatus. Segmenta foliorum
circa 14, cuneatissima, 2—6 costata, subaequalia; medianum
circa 27 cm. longum, 4 cm. latum, 3-6 costatum, apice
obsolete dentatum costis approximatis; intermedia mediano
aequilata vel fere, paulo breviora, oblique eroso-dentata, |
costis 8-4 nonnihil remotis; basilaria minima, 2—3 costata,
12-14 cm. longa, 2-5-3 cm. lata, apice intermediis conformia.
Spadix unicus tantum visus, haud flexuosus, 60 cm. longus,
in inflorescentias partiales duas simplices, 5-8 cm. longas,
ferrugineo tomentosas, supra spathae apicem remoto
orientes divisus, basi cum spathis 3 involutus. Spathae
fugaceo fusco-furfuraceae, tubulosae, infima bicarinata.
Flores pluseriatim dispositi, solitarii, in alabastro globoso-
ovati, circa 4 mm. longi; super pulvinum prominentem basi
bracteolatum siti. Calyx lato campanulatus, ferrugineo-
pilosus, haud striatus, ad medium in lobos 3 rotundatos
divisus, basi truncatus. Corolla calyce duplo longior,
striata, apice acuta, puberula. Annulus staminalis in
filamenta 6 aequalia abrupte subulata divisus. Ovarium
loculis saepe 3 fertilibus praeditum, minute pubescens.

MALAY PENINSULA: Kemaman, Sungai Nipa (Corner,
30520).

Gardens Bulletin, S.S.

51

Fig. 3. Licuala kemamanensis (Holotypus: CORNER 30520).

a. Spadix. b. Frons. c. Flos tres carpella fertilia et unum
lamentum staminale exhibens. d. Flos cum uno carpello fertile.

Tol. XI. (1940).

52

This species appears to be very near to L. ferruginoides,
a Sumatran species described to have simple partial inflores-
cences but known also to bear sometimes two spikelets to
each spadix branch. L. ferruginoides is however a very tall
species having much larger leaves and flowers, a zigzag axis
to the spadices, the spikelets arising a little below the apex
of the leaves.

The only peninsular species having simple partial
inflorescences to simulate L. kemamanensis is L. Kunstleri,
but the latter is readily distinguished from the former by
the very long pedicels of its flowers, turbinate calyx, and
glabrous ovary.

8. Licuala Kiahii Furtado sp. nov. Fig. 4.

Palma humilis acaulescens. Petiolus 60-70 cm. longus,
praesertim ad basin versus rubescens, prope basin tantum
spinis reduncis, remotis, 1-2 mm. longis armatus. Segmenta
foliorum 5, rarissime 6-7, inaequalia; medianum 10-14
costatum, 20-25 cm. longum, apice 8-10 cm. latum, super-
ficialiter obtuseque dentatum, cuneatissimum, sessile; altera
3—5 costata, mediano minora, oblique eroso-truncata, arcuato
cuneata. Spadix petiolo duplo vel triplo brevior, basi cum
pedunculo circa 20 cm. longo fusco-furfuraceo, spathis 3—4
tubulosis induto, suffultus, apice in inflorescentias partiales
2-8, simplices vel bifurcatas, furfuraceas, 3—4.5 cm. longas
divisus. Flores in glomerulis 1-3, remotis, alterne vel
spiraliter dispositi, sessiles, pulvino inconspicuo. Calyx
furfuraceus, pubescens, in lobos acuminatos profunde
partitus.Corolla calyce parum longior, pubescens, in lobos
altos, acuminatos divisa. Annulus staminalis in sex fila-
menta abrupte subulata divisus. Ovariwm glabrum;
carpellis plerumque 1-2 vel saepius omnia evolutis. Fructus
elongato-clavatus, saepe parum curvatus, circa 22-23 mm.
longus, 35-45 mm. in diametro; semine conforme, osseo;
fossa integumentali embryo dorsali supra basin objecta,
apicem versus mox ascedente cylindracea.

MALAY PENINSULA: Johore, in collibus apud ripas
Sungai Kayu (Kiah, 32137, Holotypus) ; in monte Gunong
Pulai (Haniff, s.n., 27 Dec. 1925).

This species is very peculiar in the genus in that it
has very long fruits borne on a compound spadix. The
only other species known to produce such long fruits is
L. mattanensis Becc. from Borneo, which however is
characterised by many-lobed leaves, a simple, unbranched
spadix having almost pedicellate flowers, and the integu-
mental process in the seed situated on the same side as the
embryo (not opposite the embryo as is the case with L.

Gardens Bulletin, SS.

53

Fig. 4. Licuala Kiahii (Holotypus: KIAH 32127).

Spadix. b. Planta cum spadice fructifero. c. Fructus.
f. Fila-

a.
d. Fructus verticaliter discissus. e. Perianthium fructiferum.

mentum staminale.

WV Ol. XS: (1940).

54

Kiahu). Otherwise L. Kiahii has many characters similar
to those of L. mattanensis, even in the habit of developing
frequently more than one fertile carpel in each ovary.

9. Licuala Kingiana Becc. in Malesia III (1889) 193;
Hook. f., Fl. Brit. Inds» Vi (1892). 4342 Rigs
Mat. II (1907) 165; Becc. in Webbia V (1921) .
38 et 51; Ridl., Flora V (1925) 30; Becc. in Cale.
Annals XITI 9(1933) 213 t. 11—H.

MALAY PENINSULA: Perak, Bukit Kapayang (Ridley
in Feb. 1904) ; Sungai Siput (Haniff & Nur, 6962) ; Gunong
Keledang (Ridley, 9804); Kinta Valley (Ridley s.n.);
Bujong Malacca (Curtis, 3162) ; Keledang Saiong (Syming-
ton, Cf. 25724 & 25612—Herb. Kepong).

The Bukit Kapayang specimen collected by RIDLEY has
one spadix bifid at the apex.

10. Licuala Kunstleri Becc. in Hook. f., Flor. Brit. Ind.
VI. (1892), 4333 Ridl., Mat. 112( 1907) 1622 Geee:
in Webbia V (1921) 30 et 44; Ridl., Flora V
(1925) 27; Bece. in Cale. Annals XIII (1933)

167 t. 94.

MALAY PENINSULA: Perak, Temango (Ridley. 14709).
Pahang, Tanjong Antan (Ridley in 1891); Sungai Lepar
(Burkill & Haniff, 17460). Selangor, Ulu Gombak, alt.
1,500 ft. (Hume, 9437); Bukit Kutu (Casdani in 1899) ;
Kuala Lumpur (Ridley in 1889) ; Gua Batu (Ridley in Dec.
1896). |

Some specimens of this species have been distributed
from Singapore under the name of L. ferruginea.

11. Licuala lanuginosa Rid]. in Journ. Roy. Asiat. Soe.
Straits Br. 44 (1905) 203, Mat. II (1907) 165;
Becc. in Webbia V (1921) 33 et 46; Ridl., Flora
V (1925) 30; Bece. in Cale. Annals XIII (1933)
181 tt. 9-IV et 50.
L. longipes sensu Ridl. op. cit. (1907 and
1925) pro parte.

MALAY PENINSULA: Johore, Gunong Panti (Ridley in
Dec. 1892; Corner 30739 & 307394) ; Gunong Belumut, alt.
2,200 ft. (Holttum, 10792); Gunong Pulai (Mat in 1892;
Ridley 12198 partim; Best 7711; Corner in 1937) ; Bukit
Tinjau Laut alt. 1240 ft. (Corner 37068).

12. Licuala longicalycata Furtado sp. nov. Fig. 5.

L. bacularize affinis ut videtur, sed planta robustiore,
foliorum segmentis latioribus, inflorescentie ramulis part-
ialis pluribus, floribus majoribus, calyce floris longe piloso,
sed fructus interdum laevi, corolla minute pubescente haec
species sat distincta. Inter species malayanas L. glabra
producit formas ~spectu calycale similes L. longicalycate,

| Gardens Bulletin, S.S.

55

Fig. 5. Licuala longicalycata (Holotypus: KiAH 32401).

a. Pars spadicis floridi. b. Alabastrum. c. Alabastrum verti-
caliter discissum ut aestivatio staminum appareat. d. Alabastrum
verticaliter discissum ut ovarium exhibeat.

e. Fructus (e specimine
SYMINGTON 47092).

f. Semen ejusdem fructus verticaliter discissum.
Wal XL \CIGLO).

56

a qua illa imflorescentie ramulis partialis paucioribus,
calyce laevi apice arcuato denticulato (haud truncato et

ef aac eae fisso), pulvinis floriferentibus brevioribus
ffert.

Caulis robustus, ad 3 m. altus. Petiolus longus 1-3 m.
longus, tertia parte basilari aculeatus, apice circa 8 mm.
latus. Segmenta foliorum circa 16, inaequalia, cuneatissima ;
centrale latissimum, pluricostatum, apice obtuse dentatum,
ad 9-16 cm. latum, 40-55 cm. longum; altera angustiora,
altius dentata, 3—4 costata. Spadix circa 70-80 cm. longus,
petiolo brevior, in 4—5 inflorescentias partiales, unasquisque
o-10 ramulosas, fugaceo fusco-furfuraceas ad apicem
spathe vel paulo infra orientes, divisus, axi flexuosus.
Flores 6-8 mm. longi, super pedicellum prominentem 1-2
glomerulati, remoti, spiraliter dispositi; in alabastro fusi-
formes, striati, dense fugaceo-furfuracei. Spathe tubulares,
fugaceo furfuracee, parum inflate, apice parce marces-
centes, basilares conspicue bicarinate. Calyx 4—5.5 mm.
longus, gramineus, pilis longiusculis remotis praeditus,
cyathiformis, basin versus nonnihil angustatus, utrinque
truncatus, apice superficialiter 3 lobatus, interdum in fructu
irregulariter fissus. Corolla minute puberula, calyce sesqui
vel duplo longior, segmentis lanceolatis, acutis, striatis,
persistentibus. Annulus staminalis 6—fidus, lobis abrupte
subulatis. Ovarium glabrum, gramineum, rarissime loculis
fertilibus 2-3. Fructus in sicco oblongus, rugosus, circa
12-14 mm. longus, 10 mm. in diam; semen oblongum, 7 mm.
longum, 5 mm. crassum; fossa albuminale indivisa,
cylindrica.

MALAY PENINSULA: Johore, Sungai Kayu, in locis
paludosis (Kiah, 32401, Holotypus); Kluang (Holttum,
9252); Arong (Symington, 47092—Herbarium Kepong) ;
Labis (Symington, 47068—-Herbarium Kepong). Negri
Sembilan, Ayer Kuning, prope Bahau (Symington, 24379).

In BEccaARI’s Key L. longicalycata would have to be
placed near L. paludosa and L. bintulensis, from both of
which it is readily distinguished by the longer calyx and the
longer flower-bearing pulvini. The species also appears to
be very near to L. bacularia which I know only from
description, and which is a much smaller species having
narrower leaf-segments, and more branches to the partial
inflorescences. The only peninsular species which may be
confused with L. longicalycata is L. glabra, especially those
forms which have a longer calyx, but in the latter species
the calyx is glabrous, arcuately denticulate and unlobed or
unsplit, even when in fruit. The form, the colour and the
pubescence of the calyx remind one of L. modesta and L.
tiomanensis, both of which bear undivided partial inflores-
cences to the spadix. I have described the young flowers

Gardens Bulletin, S.S.

57

from SYMINGTON 47068 where there is a portion of very
young inflorescence mounted together with an older one;
the veins and the thick furfur that is present on the flowers
in bud disappear when the flowers are developed. The
fruits are described from SYMINGTON 47092.

13. Licuala longipes Griff. in Calc. Journ. Nat. Hist.
V (1845) 330 et Palms Brit. Ind. (1850) 125 tt.
204 & & Bs. Hook..f., Fl. Brit. Ind. VI (1892)
431; Ridl., Mat. II (1907) 162; Becc. in Webbia
W092 8) 26 et 42; Ridl., Flora: V (1925) 27;
Bece. in Cale. Annals XIII (19338) 155 t. 8-V.

MALAY PENINSULA: Kemaman, Ulu Bendong, Kajang,
alt. 700 ft. (Corner, 30111). Pahang, Titi Bungor in
Temerloh (Henderson, 10559) ; Mentakab (Holttum, 24560).
Negri Sembilan, Gunong Angsi (Nur, 11571; Ridley,
10121) ; Tampin (Burkill, 1417) ; Gunong Tampin (Burkill,
41171; Holttum 9557); Beremban along Sungai Bendol
(Furtado, 33125). Malacca, Ayer Panas (Goodenough,
1406; Ridley & Goodenough, 1577); Bukit Besar Ophir
(Ridley, 10120 & 3474); Selandan (Ridley, 10793) ; Bukit
Tampin (Goodenough, 1962).

Distribution: recorded from Mergui in Lower Burma.

GRIFFITH cites the syntypes of this species thus:
“Malacca, solitary in dense forests, Ayer Punnus (Rhim),
Goonoong Miring, and Mount Ophir, but not above an
elevation of a thousand feet. Tenasserim coast in forests
near Lainear to the south of Mergui.”

BECCARI found only one syntype in the Calcutta
herbarium with no indication as to its locality, and since
he had not seen any specimen definitely coming from the
Malay Peninsula, he gave Tenasserim as the probable origin
of this GRIFFITH’S specimen. However from the description
and the drawings given by BrEccari I think the species is
common in the Peninsula, especially in the regions where
GRIFFITH saw it.

From the notes given by collectors it appears that this
species is very common in Malacca and Negri Sembilan;
also that though it is easy to find plants in flower, it is rare
to find them in fruit. There are a few specimens which
have been collected above a height of 1,000 ft., and in some
of these the flower-bearing pulvini are quite prominent (e.g.
BURKILL 1171 and HOLTTUM 9957); but there are transi-
tions between these and the usual forms where the pulvini
are nested in depressions.

There is some variation in the dentation of the calyx;
some split very early during the development of the flowers
and here the lobes are rounded at the apex and are often
bifid; in others this lobation is retarded and the calyx looks

Vol. XI. (1940).

58

almost truncate and later split irregularly as in L. malajana,
a species which appears to have very close affinities with
L. longipes.

GRIFFITH records that this species is known to Malays
as Palas Batu, a name recently recorded also by a Malay
collector who notes that the leaves are used for thatching
house roofs.

14. Licuala malajana Becc., Malesia III (1889) 197;
Hook. f., Fl. Brit. Ind. VI (1892) 431; Ridl.,
Mat. II (1907) 161 pro parte; Becc. in Webbia V
(1921) 35 et 55; Ridl, Flora’ V-(1925)' 26 op.
Becce. in Cale. Annals XIII (1933) 196 tt. 14-IV
et 87.

MALAY PENINSULA: Trengganu, Brang in Tersat, alt.
2,500 ft. (Kiah & Moysey, 33398). Kemaman, Kajang at
Ulu Bendang, alt. 500 ft. (Corner, 30072 A) ; Sungai Nipa
(Corner, 30547).

The following specimens have more hairy calyces, but
seem to belong here:

Kelantan, Kuala Krai (Haniff & Nur, 10104). Pahang,
Bukit Senai (Henderson, 19434). Selangor, Pahang Track
(Ridley in 1897). Johore, Gunong Muntahak, alt. 600 ft.
(Nur, 19968).

I have not seen any authentic specimens named by
BECCARI and so my identification of the species is based
entirely on the description and plates given by BECCARI.
The specimens doubtfully cited here come very near to some
forms of L. modesta which however produces very much
shorter spadices bearing flowers in groups of two or three
on very prominent, almost pedicelliform tubercles. The
typical calyx of L. malajana has some resemblance to that
of L. Ridleyana, but the latter species produces unbranched
partial inflorescences.

15. Licuala mirabilis Furtado sp. nov. Fig. 6.

A L. Kunstleri, cui peraffinis, spathe limbo pedicellum
axillaris spice plerumque valde superante, spathis superne
inflatis, spicis (=inflorescentiis partialibus) 2-5 ramulosis
haec species sat distincta. Secundum clavem Beccari
analyticam prope L. baculariam ponenda, a qua spadicibus
quam petioli valde brevioribus facile distinguitur.

Palma solitaria vel sobolifera, cum caule 1-1:'5 m.
longo 3-4 m. alta. Petiolus circa 15 m. longus, vagina
fibrosa mox marcescente decidua preditus, in 2/3 partibus
basalibus aculeis reduncis, inferioribus robustis 4 mm. altis,
1-2 mm. crassis, superioribus obscuris armatus. Segmenta
frondium 20-27, basin versus valde angustata, fere petiolu-
lata, apice 3-5-7 cm. lata, 5-15 mm. profunde dentata, 2-3,
raro 4-6, costata; medianum alteris parum majus 3-6

Gardens Bulletin, S.S.

fy
Y

Fig. 6. Licuala mirabilis (Holotypus: FURTADO 33054).

a. Pars spadicis floridi. b. Frondis lamina cum parte petioli.
c. Fructus. d. Semen verticaliter discissum. e. Alabastrum. f. Pul-
vinus cum flore dua carpella fertilia exhibente. g. Filamentum

staminale.

Vol. XI. (1940).

60

costatum, minus profunde dentatum. Spadices petiolo multe
breviores, ad 45 cm. longi. Spathx fugaceo furfurace basi
tubulosz superne subito inflate, apice acute, in uno latere
fisse, partem pedicellarem spicarum axillarium plerumque
superantes. Inflorescentia partialis in spiculas 2-5 axi
pubescentes divisa, pedicello ad 2-5 cm. longo, infra limbum
spathe axillantis plerumque oriente, saepe quam limbus
spathe minore. Flores conspicue pedicellati, geminati vel
solitarii. Calyx turbinato campanulatus, pubescens, apice
membranaceus, lobatus vel partim truncatoides, in basin
cylindricam induratam abrupte constrictus. Corolla calyce
duplo longior, dense sericea. Annulus staminalis in fila-
menta 6, aequalia superne abrupte subulata. Ovarium
glabrum. Fructus globosus, utrinque acutus, 10-12 mm.
longus, 6-8 mm. crassus; albumine carneo, albescenti, fossa
integumentale cylindrica, vertice paulo dilatata, interdum
biapicali.

MALAY PENINSULA: Kedah, Ayer Térjang Valley in
Baling (Furtado, 33054—Holotypus) ; Yan (Ridley in June
18938). Province Wellesley, Permatang Bertam? (Ridley
in 1895) ; Bukit Mertajam, alt. circa 450 m. (Burkill, 9021).
Perak, Bukit Merbau Balong in Kroh (Furtado, 33037).

This species is closely allied to L. Kunstlevi, but the
branching of the partial inflorescences distinguishes them
at once, L. Kunstlerit having always simple branches to the
spadices. From L. ferruginea, with which L. mirabilis has
been confused, it is distinguished by the non-marcescent
apex of the spathes, the pedicelled flowers and the
membranous calyx having an attenuate and constricted base.

16. Licuala modesta Becc., Malesia III (1889) 195;
Hook. f:, Fl. Brit. Ind. VI (1892) 43a Baume
Mat. II (1907) 163; Becc. in Webbia V (1921)
35 et 48; Ridl.,. Flora V (1925): 28; Becca
Cale. Annals XIII (1933) 198 tt. 14—-VI et 88.

L. Wrayi Bece. ex Ridl. in Journ. Roy. Asiat. Soc.
Straits Br. 82 (1920) 201; Becc. in Webbia V (1921) 37 &
50; Ridl., Flora V (1925) 28; Becc. in Cale. Annals XIII
(19338) 210 tt. 13-I et 46. Syn. nov.

MALAY PENINSULA: Perak, Taiping Hills (Hervey, in
1889; Ridley, 14721, in Dec. 1902 and in Feb. 1904; Furtado,
37103); Kuala Kangsar (Ridley, 3167); Gunong Hijau
(Fox sub Curtis no. 3529; Burkill & Haniff, 12873 & 12589;
Anderson, 164 & 171) ; Tea Gardens (Ridley, 11406 and in
Feb. 1891); Maxwell’s Hill (Burkill & Haniff, 12948, &
12551; Ridley in June 1893 & Feb. 1892; Curtis, 2077
partim, ex altera parte=Calamus sp.) ; Gunong Pondok, alt.
1,000 ft. (Henderson, 23804) ; Box Hill, alt. 4,200 ft, (Fox,
163) ; Larut Hill (Anderson, 177).

Gardens Bulletin, SS.

61

I am unable to detect any appreciable differences
between the forms described by BECCARI and RIDLEY as
L. Wrayi and L. modesta. The species produces very short
inflorescences; in some there may be a few partial inflores-
cences separated by more or less elongated tubular spathes ;
in others the number of branches and the size of the spathes
are reduced, so that the entire spadix may appear to bear
directly the spikelets on its axis, though one or two
liguliform spathes about 3-6 cm. in length will indicate the
reduction that has taken place in the spathes; and in still
others the ligule may be further reduced into small,
inconspicuous bracteoles. If the definition of LICUACELLA
were accepted, the specimens bearing these small bracteoles
could not be kept in the subgenus that includes L. modesta,
a reason why BECCARI was obliged to erect L. Wray.

Some of the specimens cited above were referred by
RIDLEY (1907 and 1925) to L. malajana and to L. pusilla.

17. Licuala Moyseyi Furtado sp. nov. Fig. 7.

A L. Corneri, cuz valde affinis, recedit haec species:
petiolo inermi; segmentis frondium inaequalibus, mediano
latissumo; ramulis floriferis furfuraceis, haud pubescenti-
bus; pedicellis fere tuberculiformibus; floribus in spicis
imferioribus plerumque 2-3 glomeratis; calyce fugaceo
furfuraceo, haud piloso, minore, apice conspicue 3—lobato.

Petiolus gracilis, inermis, 75-90 cm. longus, ad basin
circa 4 mm. latus, apice paulo angustior. Segmenta
foliorum 8-10 dentibus 5-10 mm. altis; medianum latis-
simum, 10-14 costatum, 22—25 cm. longum, apice 10-15 cm.
latum; altera angustiora, minora, 3-4, raro 2-5 costata;
basilaria minima, apice oblique erosa. Spadices 60-75 cm.
longi; in parte basilari haud ramosa, 35-40 ecm. longa,
spathis 2-3 involuti; in altera parte in ramos floriferos 3-4,
simplices, fusco furfuraceos divisi. Spathx tubulares,
fugaceo fusco furfuracez, superiores obscure biangulate,
haud inflate, ad 6 mm. crasse, apice oblique truncate ;
infima latior conspicue biangulata. Flores in tuberculis
prominentibus insidentes, 1-3 glomerati, 3-4 mm. longi,
clavati. Calyx fugaceo fusco furfuraceus, conspicue 3—
lobatus, in alabastro apice globosus, basin versus pedicelli-
formiter contractus. Corolla juvenilis tantum visa, haud
pubescens. Annulus staminalis in filamenta aequalia
abrupte subulata divisus. Ovarium glabrum.

_ MALAY PENINSULA: Trengganu, Gunong Padang, alt.
circa 13800 m. (Kiah et Moysey, 31840).

This species may easily be confused with L. Corneri,
for both have simple, unbranched partial inflorescences and
they are the only species in the peninsula having a long
pedicelliform base to the calyx. However L. Moyseyi can

Vol. XI. (1940).

62

Fig. 7. Licuala Moyseyi (Holotypus: Kian & Moyszy 31840).

a. Petioli pars inferior. b. Frondis lamina cum petiolo. c. Spadix.
d. Pars ejusdem apicalis. e. Pars spicae ut dispositio florum appareat.
f. Alabastrum verticaliter discissum. g. Alabastrum sine perianthio
ut aestivatio staminum appareat. h. Calyx fructiferus.

Gardens Bulletin, S.

63

be distinguished from L. Corneri by its thornless petiole;
unequal leaf-segments, the median being 2—3 times broader
than the others; furfuraceous, not hairy spikelets; smaller,
almost tubercular flower-pedicels ; flower-glomerules consist-
ing frequently of 3 flowers; smaller, not hairy, and distinctly
3—lobes calyx.

The flowers present in the specimen are not fully
developed, the reason why the description of the petals is
so meagre. One fully developed calyx was found on a
spadix from which all fruits had fallen off; this calyx is
represented in the drawing to show the depth of the calyx
lobes.

18. Licuala pahangensis Furtado sp. nov. Fig. 8.

L. malajana sensu Ridl., Mat. II (1907) 161 et Flora
VY ALOZb) 26 p.p.

L. spinosa sensu Ridl. in Trans. Linn. Soc. Lond. III
(1893) 391 p.p.

Inter species sapdice composito ramulis_ floriferis
simplicibus ponenda; L. Ridleyanz et L. peninsulari valde
affinis, sed calycibus striatis, glabris, apice alto lobatis, basi
truncatis vertice basali caudiculatis; petalis glabris; sat
distincta.

Petiolus 50-100 cm. longus, tertia parte basali aculeatus.
Segmenta frondium subaequalia; centralia 4—6 costata, 3-5-5
cm. lata, circa 45 cm. longa, paulo latiora et minus profunde
dentata alteris 2-3 costatis paulo brevioribus. Spadices
petiolis fere aequilongi, vel eis paulo breviores, ramis
simplicibus, ad 15-20 cm. longis, supra spathe apicem
remoto orientibus. Spathx tubulose, fugaceo furfuracee.
Flores solitarii, pluseriatim dispositi, circa 4-5 mm. longi.
Calyx membranaceus, valde striatus, glaber, apice in 3 lobos
asymmetros divisus interdum irregulariter fissus, basi
truncatus vertice caudiculatus. Corolla glaber, calyce duplo
longior. Annulus staminalis in filamenta versus apicem
arcuatim angustata divisus. Ovarium glabrum, loculis
saepe duobus vel tribus fertilibus. Fructus maturus ignotus.

MALAY PENINSULA: Pahang, ad radicem montis
Gunong Senyum (Henderson, 22305-—Holotypus); Titi
Bungor, Temerloh (Henderson, 10623); Tanjong Antan
(Ridley in 1891) ; Pulau Chengai (Ridley in 1891).

From the species bearing simple spadix-branches L.
pahangensis is readily distinguished by the striate, glabrous
calyx which is deeply lobed at the apex and truncate at the
base, where it has a small tail-like projection. Otherwise
this species may be easily confused with L. peninsularis and
L. Ridleyana.

Vol. XI. (1940).

64

Qa

Fig. 8. Licuala pahangensis (Holotypus: HENDERSON 22305).

a. Pars spadicis. b. Alabastrum. c. Alabastrum verticaliter

discissum. d. Alabastrum, petala resecta, ut aestivatio staminalis
DAGGETT EN, REID nn ne ee ig i vomblel sharhasleynlur obs!

65

19. Licuala paludosa Griff. in Calc. Journ. Nat. Hist.
V (1848) 2383 et Palms Brit. Ind. (1850) 118
tt. 221 A, B, et C; Hook. f., Fl. Brit. Ind. VI
(1892) 480; Ridl., Mat. II (1907) 160 p.p.;
Becc. in Webbia V (1921) 24 et 48; Ridl., Flora
V (1925) 25 p.p.; Bece. in Cale. Annals XIII
(1933) 159 tt. 8-XI, et 98.

L. paniculata Ridl. in Journ. Roy. Asiat. Soc. Straits
Br. XLI (1903) 42, et Mat. II (1907) 163; Becc. in Webbia
V (1921) 29 et 43; Ridl., Flora V (1925) 30; Becc. in Cale.
Annals XIII (1933) 161 tt. 10-I et 46. Syn. nov.

L. patens Ridl. in Journ. Roy. Asiat. Soc. Straits Br.
LXXXII (1920) 202 et Flora V (1925) 25. Syn. nov.

L. spinosa sensu Ridl. in Kew Bull. (1926) 91.

MALAY PENINSULA: Perak, Hermitage Hill (Ridley in
1892, Holotype of L. paniculata) ; Sungai Krian Estate, alt.
sea level (Spare, 33288) ; Gunong Bubu alt. 4,000 ft. (Wray,
3928, Haptosyntype of L. patens). Selangor, Sungai
Tinggi, near Kuala Selangor (Nur, 34076). Johore, Hadji
Senawi (Ridley, 11212).

SUMATRA, Siberut in Mentawi Islands (Boden-Kloss,
14614).

Distribution: Borneo and Indo-China.

RIDLEY has referred to this species many specimens of
L. spinosa. Perhaps he had these specimens in mind when
he erected L. patens.

3 L. paludosa is easily distinguished from L. spinosa by
its calyx being minutely puberulous or almost glabrous
outside and irregularly and shortly lobed at the apex. I
do not find any important difference between the specimens
collected in the lowlands and those collected in the
mountains, and so I have reduced L. patens to a synonym
of L. paludosa. There are some small differences noticeable
between the types of L. paniculata and L. paludosa, but
there are transitions; but even without these transitions the
differences are so trivial that I have no hesitation in
reducing L. paniculata to L. paludosa. (These differences
have been noticed by BECCARI in Calcutta Annals l.c.).

In erecting L. paniculata (1903) RIDLEY remarked that
it and L. penduliflora from New Guinea were the only
species known to him to have a “panicled” inflorescence,
which, according to him, characterise the section PERICYCLA.
Since he did not include in this section L. paludosa, L.
longipes and L. patens, even in his most recent work
(1925), it is not easy to understand RIDLEY’s conception
of the section PERICYCLA (sometimes spelt by RIDLEY as
Pericyclus).

Vol. XT. (1940).

66

BECCARI has reduced L. amplifrons Miq. to L. paludosa,
but the specimens growing under that name in the Botanic
Gardens, Buitenzorg, Java, appear to be different, though
BECCARI quoted material from these plants under UL.
paludosa. In 1936 I collected herbarium material from
these plants in Buitenzorg and distributed it under the
Singapore Field no. 31102 and 31129. The plants can be
distinguished from the typical L. paludosa in being solitary,
bearing shorter petioles, fewer branches to partial inflores-
cences, longer fiower-bearing pulvini, and sometimes the
flowers borne in pairs on each pulvinus. The material looks
so different that a further inquiry into the status of L.
amplifrons Miq. seems to be necessary.

20. Licuala pusilla Becc., Malesia II] (1889) 194;
Hook. f., Fl. Brit. Ind. VI (1892) 433; Ridl., Mat.
Il (1907) 164; Becce. in Webbia V (1921) 31
et 45; Ridl., Flora V (1925) 28; Becc. in Cale.
Annals XIII (1933) 174 t. 85.

L. acutifida sensu Ridl. op. cit. (1907) 163 et (1925)
27. p. parte.

apis PENINSULA: Kelantan, Gua Ninek (Henderson,
19686). Selangor, Rantau Panjang (Hume 7619; Ridley
in Aug. 1909); Kanching (Foxworthy and Burkill on 30
Nov. 1921). Negri Sembilan, Gunong Angsi (Ridley in
Feb. 1904). Malacca, Sungai Udang (Goodenough, 1360
as Palas Padi).

Of this species I have not seen any specimens named
by BECCARI, and so my identification of it is based entirely
on BECCARI’S description and plate. The young floriferous
branches are covered with whitish tomentum.

This species appears to be very near to L. triphylla
some forms of which bear many, narrow almost linear
segments to the leaves, the median lobe being sessile.
Further investigation in the field may prove L. pusilla to
be only a form or variety of L. triphylla.

21. Licuala Ridleyana Becc. in Webbia V (1921) 31
et 44 et in Cale. Annals XIII (1933) 170 tt.
10-III et 34.

L. malajana sensu Ridl., Mat. II (1907) 161 et Flora
V (1925) 26 p.p.

MALAY PENINSULA: Perak, Bujong Malacca (Ridley,
9805, et 9806, Haptoparatype of L. acutifida var penin-
sularis) .

From the description and drawings given by BECCARI
and from the Singapore duplicates of the specimens cited
by BECCARI, the flowers of the specimens cited under L.
Ridleyana and L. acutifida var. peninsularis (see L. confusa)
appear to have been interchanged either by RIDLEY at the

Gardens Bulletin, S.S.

67

time of distributing his collections or by BECCARI at the
time of drafting his description. In the Singapore her-
barium spadices of RIDLEY 9806 (cited by BECCARI under
L. acutifida var. peninsularis) and RIDLEY 14112 (Hapto-
holotype of L. Ridleyana) still bear flowers; but RIDLEY
9806 has flowers depicted for L. Ridleyana and RIDLEY
14112 has flowers depicted for L. acutifida var. peninsularis.
As I have typified the species on the characters of flowers,
which are more stable and of better diagnostic value than
the characters of leaves, spathes, etc., investigators should
note these points in case discrepancies be noticed in the
original specimens of these two taxonomic units.

As described by BECCARI, L. Ridleyana approaches very
near to L. malajana which has however branched partial
inflorescences. (see also observations under L. confusa).

22. Licuala Scortechinii Becc., Malesia III (1889) 192;
Hook. f., Fl. Brit. Ind. VI (1892) 434; Ridl.,
Mat. II (1907) 164; Becc. in Webbia V (1921)
37 et 49; Ridl., Flora V (1925) 30; Becc. in
Cale. Annals XIII (1933) 208 tt. 138-IV, 59 et
59 bis.

MALAY PENINSULA: Kedah, Gunong Jerai —Kedah
Peak (Haniff in May 1904; Ridley, 5201; Robinson and
Kloss, 6016).

This species was based on a specimen collected by
SCORTECHINI in Perak where it has not been found again.

23. Licuala spinosa Wurmb in Verh. Bat. Genootsch.
II (1780) 469; Bl., Rumphia II (1844 ?) 39 tt
82 et 88; Hook. f., Fl. Brit. Ind. VI (1892) 4381;
Ridl., Mat. II (1907) 160; Bece. in Webbia V
(1921) 34 et 47; Ridl., Flora V (1925) 25; Becc.
in Cale. Annals XIII (1933) 186 t. 84.
MALAY PENINSULA: Perlis, Mata Ayer (Henderson,
23061). Kelantan, Sungai Keteh (Nur, 12077). Lankawz1,
common (Curtis, 2129); Coah (Curtis, 3419). Pahang,
Pekan (Ridley on Aug. 20, 1889) ; Tasek Bera (Henderson,
24433) ; Pulau Tioman (Nur, 18901; Burkill in June 1915).
Perak, Dindings (Ridley, 3170); Matang Jambu (Wray,
2527). Malacca, Bukit Bruang (Hassan, 25). Johore,
Sungai Segal in Muar (Fox, 11303) ; Sungai Rhu (Corner,
28485) ; Sungai Sedili in Mawai (Holttum & Corner, on
16 May 1982); Pulau Tinggi (Feilding in 1892).
Singapore, Pulau Ubin (Ridley, 5121 et 3166; Hullett or
9 Nov. 1884) ; Pulau Battam (Ridley in 1890 and in 1891) ;
ony Brani (Ridley in 1900); Changi (Ridley in April
Distribution: Lower Burma, Siam, Indo-China, Anda-
mans, Sumatra, Java, Borneo, and the Philippines.

Vol. XI. (1940).

68

24. Licuala tiomanensis Furtado sp. nov. Fig. 9.

Inter species inflorescentias partiales simplices gerentes
calyce majore superficialiter lobato dense adpresse piloso sat
distincta. Facie florum L. modeste similis, sed spadicibus
longioribus, inflorescentuis partialibus remotioribus simplici-
bus dissimilis.

Palma humilis. Petrolus unicus tantum visus, 22 cm.
longus, infra medium aculeatus. Segmenta foliorum multa,
inaequalia, cuneata; medianum maximum, circa 28 cm.
longum, 9 cm. latum, pluricostatum; basilaria 2—costata,
minima, 15-17 cm. longa, circa 2 cm. lata; intermedia 3—4
costata. Spadix circa 80 cm. longus, in ramos 2-3, pilosos,
simplices, infra spathe apicem orientes divisus. Spathz
tubulose, fugaceo-furfuracese. Flores immaturi lanceolati,
6-8 mm. longi, super pulvinos prominulos plerumque
geminati. Calyx 4 mm. longus, dense pilosus, apice
truncatus, breviter lobatus vel in 3 lobos obscure fissus.
Corolla ante anthesin calyce 1/3 longior, puberula, lobis
acutis. Annulus staminalis conspicuus, in 6 filamenta
abrupte subulata divisus. Ovarium glabrum.

MALAY PENINSULA: Pahang, Bukit Kajang in insula
Tioman, alt. 3,300 ped. (Nur, 18611).

This species resembles L. Ridleyana, L. peninsularis,
etc. in having simple branches to the spadix, but in the
characters of the flowers it is easily confused with L.
modesta. Unfortunately L. tiomanensis is known only from
a single specimen containing only one leaf which may not
be typical of the species. Fruits are not known.

Nur, 18910 collected also on Gunong Kajang in Pulau
Tioman at an alt. 2,500 ft. seems to belong here, but no
flowers are present in the Singapore herbarium, though
from the collector’s notes it appears that the flowers were
present in the collection. The petiole is very much longer
in this specimen and may have been over 314 feet long; the
partial inflorescences arise very high above the apex of
the spathes. The collector notes that the plant which
supplied this specimen is 15-20 feet in height.

25. Licuala triphylla Griff. in Cale. Journ. Nat. Hist.
V (1845) 332 et Palms Brit. Ind. (1850) 126
t 225; Hook. f., FI, Brit. dnd VY (1892) 4225
Ridl., Mat. II (1907) 163; Becc. in Webbia V
(1921) 33 et 46; Ridl, Flora VY (1925)8232.
211; Bece. in Cale. Annals XITI (1933) 182 Tt.
9-V et 96.

L. pygmaea Merr. in Univ. Calif. Publ. Bot. XV (1929)

20. Syn. nov.

Gardens Bulletin, S.S.

69

Fig. 9. Licuala tiomanensis (Holotypus: Nur 18611).

a. Pars spadicis. b. Frondis lamina cum parte petioli. c. Pars
spicae cum alabastris. d. Flos ad anthesin. e. Flos verticaliter
discissus ut filamenta staminum et ovarium appareant.

70

L. triphylla var. integrifolia Ridl., Mat. II (1907) 164;
Bece. in Webbia V (1921) 33 et 46; Ridl., Flora V (1925)
28; Becce. in Cale. Annals XIII (19383) 184. Syn. nov.

MALAY PENINSULA: Kedah, Bukit Seblak at Weng
(Furtado, 33069). Pahang, Gunong Senyum (Henderson,
22356) ; Batu Talam near Raub (Burkill & Haniff, 16959) ;
Lubo Pellang (Ridley in 1891, the holotype of the var.
wntegrifolia) ; Pulau Tijou (Ridley in 1891); Dong near
Raub (Burkill & Haniff, 16911). Perak, Kroh (Furtado,
33041) ; Bujong Malacca (Curtis in Aug. 1898); Gunong
Tungul in Dindings (Ridley in May, 1896 and July 1898).
Negri Sembilan, Tampin (Burkill, 3212; Nur 1308).
Malacca, Chabau (Alvins, 2306 as Palas Revang); Ayer
Panas (Ridley & Goodenough, 1628); Malacca Town
(Alvins, 655, as Gurcheng or Palas Tikus) ; Bukit Bruang
(Curtis, in May 1901). Johore, Sungai Tebrau (Ridley,
13234) ; Bukit Keara in Muar (Fox, 11301; Ridley in 1902) ;
Sungai Buloh Kasap (Corner on 5 Jan. 1936) ; Bukit Tinjau
Laut (Corner, 37079). Singapore, Changi (Ridley in 1891
& 1894).

BRITISH NORTH BORNEO: Tawao (Elmer, 21635,
Haptosyntype of L. pygmaea).

Distribution: Bangka (?) fide Beccari.

L. triphylla is a very variable species as regards the
number of the leaf segments and the size of the petiolule
of the median segment, some median segments being quite
sessile. Owing to these variations field studies are necessary
to see whether L. pusilla is a special form of this species.
L. triphylla flowers when the plants are very small, so that
leaves showing no divisions or leaves partially divided into
2 or 3 lobes may be found. In the holotype of L. triphylla
var. integrifolia some leaves have very narrow segments
at the base; these have been ignored by RIDLEY in describing
the variety. Transitions from this to many-segmented
leaves are many.

Many leaves bear sessile median segments, so that this
cannot be taken as an invariable character of the species.
GRIFFITH’S plate depicts leaves with sessile median lobes and
so also RIDLEY’s figure which appears to have been based
mainly on GRIFFITH’S plate.

I cannot separate L. pygmaea from L. triphylla on any
characters. The median segment in the duplicate specimen
in Singapore shows a distinct tendency to petiolulation.

6. Summary and Conclusions

This paper contains the results of a study of the Licuala
species indigenous in the Malay Peninsula.

Gardens Bulletin, SS.

71

Though Licualas present little difficulty for. collecting
herbarium specimens, yet they are not well represented
in the herbarium. The principal reason for this is that
some experience is required before the collector is able to
distinguish between the different species he meets with in
the field, and so very often rare species, being mistaken for
the common ones, are omitted in the collection. Hence some
hints are given in this paper to enable the collector to
distinguish readily between the species of a given area.

An inquiry into the affinities of the Malayan species
revealed the necessity of revising the subdivisions of the
genus. The previous subdivisions were mostly based on
the branching of the spadices, a character very variable and
therefore unsatisfactory for the purpose. It has been found
also that all Licwala species produce a dorsal embryo in the
fruit so that the definition that PERICYCLA includes species
with basal embryo is untenable. The only stable character
which can be employed to place allied species into large
groups is found in the andrecium, and in the nature of
its divisions.

As revised here, the genus Licuala consists of three
subgenera: LIBERICULA (new), EU-LICUALA and PERICYCLA,
the second subgenus being again subdivided into four
sections: WURMBIA (new), BONIA (new), BECCARIA (new)
and DAMMERA (new state). The subgenus LIBERICULA
and the sections BONIA and BECCARIA are monospecific,
containing respectively: L. peltata (in monsoon forests in
North-East India, Assam, Burma, Andamans and Thailand
(Siam) ), L. fatua (Indo-China) and L. reptans (Borneo).
Of the section DAMMERA only two species are known, both
from New Guinea. The subgenus PERICYCLA has been
re-defined to include all the species which are allied to
L. penduliflora (Bl.) Migq., the type of the subgenus; these
inhabit the tropical and sub-tropical regions in Oceania,
one occurring in Australia, a few in the small islands west
of the Solomon Islands, and the rest in New Guinea. The
section WURMBIA includes the great majority of the Licuala
species, the greatest number of these being limited to the
Malay Peninsula, Sumatra and Borneo, though some also
occur outside this region in Burma, Indo-China, South
China, Java, Philippines, Celebes and New Guinea. Usually
the distribution of the individual species of this section is
very limited; L. spinosa, the type of the section, is the most
widespread, being also of a frequent occurrence throughout
the Malay Peninsula.

There are in all twenty-five species indigenous in the
Malay Peninsula. Of these L. peltata is the only species
that is not a member of WURMBIA. Besides L. peltata and

Vol. XI. (1940).

72

L. spinosa, four peninsular species are known or reported
to occur wild outside the Malay Peninsula, namely L.
ferruginea, L. longipes, L. paludosa and L. triphylla; they
are also fairly widespread within the Peninsula itself. The
other nineteen are all endemic, most of them being
apparently restricted to small regions in the Peninsula.
Much of this endemism may be more apparent than real,
because, owing to the difficulty of distinguishing Licualas in
the field, the species are not well represented in the
collections. Furthermore, recent collections from Sumatra,
Borneo, and Siam have not yet been worked out.

The new name L. Beccariana has been proposed for the
New Guinea species, previously known as L. racemosa Becc.
(DAMMERA), non L. racemosa Bl. (WURMBIA). In addition
nine new species have been described from the Malay
Peninsula: L. confusa, L. Corneri, L. kemamanensis, L.
Kiahii, L. longicalycata, L. mirabilis, L. Moyseyi, L.
pahangensis, and L. tiomanensis.

The following names have been reduced in the
synonymy: L. acutifida var. peninsularis (L. confusa,
doubtful), L. longepedunculata (L. glabra), L. paniculata
and L. patens (L. paludosa), L. pygmaea (L. triphylla), L.
racemosa Becc., non Bl. (L. Beccariana), L. triphylla var.
integrifolia (L. triphylla) and L. Wrayi (L. modesta).

It has been shown that L. ferruginoides produces at
times branched partial inflorescences (vide observ. sub. L.
ferruginea) and that a further inquiry is needed before
Beccari’s reduction of L. amplifrons to L. paludosa can be
accepted as final.

7. Index to the Collector’s Numbers

Alvins: 655 & 2306(25). Anderson: 164(16); 171 &
177(16). Best: 7711(11); 18884(5). Boden-Kloss:
14614(19). Burkill: 1171 & 1417(18); 3212(25);
9021(15). Burkill & Haniff: 12551 & 12589(16);
12873 & 12948(16); 16911 & 16959(25); 17460(10).
Burkill & MHolttum: 7842(5); 8426(6). Corner:
28485 (23) ; 28688(4) ; 30072(3) ; 300724 (14) ; 30111
(13); 30398(4); 30520(7); 30547(14); 307389(11) ;
37068(11); 37079(25). Curtis: 1010(1); 2077(16) ;
2129 (23) ; 3162(9) ; 3419(23). Derry: 633(5). Elmer:
21635(25). Fox: 163(16); 3529(16); 11301(25);
11303(23). Furtado: 33037 & 33054(15) ; 33069(25) ;
33125(18) ; 371038(16). Goodenough: 1360(20); 1406
(13); 1962(13). Haniff & Nur: 4720(A); 6962(9) ;
8111(5); 10104(14). Hassan: 25(23). Henderson:
10559(13) ; 10623(18) ; 19434(14); 19686(20) ; 22305
(18). 223856(25) ; 23061 (23) ; 238804(16) ; 24433 (23) ;

Gardens Bulletin, S.S.

73

29725(5); 33041(25). Holttum: 9252(12); 9260(4) ;
9557(13); 10601(16); 10605(4); 10792(11); 24560
(18); 24944(4). Hullett: 852(5). Hume: 7619(20) ;
7962(2) ; 9437(10). Kiah: 24256(5) ; 32137(8) ; 32401
(12). Kiah & Moysey: 31840(17); 33398(14) ;
33399(6). Kloss: see Boden-Kloss. Lobb: 280(1).
Moysey: see Kiah & Moysey. Nur: 1308(25); 11571
(18); 12077(238); 18611(24); 18901(23); 18910(24
obs.) ; 19968 (14) ; 25167(4) ; 34076(19). Ridley: 3166
(23); 3167(16); 3168(4); 3170(23); 3473(5); 3474
(18) ; 3512 (4) ; 5121 (23) ; 5201 (22) ; 6676(4) ; 7894 (5) ;
9804(9); 9805(21); 9806(21) 10120(13); 10121(18) ;
10239 (2 obs.) ; 10329(2 obs.) ; 10793(18) ; 11212(19) ;
11406(16) ; 12117(6); 12198(11); 18234(25); 13519

(ay Tati 2(2y 3s 14709(10); 14721 (16) ; 315881 (6).
Ridley & Goodenough: 1577 (13) ; 1628(25). Robinson
& Kloss: 6016(22). Spare: 33288(19). Symington:

24379 (12) ; 24880 (4) ; 25612(9) ; 25724(9) ; 47068 (12) ;
47092 (12).

8. Index to the Botanical Names

Sectional and Subgeneric names in small CAPITALS.
in bold faced type. Synonyms in italics.

New names

BECCARIA Licuala:
BONIA mirabilis sp.n.
DAMMERA tr. nov. modesta

Moyseyi sp.n.
pahangensis sp.n.

DAMMERA subg.
EU-LICUALA

LIBERICULA paludosa
LICUACELLA paniculata
Licuala: patens
acutifida peltata
acutifida var. peninsularis pusilla
(obs.) pygmaea

amplifrons (obs.)
Beccariana nom. nov. (gen.)
confusa sp. nov.

Corneri sp. nov.

racemosa (gen.)
Ridleyana (obs.)
Rumphii (gen.)

ferruginea Scortechinii
ferruginoides (obs.) spinosa

glabra tiomanensis sp.n.
glabra var. selangorensis triphylla

grandis (gen.) triphylla var. integrifolia
kemamanensis sp. nov. Wrayi

Kiahii sp.n. LICUALINA

Kingiana LICUALOPSIS
Kunstleri Pericycla emend. (obs.)
lanuginosa PERICYCLA
longepedunculata WURMBIA
longicalycata sp.n.

malajana

Vol. XI. (1940).

74

THE FERN GENUS DIPLAZIUM IN THE
MALAY PENINSULA

by R. E. HOLTTUM

General

The ferns of the genus Diplazium occurring in the
Malay Peninsula are all, with the exception of D. esculentum
and D. polypodioides, found on the ground in shady forest,
many of them only in moist places near streams. They are
especially characteristic of the valleys at medium elevations
in the mountains, though a few species are common in the
more open parts of the forest. There are no epiphytes and
no climbing species among them. Nearly all have short
erect rootstocks (sometimes developing a short trunk) and
tufted fronds, sterile and fertile fronds being almost alike.

There is comparatively little difference of external form
among the species of the same degree of division (2.e.,
among pinnate and among bipinnate species), and there is
of course a certain amount of variation according to age
and infiuences of habitat, so that the clear distinction of
species is at first sight difficult, and there has in fact been
much confusion in the past. I have found, however, that
the characters of the scales are most useful in distinguishing
species of similar habit, and I believe that I have arrived
at a clear distinction of our local species, with the possible
exception of D. Prescottianum.

I have been unable to identify four of our species with
any described from places outside the Peninsula, and so
have described them as new. There are undoubtedly a
very large number of species within the Malayan region,
but I think that a good many have been described more than
once. The descriptions are, however, inadequate as a basis
for comparison, and much herbarium material is incomplete
(part only of a leaf, or lacking in scales, or representing
an immature stage). A study of much material over a
wide area is necessary before a proper account of the genus
in the Malayan region can be given.

Diplazium cordifolium provides a remarkable case of
polymorphism leading to confusion of names. Young
plants develop rather large simple fronds which are often
fertile, and this stage may persist indefinitely, especially
in the lowlands. Mature plants, however, .often bear
pinnate fronds with several pairs of pinnze. The two
stages have been separately named; and the fact that other
superficially similar simply pinnate species have sometimes
a few veins anastomosing has also led to confusion between
these and the mature form of D. cordifolium. The young

Gardens Bulletin, S.S.

75

stages of species may be characteristic, and so useful
diagnostically, in this genus as in other fern genera, but
they have been little studied. I have noted especially the
early stages of D. bantamense and D. riparium.

In the following account of the species found in the
Malay Peninsula, I have not attempted to give full lists
of synonyms. There has been so much confusion in the
application of names that such lists of synonyms would
be difficult to compile accurately. I believe that the names
used are all correctly applicable to our species; but I differ
in some cases from the usage of the recent book by Backer
and Posthumus on the ferns of Java. In such cases, the
synonymy is briefly discussed.

As regards citation of specimens, this is done only
where I believe there has been confusion in the past, or in
the case of new species; in cases of well characterized species
such as D. accedens and D. cordifolium there seems no need
for the publication of lists of specimens. Specimens
collected in our Singapore Field Number series (started
by Burkill in 1912) are cited asS.F.N. Duplicates of many
of these will be found in other herbaria, in some cases under
other names than those now given. A list of numbers at
the end of this paper may therefore prove helpful. It should
be noted that the earlier distribution labels did not bear
the words “Singapore Field Number’’.

A fairly full description is given for each species. To

aid in the identification of specimens, the most distinctive
_ characters in each description are italicised.

Distribution

Until a full investigation is made, it is impossible to
give a complete picture of the distribution of the species
of Diplazium found in the Malay Peninsula. There are
certainly cases in which one name is used in different
countries to cover different species, and conversely the same
species sometimes goes by different names in different
countries. From the material at my disposal in the
Singapore herbarium, it is, however, possible to make a few
generalizations.

A few Malayan species certainly have a wide distribu-
tion, notably the two that grow in open places, D.
polypodioides and D. esculentum. These both appear to
extend from N. India and S. China right through the
Malayan region to N. Guinea or beyond. D. polypodioides
has various synonyms, and it appears that specimens from
India and China need checking against the Java type.

D. accedens may have an even wider distribution if it
should be found indistinguishable from D. proliferum, of
which the type comes from the Mascarene Islands. A

Vol. XI. (1940).

76

similar case is that of D. silvaticum (type from Mauritius),
but this is a less well characterized species, and its true
Asiatic distribution is in doubt.

D. sorzogonense is another interesting case of a widely
distributed species. This has undoubtedly been confused
with D. speciosum and probably also with other species, but
I have seen quite typical specimens from the Philippines
and New Guinea. It is perhaps significant that this species
is found in lowlands and mountains alike in Malaya, and
is not confined to the most moist and shady places. D.
speciosum, on the other hand, is certainly more restricted
in distribution, and in Malaya is only found in the
mountains, the typical form only (hitherto) on G. Tahan
on the east side of the Peninsula.

D. montanum is a case of more restricted distribution.
Our ferns which I now so name were formerly called D.
palidum. The latter species, however, as found on the
mountains of Java and Sumatra, is clearly distinct, and
I find that our ferns agree with the type of D. montanum
from Sumatra. D. montanum appears to occur at lower
altitudes in east and south Sumatra; 2.e., those parts
adjacent to Malaya.

D. heterophlebium is our only case in this genus of a
species finding the southern limit of its distribution in
Malaya.

So far as our present knowledge goes, the following
Species appear to be confined to the Malay Peninsula: D.
insigne, D. subintegrum, D. velutinum, D. procumbens, D.
Kunstleri, D. simpliciventum. All these are ferns of the
moist, shady mountain valleys.

D. insigne is a remarkable fern of the alliance of D.
accedens but very large. D. subintegrum is allied to D.
pallidum but in my opinion quite distinct. D. velutinum
is probably the most local of all, a derivative of D.
tomentosum. D. procumbens, D. Kunstleri and D. simpli-
civenium are distinct members of the very large bipinnate
group, the Peninsular species of which have been mostly
hitherto lumped under the names D. dilatatum or D.
latifolium.

Characters of the Genus

There has been much discussion on the status of the
genus Diplazium and on its relationship to Athyrium. I
do not feel that I can offer any new evidence, more especially
as the typical species of Athyrium (in the stricter sense)
occur outside Malaya. In view of Christensen’s remarks
in this Bulletin, Vol. 7, p. 268, and in view of the fact that
in Malaya Diplazium and Athyrium are quite distinct, I
here maintain the genus Diplazium, as in Christensen’s
Index Filicum. 7

Gardens Bulletin, S.S.

ref

The principal characters of the genus may be sum-
marized as follows :—

Stock usually short, erect, sometimes forming a short
trunk, in a few cases creeping, but (in Malayan species)
the fronds always tufted; roots rather coarse, black, some-
times forming a very close mat on the surface of the ground.
Scales on young fronds orbicular to linear, edges entire or
toothed, the cells usually elongated, walls of rather uniform
thickness, colour throughout medium to dark brown (never
clathrate), the teeth where present always formed of two
adjacent cells, the ends of which may separate; edges of
scales also provided sometimes with short glandular hairs.
Stipes relatively stout, sometimes bearing papillae or spines
near the base or throughout, these prominences being the
bases of fallen scales. Rachises glabrescent, minutely scaly
or in a few species hairy on the rounded lower side, papillate
in the grooved upper side. Midribs of pinnae and pinnules
grooved above, the edges of the groove often strongly winged,
the wing often interrupted and/or extended at the junction
of the midrib of a pinnule or the main vein of a lobe.
Lamina simple, pinnate or bipinnate, texture mostly thin,
pinnae and pinnules of bipinnate fronds with subequal base,
the venation never anadromous. Veins in most species free,
in a few species adjacent vein-groups anastomosing more or
less freely. Sori elongated along the veins, with lateral
indusium; the lowest acroscopic sorus in any group usually
double (diplazioid), and occasionally other sori double also;
double sori usually with two quite separate indusia, never
grading into a horse-shoe shape with indusium continuous
all round.

As compared with the above description, most species

of Athyrium have an anadromous venation and division,

A mm
° om 2” Macon

Fig. 1. Scales from base of stipe. On left, D. crenatoserratum,
outline of single scale and apex enlarged. Centre,
D, silvaticum, same, showing forked teeth. On right, D.
accedens, middle part of scale; shaded cells at edges
are black.

Vol. XI. (1940).

78

the pinnules often with very unequal bases, the scales are
always relatively broad and entire, usually rather pale and
thin, and the sori are always short and broad, the diplazioid
ones often of a horse-shoe shape, the two sides of the sorus
being connected.

In the character-of grooved upper surface of coste
and costules and the winged margins of the groove, both
Diplazium and Athyrium agree with Eu-Dryopteris and
Polystichum rather than with Thelypteris; and the teeth
on the scales on certain species of Diplazium are similar
in structure to the teeth of Dryopteris and Polystichum.
No scales in Athyrium or Diplazium (so far as I have
observed) have hairs on the margin as found in many
species of Thelypteris and Cyclosorus.

The markedly anadromous leaf-architecture found in
many species of Athyrium is matched in the group of species
formerly calied Polystichum but now separated by Ching
as Rumohra. Such leaf-architecture does not occur in
Thelypteris.

It appears therefore that Athyrium and Diplazium are
more nearly related to Eu-Dryopteris than to -Thelypteris.

Though the leaf-architecture in many species of
Athyrium is anadromous, there are clearly catadromous
species (e.g. A. silvaticum) commonly referred to that
genus. It seems possible that the type of leaf-architecture
may afford a useful character in subdividing the genus.

The true Athyriums in the Malayan region are all
species of rather high mountains (which is probably the
reason why they are little represented in Malaya, which
has much less high land than the great Sunda Islands),
whereas species of Diplazium, as noted above, are charac-
teristic of lowlands and medium elevations in the mountains.

Key to the Species of Diplazium of Malaya

N.B.—This Key applies only to memhaisitks: plants.
Fronds simple.
Fronds to 4 cm. wide, narrow at

base Dy .. D. subserratum
Fronds more than 4 cm. wide,
broadly cordate at base .. D. cordifolium

Fronds simply pinnate.
Scales entire, not toothed
Veins anastomosing copiously
veins anastomosing near costa D. heterophlebium
veins anastomosing only in
marginal third of lamina .. D. cordifolium

Gardens Bulletin, S.S.

79

Veins free or only slightly anasto-
mosing
fronds to about 15 cm. wide, |
costz hairy beneath .. D. tomentosum
fronds usually wider, costz not
hairy beneath
pinne cut half way or
more to costa
sori impressed, rachis
fibrillose, veinlets to
10 pairs .. D. sorzogonense
sori not impressed,
rachis not fibrillose,
veinlets to 7 pairs D. malaccense
pinns cut less than half
‘ way to costa
pinne to 5 cm. or
more wide
veins at 45° to
costa, pinne to 12
pairs D. xiphophyllum
veins at more than
60° to costa, pin-
heto5pairs .. D. riparium
pinne under 5 cm.
wide
lower veinlet of
group simple or
once forked .. D. montanum
lower veinlet more
copiously branch-
ed
2 veins of each
group sori-
ferous .. D. subintegrum
4-6 veins of
each group
soriferous D. Prescottianum
Scales toothed
Veins much anastomosing .. D. accedens
Veins free or only occasionally
joining near margin
pinne subentire, more than
2 cm. wide D. bantamense
pinne lobed, or if entire, less
than 2 cm. wide
pinne lobed 34 or more
towards costa .. D. speciosum

Vol. XI. (1940).

80

pinne lobed less than 34,
towards costa
pinne thin in tex-
ture, not auricu-
late below, scales
strongly toothed
pinne thicker in
texture, often
auriculate below,
scales irregularly
and “shordgly
toothed
Fronds bipinnate.
Veins anastomosing
stipes smooth, pinnules usually
less than 2 cm. wide, lobed
about 14 to costa .. ee
stipes spiny, pinnules always
more than 2 cm. wide, toothed
only Ne
Veins free
scales + round, peltate, entire
sori very short, basal on veins
sori longer, medial on veins ..
scales much longer than broad
frond to about 30 cm. long
frond much longer
lobes of pinnules 3—4 mm.
wide, cut more than 34, to
midrib
lobes of pinnules 7 mm. or
more wide, cut less than 34,
to midrib
rhizome procumbent,
fronds to 80 cm. long
rhizome erect, fronds to
more than 100 cm.
long
lower pinnules
sessile, to c. 2 cm.
wide, veins to 5
pairs, all simple
lower pinnules on
stalks to 5 mm.
long, to more
than 2 cm. wide,
veins 5-8 pairs,
some forking

Gardens Bulletin, S.S.

So

D,

. sulvaticum

. crenatoserratum

. esculentum
. nsigne

. latisquamatum
. Kunstleri

. velutinum

. polypodioides

. procumbens

. simplicivenium

dilatatum

81

1. Diplazium accedens Bl. Enum. 192. 1828.

Stock stout erect, sometimes a short trunk. Stipes
stout, green, covered with soft green protuberances near
base, to c. 60 cm. long; when young covered with narrow
scales, each on the end of a protuberance, the scales to 1:5
mm. wide at base, dull brown, thick in texture, with narrow
black toothed edge (see fig. 1). Lamina simply pinnate,
to 120 cm. or more long and 50 cm. wide, pinne about
15-jugate; apex of frond like the lateral pinne or with 1
or 2 deep lobes at the base; rachises and coste glabrescent,
smooth, buds common in the axils of the pinne. Lower
pinne on stalks to 8 mm. long, upper sessile, gradually
smaller. Largest pinnxe to 40 X 8 cm., base truncate or
very broadly cuneate, sides parallel for 2/3 or more of the
length of the pinne, then tapering to acute apex, edges
entire or very shallowly lobed, the lobes slightly toothed.
Main lateral veins in the pinne about 8-12 mm. apart on
the costa, at about 60° to the costa, each bearing up to 10
pairs of lateral veinlets, the lateral veinlets of adjacent
groups anastomosing when they meet and so producing a
series of parallelogram-shaped areolx; anastomosis of small
veins near edge of lamina irregular. Sori on some or all
the veinlets, often throughout their length.

A fern of shady mountain stream-sides found in Malaya
only on the Main Range and Taiping Hills. Sometimes in
very large fronds there are extra areoles between the normal
groups adjacent to the main lateral veins; a specimen having
this character formed the basis of Athyrium Ridleyi
Copeland, which I include in the present species.

This species is closely related to D. proliferum (am.)
Thouars, and the two should perhaps be united; if so, the
Species in this broader sense would extend from the
Mascarene Islands to the Pacific. A study based on
material from all parts of this wide range is required to
decide on the status of D. accedens and D. proliferum.

As regards the name D. accedens, the situation is a
little complicated. It appears that the three names D.
accedens, D. repandum and D. Swartzti, all published in
the same book by Blume, are to be regarded as synonymous.
Later authors have in many cases regarded all as synonyms
of Lamarck’s earlier name, or of Asplenium decussatum
Sw.; the names Swartzi and accedens have been taken up
and used in the genera Callipteris, Asplenium and Athyrium,
but the name D. repandum appears to have been almost or
entirely ignored. Backer and Posthumus (1939), however,
have revived the name D. repandum, apparently on grounds
of page priority (though they do not state this) which is
not admitted by the Rules. In their synonymy, however,

Vol. XI. (1940).

82

they include Diplazium proliferum Thouars, an older name;
their use of the name D. repandum is therefore contrary
to the rules.

It appears to me that, if the Malaysian fern is to be
regarded as specifically distinct from that of the Mascarene
Islands, the name Diplaziwm accedens Bl. is the appropriate
one to use, as this name has already been used by
Christensen in the 3rd Supplement to his Index Filicum
(1934), and had previously been used in a varietal sense
for Java ferns by van Alderwerelt van Rosenburgh.

2. Diplazium bantamense Bl. Enum. 191. 1828.

Stock short suberect. Stipes to 60 cm. long, glabrescent,
clothed at the base with brown scales about 1 cm. long and
15 mm. wide at base, edges finely toothed. Frond simply
pinnate, to about 60 cm. long and 25 cm. wide, pinnz up
to about 8 pairs, terminal pinne similar to the others or
occasionally with a large lobe at the base; a bud often
present on the rachis at the base of one of the upper pinne.
Basal pinne on stalks 3 mm. long, base slightly unequal,
on the lower side rounded and on the upper side cuneate
at 45° and decurrent; bases of the upper pinne more equally
rounded. Sub-basal pinnz largest, up to about 20 by 7
cm., commonly about 12 by 4:5 em., widest a little above
the base, then narrowed gradually to the acuminate apex;
edges entire or toothed only near the apex, texture firm;
veins usually distinct on both surfaces but sometimes
obscure, at rather more than 45° to the costa, forked at
or near the costa, the upper branch simple and often
soriferous almost throughout its length (diplazioid sorus),
the lower branch forked again 1-3 times, 1—3 of the branches
often bearing short simple sori towards the margin;
occasionally a slight anastomosis of veins towards the
margin; indusia narrow, persistent. |

Young plants have fronds with simple lamina to about
15 by 5 cm., deeply cordate or almost sagittate at the base,
later fronds having one or more pairs of lateral pinnez below
the terminal lamina. This behaviour is closely similar to
that of D. cordifolium; but D. cordifolium develops a much
larger simple lamina, which usually becomes fertile, before
producing fronds with free lateral pinne.

The name D. bantamense has in the past been given
to all Peninsular specimens of Diplazium with free veins
and rather broad subentire pinnze. These I now separate
into D. bantamense with toothed scales, D. ziphophyllum
and D. riparium with entire scales. I find that D.
bantamense as above described agrees well with Java
specimens, and hence, though I have not seen the type,
I feel confident that I have selected the correct Peninsula

Gardens Bulletin, SS.

83

fern to bear this name. It is a fern of moist shady forest,
chiefly in the hills, sometimes near streams in lowland forest.
In the shape of its pinne, it most nearly resembles the
pinnate form of D. cordifolium, but the much anastomosing
veins of the latter distinguish it.
Specimens in Singapore Herbarium :—
TRENGGANU. Kuala Berang, on _ streambank,
Holttum s.n., 14.5.1925.
PAHANG. Fraser’s Hill, 4,000 ft. S.F.N. 10018
(Holttum & Burkill) ; S.F.N. 11473 (Holttum) ;
S.F.N. 36505 (Holttum). Cameron Highlands,
Robinson Falls, 4500 ft., S.F.N. 17978 (Hender-
son); 4,800 ft., S.F.N., 23340 (Holttum).
Sungei Teku, 500 ft., S.F.N. 20554 (Holttum).
PERAK. Waterloo, Curtis 1361 p.p. (narrow toothed
pinne, and apical pinna lobed; a doubtful
specimen). Ulu Temango, Ridley 14229.
Taiping, 300-500 ft., King’s Collector 8388.
Upper Perak, 300 ft., Wray 3616.
NEGRI SEMBILAN. Ulu’ Bendul, S.F.N. 9876

(Holttum).

SELANGOR, near the Gap, 2,700 ft., S.F.N. 37159
(Addison).

PENANG, near Tiger Hill, 2,500 ft., S.F.N. 19346
(Holttum).

3. D. cordifolium Bl. Enum. 190. 1828.

Stock short erect. Stipes to 50 cm. or more long,
clothed at the base with brown scales about 1:0 cm. long and
15 mm. wide at base, edges of scales entire. Fronds of
two kinds, either simple with cordate base, or pinnate with
ovate entire pinne; transition forms also occur, but fully-
grown plants may bear simple fronds. Lamina of simple
fronds to 35 cm. long and 12 cm. wide, the base cordate
(often with a bud on the costa close to the base) the apex
more or less acuminate, the margin entire; texture thin but
firm, veins clearly visible on both surfaces; veins at more
than 60° to the midrib, forked close to the midrib, the lower
branch forked again, lateral members of the groups so
formed anastomosing about half-way towards the margin,
the outer part of the lamina containing a series of narrow
areoles about 1 mm. wide and 1 cm. or more long. Sori on
the outer veins of each group and also on some of the inner
veins, anastomosing with the veins, and reaching almost to
the edge of the lamina; indusia narrow, persistent. Pinnate
fronds with 1-8 pairs of pinnz below the terminal lamina,
which is smaller as the number of lateral pinne increases.
Lateral pinne sessile or the lowest very shortly stalked,
varying much in size, the basal pinne largest, upper ones

Vol. XI. (1940).

84

gradually smaller; base of pinnz subequally rounded,
margins quite entire, apex shortly acuminate; largest pinne
usually 12-16 cm. long and from 2-5 to 6-5 cm. wide; veins
as in the simple lamina above-described, but more oblique
to the midrib of the pinne (at about 45°).

This species is frequent in both lowland and mountain
forest in Malaya but is never found in the open. In the
lowlands it rarely (if ever) develops beyond the stage of
simple fronds, and in the mountains it appears that some
plants remain for a considerable time at this stage, which
is usually fully fertile. It is common, however, in mountain
forest to find plants which have one or more pairs of lateral
pinne below the cordate terminal lamina, and all stages of
transition can be found up to fronds with 7 or 8 pairs of
lateral pinnze. Specimens of the latter have usually been
named D. fraxinifolium Pr. or D. lineolatum Bl., simple
fronds being called D. cordifolium.

Backer and Posthumus (Varenflora voor Java, p. 123),
recognizing that both simple and pinnate forms belong
to one species, use the name D. fraxinifolium Presl. Presl’s
name was published prior to D. cordifolium, but the
description is very inadequate (anastomosis of veins is not
even mentioned) and I think it better to use Blume’s name,
which is well established. Owing to the polymorphism of
this species, and to the superficial resemblance of pinnate
forms to D. bantamense Bl., there has been much confusion
of names.

The parallel between D. bantamense and D. cordifolium
is a close one. Both ferns have at first simple fronds, and
later one or more pairs of lateral pinne which are broad
and sub-entire. There are, however, several clear differ-
ences. The scales are toothed in D. bantamense and entire
in D. cordifolium; the veins are almost always quite free
in D. bantamense but copiously anastomosing in D. cordi-
folium; and the initial stage of simple lamina is much more
developed in D. cordifolium, being often fertile, whereas the
simple lamina of D. bantamense is only found in young
plants and is rarely fertile.

I think it unnecessary to cite the many specimens of
this species which have been obtained in all parts of Malaya.

4. D. crenatoserratum (Bl.) Moore, Index 121. 1859.
Asplenium Bl. Enum. 177. 1828.

Stock short, erect. Scales on young fronds and bases
of stipes to about 4 by 1:5 mm., acuminate, edges irregularly
and shortly toothed, teeth not forked (see fig. 1). Stipes
to about 30 cm. long. Fronds simply pinnate, commonly
about 30 cm. by 15 cm., sometimes larger, pinne numerous,
close, the upper adnate, grading into the lobed deltoid apical

Gardens Bulletin, SS.

85

lamina, the lower longest, on stalks to 5 mm. long, commonly
about 15 by 2 cm., sometimes larger, texture firm. Pinnz
of fronds on young plants with bluntly rounded apex, edges
almost entire, base truncate and slightly auricled above.
Pinnz on large plants acuminate, edges lobed 14-14 way
towards costa, base strongly auricled above and less so
below; lobes close, usually irregular in length; veins pinnate
in the lobes, to about 4 pairs, often all fertile, the sori from
the costa almost to the edges, the sorus on the lowest
acroscopic vein diplazioid, the others not.

This is a very common fern of lowland forest, varying
much in size and in the degree of lobing of its pinne.
Young (or small) plants with blunt entire pinne may be
fertile. The largest plants approach D. silvaticum, but
differ in the points mentioned under that species. It
approaches also D. tomentosum, some specimens having
many short hairs, like those of D. tomentosum, on the base
of pinna-stalks beneath, and to much less extent on the
main rachis; but D. crenatoserratum is never so tomentose
as typical D. tomentosum. Typical specimens of the two
species are very different, but some are almost intermediate;
distinguishing characters are given under D. tomentosum.

5. Diplazium dilatatum Bl. Enum. 194. 1828.

A very large fern, with stout erect stock. Stipes 8
mm. or more in thickness near base, slightly rough with
raised bases of fallen scales; scales to about 1:5 cm. long,
hardly 1 mm. wide at base, brown with black toothed edges.
Lamina bipinnate-tripinnatified, pinne to 75 em. long and
27 cm. wide, pinnules to about 12—jugate below deltoid
lobed apex of pinna. Rachis glabrescent, costz with scattered
very narrow brown scales, especially near base and on stalks
of pinnules. Largest pinnules to 18 by 4 cm. (commonly
narrower) on stalks 5 mm. long, base truncate-subcordate,
apex acuminate, cut half-way to costa or a little more,
segments 8-10 -mm. wide at base, slightly oblique, edges
slightly toothed, apex rounded; smaller distal pinnules
sessile, lobed less deeply ; texture thin but firm, colour when
dry dark. Veins in each lobe pinnate with 5-8 pairs of
veinlets, lowest acroscopic veinlet often simple, the rest
usually forked (sometimes simple in fronds of small or
young plants), sori from near base along 14-2/3 length of
the vewmlets; sorus on basal acroscopic veinlet usually
diplozioid, rest not.

This very fine species has only been found in Malaya
at Cameron Highlands. The Highlands specimens closely
match a fern from Java (G. Ardjcene, 1800 m. Posthumus
3939) and also the Sumatran specimens quoted below. I

Vol. XI. (1940).

86

think they are quite distinct from the more numerous
specimens, mostly from lower elevations, referred in this
paper to D. simplicivenium.
Specimens in Singapore Herbarium:
PAHANG. Cameron Highlands, 4,600 ft., S.F.N.
23620 (Henderson) ; S. Terla, 3,900 ft., S.F.N
31334 (Holttum), young plant.
SIAM. Doi Chiang Dao, Eryl Smith 1198.
SUMATRA. Brastagi, 5,000 ft. S.F.N. 15403 (Holt-
tum). Dolok Singgalang, N. v. h. Tobameer
1650 m., Lorzing 8864.
Copeland Pterid. Philip. Exsicc. 178 (Mt. Matutum,
April 1917) seems to me closely similar to the above
specimens and I would refer it to the same species.

6. Diplazium esculentum (Retz.) Sw. Schrad. Journ. 1801,
- : 312. 18038. Hemionitis esculenta Retz. Obs. Bot.
. 1791.

Stock erect. Stipes to 50 cm. or more long, glabrescent
except for a few brown scales at base; scales about 1 cm.
long and 1 mm. wide, the edges finely toothed. Lamina
bipinnate, large, often ‘1 m. or more long and about half as
wide. Pinnz to 50 cm. or more long, bearing numerous
pinnules which are often rather distant from each other.
Pinnules varying much in size and cutting, usually less than
2 cm. wide, the lowest stalked 2 mm., the rest sessile,
truncate or broadly cuneate and usually more or less
auricled on one or both sides at the base, edges crenate or
lobed up to about 14 distance to costa, the lobes or crenations
slightly toothed. Veins in pinnate groups in the lobes,
about 8-10 pairs of side veins, the lower 2-3 pairs of
adjacent groups anastomosing, forming an irregular inter-
mediate excurrent vein leading towards a sinus between
adjacent lobes. Texture thin; rachises and coste glabres-
cent, or occasionally pubescent beneath with numerous pale
brown hairs with dark cross-walls, and scattered or
abundant small ovate toothed scales on costz. Sori occupy-
ing almost the whole length of the veins, often also on part
of the joint excurrent vein, often more than one diplazioid
sorus in each group. |

This is a common fern of wet ground in open places
in the lowlands of Malaya, and is well known as an edible
plant. It never occurs in shady forest. It is a straggling
and untidy species, lacking the beauty of most of the larger
species of this genus.

The excurrent vein formed by union of the lateral veins
of adjacent groups does not run to the sinus, stopping there
in a thickened single unit, as in Cyclosorus (Dryopteris
p-p.), but divides just below the sinus, part running close

Gardens Bulletin, S.S.

87

to the edge of the lamina on either side above the sinus.
Thus, though the venation of D. esculentum is superficially
similar to that of species of Cyclosorus, it is not identical
and doubtless had a quite separate origin.

7. Diplazium heterophlebium (Mett.) Diels Nat. Pfl. 1, 4:
228. 1899. Asplenium heterophlebium Mett.; Baker,
Syn. Fil. 248. 1867.

Stock short erect. Stipes slender, to 30 cm. long,
glabrescent, when young clothed at the base with thin entire
brown scales to about 1 cm. long and 1:5 mm. wide at the
base. Lamina to at least 35 cm. long and 25 cm. wide, the
apical part (in young plants the whole lamina) deeply lobed,
and below this a few pairs of adnate or sessile pinnz. Lobes
of apical lamina falcate acuminate entire, to about 15 cm.
long and 4 cm. wide; pinnz to about the same size, their
edges entire or more or less deeply undulate. Main lateral
veins in the pinne pinnate, the veinlets of adjacent groups
anastomosing rather irregularly from about 14 distance -
from costa to margin, the areole smaller towards the
margin, with free short veins with thickened ends just
within the margin. Sori of small fronds few, often one to
each vein-group, on the basal acroscopic vein, often all simple
but sometimes diplazioid; on larger fronds often several
sori to each vein-group, some of them anastomosing with
the veins. Texture thin, rachis and coste glabrous or
slightly scaly beneath with small narrow entire scales.

In Malaya, this species has only been found in the
valley of the Bertam River, below Cameron Highlands, at
3,000—3,500 ft. altitude. The specimens there gathered
differ from Himalayan specimens in having broader fewer
pinne, all (or almost all) adnate to the rachis, with ampler
venation and more irregular anastomosis. It is possible
that fully grown plants have not yet been noticed. Further
collecting in the valleys in the north of the main range will
no doubt reveal more localities for this and other ferns of
northern distribution which find their southern limit in
Malaya. The thin pale entire scales show that this species
is not closely related to D. accedens, though the anastomos-
ing venation might lead one to think them allied.

8. Diplazium insigne Holttum, Gard. Bull., S.S., 9: 123.
1937.

Stock stout, short, erect. Stipes stout, c. 1 m. long,
spiny towards the base, the spines 2 mm. long, each at first
bearing a scale; scales dull brown 1:5 cm. long’ by 1:5 cm.
wide at the base, with a narrow black toothed edge,
deciduous. Lamina to 1:5 m. long, bipinnate. Lowest
pinne about 28 cm. long and 6-5 cm. wide, narrowed and

Vol. XI. (1940).

88

stalked at the base, the margins lobed half-way to the costa,
the apex acuminate. Middle pinnzx largest, to 60 cm. long
and 16 cm. wide, pinnate; pinnules slightly oblique, adnate
to the rachis (the lowest ones narrowly, the upper fully
adnate and grading into the lobed apical lamina of the
pinna), to 9 cm. long and 2:3 cm. wide, cuneate at the base
at an angle of about 45° on each side, narrowed gradually
from the base and then suddenly at 1-5-2 cm. from the apex,
margins slightly serrate; texture firmly herbaceous; veins
anastomosing as in D. accedens. Sori few or copious,
the lowest acroscopic sorus in each vein-group usually
diplazioid, and occasionally a few others also.

This species is evidently closely allied to D. accedens,
and the apex of a frond might pass for that species, but
its copious bipinnate form marks it as a quite distinct
species. It is evidently similar to D. Smithianum from
Ceylon, but appears to be much larger; probably both D.
Smthianum and D. insigne are local derivatives of D.
accedens (or D. proliferum).

Since the original description of D. insigne was
published, I have discovered a collection made by Curtis on
Taiping Hills at 2,000 ft. (No. 13852), labelled D. Smith-
ianum, which is clearly D. insigne. Otherwise no new
locality has yet been found for D. insigne except the type
locality in a valley below Fraser’s Hill (Pahang side).

9. Diplazium Kunstleri Holttum sp. nov. Fig. 2.

Rhizoma repens? Stipites ad 1 m. longi vel ultra, in
sicco olivaceo-straminei, non muricati, basin versus
squamis rotundatis peltatis deciduis muniti. Rachides
leeves, pallide, plerumque ad basin pinnarum superiorum
prolifere. Lamina bipinnata, ad 120 cm. longa et 80 cm.
lata vel ultra; pinne fere 12—jugate, infime leviter reducte,
superiores in lobos apicis laminze sensim mutate. Pinne
maxime c. 65 cm. longz, c. 25 cm. late, in stipitibus 9 cm.
longis stantes (stipites pinnarum superiorum sensim
breviores). Pinnule infime inter se 5 cm. distantes;
pinnule maxime 16 cm. longe et 3:55 cm. late, stipitate
3 mm. vel ultra (ad 8 mm.), basi late cuneate vel
truncate, apici acuminate dentate, cetera margine lobatze
2/3 costam versus; lobi leviter obliqui, 5-7 mm. lati, basi
dilatati, margine leviter serrati, apice rotundati; costule
falcatee, squamis parvis latis peltatis paucis munite;
venule in uno lobo 7—9-jugate, oblique, plerumque furcate
(vel in pinnulis parvis simplices), utrinque distincte sed
haud prominentes; lamina textura quam species affines
crassior. Sori medium venularum solum_ occupantes;
indusia angusta, in soris maturis haud evidentia.

TYPUS: Ginting Simpah, Selangor, 2,500 ft. alt., S.F.N.
31194, leg. Holttum. |

Gardens Bulletin, SS.

89

Fig. 2. Diplazium kunstleri, type. On left, apex of frond; on right,
above, single pinnule from middle pinna; below, two
fertile lobes.

D. Kunstleri is very distinct among the Malayan
bipinnate species of Diplazium in the combination of the
following characters: round peltate scales (otherwise found
only in D. latisquamatum) ; pinnules cut more than half-way
to the costa, with broad lobes; medial sori which never
reach base or apex of the veins; and proliferous rachis. I
ae the type growing in a rocky streambed in shady

orest.

The species is named after Hermann Kunstler (‘‘King’s
Collector” of herbarium labels), who collected many
interesting Malayan plants for the first time. He collected
D. Kunstleri “near little stream, dense jungle’, near Ulu
Kerling, under No. 8797. Other collections in the Singapore
herbarium are: 15th mile, Pahang Track, Ridley 8652;
Rawang, Ridley 7848, 7852. Thus all specimens have been
found on the Main Range in Selangor, at medium altitudes.

Vol. XI. (1940).

90

10. Diplazium latisquamatum Holtt. Gard. Bull., S.S., 9:
124.1937. . Fie. 2

Stock short erect. Stipes to 90 cm. long, throughout
or towards base only papillate with rather soft green
papille, which in young fronds bear round or ovate, peltate,
entire, nearly black scales, to about 5 mm. long and 2 mm.
wide; old stipes and rachises bearing scattered small
rounded scales only. Lamina to about 1:5 m. long and 1 m.
wide, bipinnate-tripinnatifid. Largest pinne to about 60
cm. long and 23 cm. wide, shortly stalked, the basal pinnules
slightly reduced, smaller on the acroscopic than on the
basiscopic side. Pinnules sessile (or the basal ones on stalks
to 2 mm. long), the largest about 12 K 2:5 cm. (commonly
about 9 X 1:5 cm.), base broadly cuneate or subtruncate,
gradually narrowed to acuminate toothed apex, margins
lobed to within 2-3 mm. of the costa, lobes slightly oblique,
mostly about 4 mm. wide, apex truncate, margins entire
or slightly toothed near apex, texture thin but firm, coste
bearing scattered small ovate or rounded scales; veins
pinnate in the lobes, the veinlets 5—6—jugate, simple or
forked. Sori occupying only the lower half or less of the
veins, the basal acroscopic sorus usually diplazioid, the
others simple, the indusia thin, broad, persistent.

Fig. 3. Diplazium latisquamatum. Above, single pinnule from middle
pinna. Below, two fertile lobes; the dotted outlines on
the left hand lobe show the position of sori removed.

Gardens Bulletin, S.S.

91

This species is very clearly distinguished from all
others in Malaya by its almost round peltate entire scales
and very short sori confined to the basal half of the veins.
In general appearance, it resembles D. procumbens, but the
scales of the latter, and its prostrate stem, distinguish it,
even in sterile plants.

D. latisquamatum has been found in Malaya in shady
forest by streams in the valleys of the Main Range at
4,000-5,000 ft. at Cameron Highlands and Fraser’s Hill
only. I have collected closely similar specimens in Sumatra,
at 6,000—7,000 ft. altitude on Korinchi Peak. I am, however,
in some doubt whether the Kinabalu (British North Borneo)
specimens cited with the original description of the species
should be regarded as identical; some have decidedly longer
and more pointed scales, and pinne and pinnules broader.

The additional specimens from Malaya and Sumatra
to be added to the list published in 1937 are:—

PAHANG. Fraser’s Hill, 4,000 ft. S.F.N. 11007 (Md.
Nur) ; S.F.N. 36504 (Holttum).

SUMATRA. Korinchi Peak 6,000-—7,000 ft., Holttum
s.n. 18/2/1933.

11. Diplazium malaccense Presl. Epim. Bot.: 86, 1849

Stock short erect. Stipes to c. 50 cm. long, clothed at
base with entire brown scales which are 10-15 mm. long
and 1-5-2 mm. wide at base, rest of stipe and rachis glabres-
cent, not fibrillose. Lamina pinnate, to about 60 by 30 cm.
Pinne to about 20—jugate, below lobed apex of the lamina.
Lower pinne on stalks 5 mm. long, bases very unequal,
lower base often much cut away, upper truncate; upper
pinne with base subequally truncate or broadly cuneate.
Pinne thin in texture, drying rather light green, commonly
to 16 cm. long and 2 cm. wide, apex acuminate, lobed not
more than 2/3 way to the costa, lobes about 6 mm. wide at
the base, a little more oblique than in D. sorzogonense, ends
rounded to subtruncate and slightly toothed; veins up to 7
pairs in each lobe, all simple; sori from near base of veins
to near margin of lamina, the acroscopic basal sorus usually
but not always diplazioid. Surface not depressed below
sori; indusium medium brown, thin but firm, usually turned
back but not rolled as in D. sorzogonense.

A species of lowland and midmountain forest, found
throughout Malaya, differing from D. sorzogonense (which
grows in similar localities) in glabrescent stipe and rachis,
less deeply lobed pinnz, and sori not at all impressed.
Specimens have been distributed under the names D.
silvaticum and D. speciosum, both of which differ in having
toothed scales.

Vol. XI. (1940).

92

A very large, perhaps monstrous, specimen was
gathered by Mohamed Nur on Gunong Angsi (S.F.N.
11563). The apical part of the frond is quite like normal
D. malaccense, but the middle and lower pinne are very
large (to 8 cm. wide) and deeply lobed (almost to the costa),
the lobes being falcate and acutely pointed, with forked
lateral veinlets up to about 20 pairs. The bases of these
pinne have 3 or 4 pairs of normal segments below the larger
pointed segments.

12. Diplazium montanum y.A.v.R. Bull. Buitenz. II Ser.,
28: 19. 1918.

Stock short erect. Scales on bases of stipes narrow,
black, entire, to about 10 by 1 mm. Stipes black at the
base, to about 60 cm. long. Frond simply pinnate. Pinnz
to about 20—-jugate, to about 18 by 2 cm., the lower ones on
stalks to 4 mm. long, upper sessile, the apical pinna like the
others (occasionally with one or two + separate round
lobes at the base) ; upper pinnez unequally cuneate at the
base; lower pinne with upper base narrowly rounded, lower
base narrowly cuneate; edges of pinne toothed, apex
acuminate toothed, texture thin; veins forked at or near
base, the acroscopic branch soriferous, simple, the lower
branch simple or sometimes forked, occasionally bearing a
short sorus. Sori on the acroscopic veinlet mostly simple,
but occasionally diplazioid, reaching from the costa almost
to the edge of the lamina.

This species differs from D. pallidum Bl. in having the
apical pinna like the others (in D. pallidum the apical
lamina of the frond is deltoid and deeply lobed), and in
having the lower pinnze with much narrower bases (in
D. pallidum the lower pinne have a broadly truncate upper
base). It appears also that D. pallidum has fertile fronds
from a young stage of development, whereas D. montanum
does not produce fertile fronds until the plants are full
grown.

D. pallidum occurs in Sumatra on the higher parts of
the mountains, matching Java specimens exactly; D.
montanum appears to occur at lower levels in Sumatra,
thus falsifying its name. In the Peninsula D. pallidum
does not occur. D. montanum occurs in forest at low and
medium elevations.

Specimens in Singapore Herbarium :—

PAHANG. Tahan River, Ridley 2167. Tembeling,
S.F.N. 24703 (Holttum). Buloh Telang, P.
Tioman, 800 ft., S.F.N. 18583 (Henderson).

PERAK. 2,000-2,500 ft., King’s Collector 10959.
Sungei Siput, Haniff 4047. Ulu Temango,
Ridley 14209.

Gardens Bulletin, S.S.

93

SUNGEI UJONG. Hullett s.n. Aug. 1880.

SELANGOR. Batu Caves, Ridley 8139.

SUMATRA (East Coast). Region of Marban, Bilah, |
near Bilah Pertama, Rahmat si Toroes 145,221
(distr. Univ. of Michigan).

SIAM. Patani, 210-350 m., on rock in stream, Eryl
Smith 1948.

13. Diplazium polypodioides Bl. Enum. 194. 1828.
D. asperum Bl. Enum. 195. 1828.

Stock stout, erect, often subarborescent; stipes stout,
to 1 m. long, clothed thickly at the base with narrow scales
of thick texture, dull brown with narrow black toothed
edges, to about 4 cm. long and 2 mm. wide at base, usually
not more than 1 mm. wide over the greater part of their
length; the scales attached to protuberances on the stipe
which remain as more or less short prickles up to 1 mm. or
rather more in height. Lamina bipinnate, deeply tripinnati-
fid, commonly about 100 cm. long and 60 cm. wide, some-
times much larger, largest pinne to about 85 by 30 cm.,
commonly less, shortly stalked, with 20-30 pairs of pinnules
below the deltoid lobed apex of the pinna. Largest pinnules
to 16 by 3:5 cm., commonly about 9 by 2 cm. or sometimes
less, sessile or the lowest shortly stalked, base truncate or
very broadly cuneate, apex acuminate, edges cut into lobes
34, or more towards the costa; lobes slightly oblique, sub-
truncate or bluntly pointed, their margins sharply toothed
throughout, 2-5 mm. wide, veins 7-12 pairs in each lobe,
usually simple, forked in the largest lobes; texture thin to
firm. Rachises and coste usually glabrescent, the costze
sometimes with small narrow scales; rachises, like the
stipes, often slightly prickly with the bases of fallen scales.
Coste above distinctly winged, the wing interrupted and
sometimes dilated at the base of each costule. Sori usually
present on almost all veins, from the base half-way or more
to the edge, the indusia thin but persistent; on short narrow
lobes with close veins the sori may be almost confluent.

The long narrow black-edged toothed scales, the usually
prickly stipes, and the deeply cut narrow toothed lobes are
characteristic of this species.

SELANGOR. Klang Gates, F.D. 14602 (Ingram).
Ginting Simpah, 1,500 ft., Hume 8551B (Herb.
F.M.S. Mus.), S.F.N. 9982 (Burkill).

NEGRI SEMBILAN. G. Angsi 950 ft. S.F.N. 11515
(Md. Nur). Senaling Inas F.R., in small clear-
ing, S.F.N. 9770 (Holttum).

PENANG. N. of Western Hill, 2,000 ft., S.F.N. 19776
(Holttum). Penara Bukit, 2,000 ft., S.F.N.

Vol. XI. (1940).

94

reir (Holttum). No loc., abundant, C. Curtis
35.
PERAK. Ulu Temango, Ridley 14206. Ulu Bubong,
King’s Collector 10849.
PAHANG. Raub, S.F.N. 14103 (Best). Telom, Ridley
13975. Kuala Tahan, Ridley 2,400. S. Sat,
S.F.N. 22089 (Henderson). Robinson Falls, -
4,000 ft., S.F.N. 23262 (Holttum). Boh Planta-
tions, 4,000 ft., S.F.N. 329387 (Md. Nur). C.
Highlands, 5,000 ft., S.F.N. 23518 (Holttum)
S.F.N. 31295 (Holttum) ; 4,800 ft., S.F.N. 23361
(Holttum).

In Malaya, this species is common by the edge of forest
in moist ground, especially by streams, never in the heavy
shade of high forest. It is especially characteristic of the
mountains between 2,000 ft. and 4,000 ft., occurring
abundantly by roadsides, but occurs also in low country
and at higher elevations. In very moist slightly shaded
situations in mountain valleys it sometimes attains a very
large size, with broader and thinner leaflets than commonly,
but I cannot see any clear specific distinction between these
and the smaller leaves of firmer texture. The upper parts
of larger fronds are much like the smaller fronds.

I have not examined any Indian specimens, and I do
not know how far the specimens described by Beddome as
D. polypodioides and D. asperum would agree with Malayan
plants. D. polypodioides, as I understand it, is certainly
a very widely distributed species. I have seen specimens
which I would so name from N. Siam, Hainan, the
Philippines and New Guinea.

As regards the identity of Blume’s two species, D.
asperum and D. polypodioides, I follow both Raciborski and
the recent work of Backer and Posthumus in uniting them,
using the name polypodioides, though I think it is a pity
that the more appropriate name asperum was not chosen.
Van Alderwerelt van Rosenburgh suggested various differ-
ences between Blume’s two species, but I do not find that
these differences are constantly associated in Peninsula
specimens. As above noted, habitat appears to have a
considerable influence on size and other characters.

14. Diplazium Prescottianum (Wall.) Moore, Index 156.
1859. Aspidium Prescottianum Wallich, Catalogue

No. 235 (nomen) ; Hooker, Spec. Fil. 3: 251. 1860.

The type of this species was collected in Singapore, and
agrees with the other Singapore specimens quoted below.
These may be described as follows. Scales on bases of stipes
brown, entire, about 1:5 mm. wide at base. Pinne to about

Gardens Bulletin, S.S.

95

14—jugate below the lobed deltoid apex of the lamina, texture
firm. Lowest pinnze on stalks to 6 mm. long, upper sessile
or adnate. Lower pinne to 24 cm. long and 1:8 cm. wide,
base unequal, rounded above, rounded or cuneate below;
upper pinne with broadly cuneate auricled upper base, and
narrowly cuneate lower base. Hdges of pinne crenate or
slightly lobed, lobes slightly toothed or entire, the auricle
at the base of the upper pinnzx more or less separated by
a sinus from the rest of the pinna (in the type with a few
similar lobes following it, cut nearly to the costa). Ves
of largest pinnzx in pinnate groups, 3-4 veinlets on each side
of the main vein, 4-6 veins in each group soriferous, sorus
on basal acroscopic vein diplazioid, others not; sori from
near base of vein to near margin.

The other specimens differ in the width of the pinne,
which are up to 3:3 cm. wide, and some of them in the entire
lack of an auricle on the upper base of the upper pinne. All
agree in the character of the scales, the stalks of lower
pinne, the slight lobing or crenation of the edges of the
pinne, and the number of veins and sori in each group.
The species appears to be related to D. montanum v.A.v'R.
(and D. pallidum Bl.), but differs in larger brown scales,
more lobed edges to pinne, and many more soriferous veins
in each group. From D. subintegrum it differs in shorter
pinna-stalks, more lobed edges, more veins and sori in each
group.

Specimens in Singapore Herbarium :—

SINGAPORE. Between Bukit Panjang and Woodlands,
Matthew; Toas, Ridley s.n. 1892.

SELANGOR. Ginting Simpah, Hume, F.M.S. Mus.
9665, 9347, 8997, 9677.

PERAK. Kinta, 1,500-2,000 ft., Kunstler 7146.
Larut, 2,000-3,000 ft., Kunstler 2391. Kinta,
Curtis 3369.

SIAM. Patani, 540-700 m., Eryl Smith 1930; 260
m., Eryl Smith 1927. Koh Chang, Marcan 1268.

15. Diplazium procumbens Holttum sp. nov. Fig. 4.

Rhizoma procumbens, c. 1 cm. diametro. Stipites ad
80 cm. longi, in sicco sordides vel straminei, glabrescentes,
basibus squamarum deciduarum leviter asperuli, in juven-
tute squamis parvis lineari-lanceolatis, fuscis, margine
leviter dentatis (non nigris) vestiti. Lamina deltoidea,
bipinnata, ad c. 80 cm. longa et c. 80 cm. lata; pinne
maxime 50 cm. longe et 20 cm. late. Pinnule sessiles vel
infimis stitipate (stipites 3 mm. longi), maximze 2:5 cm.
late, basi truncate, apice acuminate, 3/4, ad costam lobate;
lobi maximi basi 7 mm. lati, oblongi, cum costa angulum
fere rectum formantes, apice rotundati, margine apicem

Vol. XI. (1940).

Fig. 4. Diplazium procumbens. On left, pinnules from different
fronds, and on right enlarged portions of each to show
veins and sori. Below, the procumbent rhizome.

versus leviter dentati. Venule ad 7—jugate, in lobi maximi
omnes furcate, in lobi parvi plerumque simplices, coste
squamulis paucis parvis lanceolatis brunneis dentatis
munite. Sori e costa in 144-2/3 longitudine venularum
sedentes, sorus infimus acroscopicus plerumque diplazio-
deus; indusia tenuia membranacea.

TYPUS: Fraser’s Hill, 4,000 ft, S.F.N. 365035
(Holttum).

This species resembles D. latisquamatum in appearance,
but differs in its narrow slightly toothed scales (round and
entire in D. latisquamatum), procumbent rhizome, hardly
muricate stipes, usually broader, more acuminate pinne,
more copiously forked veins and longer sori with very thin
indusia. It is abundant in the moist shady valleys, near
streams, on the Pahang side of the divide at Fraser’s Hill,
and has not yet been found elsewhere.

Gardens Bulletin, S.S.

97

Specimens in Singapore Herbarium :—

Fraser’s Hill, 4,000 ft., S.F.N. 8844 (Burkill &
Holttum), S.F.N. 11478 (Holttum), S.F.N..
21645 (Holttum), S.F.N. 37201 (Addison), s.n.
Corner 12.8.37.

16. Diplazium riparium Holttum sp. nov. Fig. 5.

Rhizoma breve erectum. Stipites ad 50 cm. longi, in
Vivo virides, in sicco brunnei fusci, basi squamis lanceolatis,
fere nigris, integris, 4-5 mm. longis, 1:5 mm. latis, vestiti,
cetera nudi. Lamina pinnata, pinne laterales ad 5—-jugate
(superiores nonnunquam leviter adnate), pinna terminalis

A
IN

Fig. 5. Diplazium riparium, type; outline of frond. On left, below,
detail of venation and sori, enlarged.

Vol. XI. (1940).

98

ceteris similis, rachis non prolifera. Pinne maxime ad 29
cm. longe et 7 cm. late, brevissime stipitate vel sessiles,
basi subeequaliter cuneate, margine fere integre, apicem
versus subito augustate et plus minusve caudate, cauda
ad 2 cm. longa; costa subtus rotundata, in sicco brunnea
vel fusca, nuda, supra in sicco canaliculata; vene angulum
plus quam 60° cum costa formantes, rarissime anastomos-
antes, in fasciculis parvis glomerate; fasciculus ex venis
tribus constatus, vena medialis furcata (sepe bis furcata) ;
sori 3—4 pro fasciculo, sori venarum lateralium e costa fere
ad marginem extensi, ceteri breviores, omnes prope
marginem terminati.

TypPpus: Selangor, Semenyih, Hume 8186 (Herb.
F.M.S. Mus.).

This species agrees with D. xiphophyllum, as I under-—
stand it, in Malaya, in its entire scales and rather large
subentire pinne. It differs from D. xiphophyllum in the
following characters: stipe and rachis darker when dry
(stramineous in D. xiphophyllum) ; scales shorter, darker ;
fronds much shorter; pinne to about 5 pairs only, shorter
and broader, almost or quite entire on old plants; anasto-
mosis of veins rare, veins at more than 60° to midrib in
mature plants, the ends of the veins uniting with a
cartilaginous margin; rachis not gemmiparous.

Young plants have the first fronds simple, later fronds
with one or more pairs of lateral pinne, which have
distinctly toothed margins, especially towards apex, fertile
from an early stage (on pinne from 9 X 2 cm.), the lateral
veins often making a smaller angle to the midrib than in
the larger pinne of mature plants.

This is a fern of the lowlands only, in forest near
streams, usually in wet places and often on stream banks.
It has also been found in the low country of E. Sumatra
and Sarawak. It has usually been referred to D.
bantamense, from which it differs in its entire scales and
lack of buds on the rachis, also in young plants not having
hastate or cordate fronds. It also resembles D. Donianum,
but this also (so far as specimens available indicate) has
toothed scales. It may be very near D. alternifolium from
dava, but appears to differ in more entire pinne and in
fewer veins in each group (D. alternifolium has usually
®2 pairs of lateral veins in each group, often all soriferous) ;
I have not seen scales of D. alternifolium.

Specimens in Singapore Herbarium :—

PERAK. Upper Perak, 300 ft., L. Wray Jr. 3590
(young), 38650. B.P.D. 300-500 ft., King’s
Collector 7894.

Gardens Bulletin, S.S.

99

PAHANG. Near K. Teku, 500 ft. (young plant),
S.F.N. 20792 (Holttum) ; S. Cheka, young plant,
S.F.N. 24780 (Holttum); Mentakab, West, .
S.F.N. 24558 (Holttum).

KELANTAN. Ulu S. Keteh, Md. Nur, s.n. 14.2.24.

TRENGGANU. K. Brang, S.F.N. 15323 (Holttum).

SELANGOR. Semenyih, Hume, F.M.S. Mus. No. 8186
(Type).

MALACCA. Ayer Keroh, Ridley s.n. Dec. 1899. Ayer

| Panas, Ridley 1587.

SUNGEI UJONG. No loc., Alvins 3282.

JOHORE. Ulu Sedili, Holttum, s.n., May 1932; foot
of G. Panti, 200 ft., S.F.N. 18301 (Holttum).

SINGAPORE. Hullett s.n. Jan. 1881.

SUMATRA. East Coast, general region of Marbau,
Rahmat si Toroes 222.

SARAWAK. No loc., Bishop Hose s.n., 1890.

17. Diplazium silvaticum (Bory) Sw. Syn. Fil. 92. 1806.
Callipteris silvatica Bory, Voy. 1: 282. 1804.

Stock short, erect. Scales on young fronds and bases
of stipes very dark, to about 10 mm. long and 1-5 mm. wide,
closely and distinctly toothed, teeth often forked (see fig. 1).
Stipes to about 40 cm. long. Frond simply pinnate, to
about 50 by 20 cm., pinne numerous, lower stalked to about
S mm., upper sessile; lowest pinne with upper base
subtruncate, lower base cuneate; upper pinnze with base
broadly cuneate to truncate, more or less auricled above but
not below; pinnz to about 12 cm. long and 3 cm. wide
(commonly less), apex acuminate, edges lobed to about 1,
distance to costa, texture thin; lobes oblique, about 5 mm.
wide, veins in pinnate groups with 3-4 pairs of lateral
veins; sori from near base of veins 34, or more towards
edge, sorus on acroscopic basal vein diplazioid, sometimes
sorus on next acroscopic vein also, other sori simple.

This species, which I believe I have correctly identified,
occurs throughout Malaya, but is not always easy to
distinguish from D. crenatoserratum. It differs from D.
crenatoserratum in having broader pinne of thinner
texture, not auriculate on the lower base, and by the larger
and closely and distinctly toothed scales, the teeth often
forked. Some specimens approach D. malaccanum in
appearance; they are distinguished by their toothed scales
(scales of D. malaccanum are entire).

D, silvaticum was originally described from Mauritius.
Malayan specimens, according to Christensen (Dansk Bot.
Ark. 7: 82. 1930), are very close to the type. This must
be reckoned one of the most widely distributed species of
the genus. Ceylon specimens in the Singapore herbarium

Vol. XI. (1940).

100

have scales like Peninsula plants, but have more ample
fronds than any Malayan specimens. How far eastwards
the species extends, I do not know. A Java specimen in
the Singapore herbarium (quoted below) agrees quite well
with Peninsula specimens.

Specimens in Singapore Herbarium :—

SINGAPORE. Hullett s.n. (2 sheets).

PAHANG. Tahan River, Ridley 5818. S. Cheka,
Holttum s.n., 26.4.31. Tembeling, S.F.N. 24704
(Holttum). Base K. Glanggi, S.F.N. 22422
(Henderson).

SELANGOR. Woods Batu Caves, Ridley s.n., Dec.
athe (?). Klang Gates, F.M.S. Mus., Hume

ris

PERAK. Upper Perak, 300 ft. Wray 3626, 3594.
Gapis Pass, Curtis 1360.

KELANTAN. Gua Panjang, Gua Ninik, S.F.N. 19498
(Henderson).

KEDAH. Langkawi, G. Raya, Curtis s.n., Sept. 1890.
48 mile Jeniang Road, S.F.N. 35995 (Kiah).

JAVA. Djember, inter G. Boto et Tjuramamis, 600
m., Backer & Posthumus s.n. 11.3.29.

SIAM. Chawng (7° 35’), S.F.N. 24382 (Kiah,
doubtful). Betong, Patani, 440 m., Eryl Smith
1929.

18. Diplazium simplicivenium Holttum sp. nov. Fig. 6.

Rhizoma breve erectum. Stipites 1 m. longi vel ultra,
basin versus basibus squamarum deciduarum asperuli;
squamez 2 cm. longe vel ultra, prope basin 1-2 mm. late,
fusce, margine dentatsze, dentes simplices. Lamina 1-5 m.
longa et 1 m. lata vel ultra, bipinnata. Pinnze adjacentes
c. 15 cm. inter se distantes, oblique; pinne maximz 65 cm.
longe et 20 cm. late; pinnule libere ad 20-jugatze, sub
apicem lobatum lamine; pinnule maxime 12 cm. longe,
2 cm. late, fere sessiles, basi truncate (non auriculate),
apicem versus acuminate, e margine 14, costam versus
lobate; lobi truncati, c. 7 mm. lati, leviter obliqui; vene
c. 5-jugate, omnes simplices; sori e basi venarum 3, vel
ultra marginem versus extensi; textura lamingz tenuis, vene
utrinque distinctissime; coste subtus (plerumque basin
versus) squamis parvis paucis brunneis, margine fere
nigris, munite. 7

TYPUS: Fraser’s Hill, 4,000 ft. alt. S.F.N. 36507, leg.
Holttum.

This species differs from D. dilatatum in the following
characters: wider and longer scales of coarser texture; veins
all simple (some may be forked on transition pinne near
apex of frond, but none on the typical larger pinne, even

Gardens Bulletin, S.S.

101

Fig. 6. Diplazium simplicivenium. Below, single pinnule from large
) pinna. Above, two lobes enlarged to show venation
and sori.

of very large fronds) ; usually not more than 5 pairs of
veins; sori occupying 34, or more of the length of the veins.
D. simplicivenium occurs in moist shady valleys near
streams; it has been found on the Main Range, the Taiping
Hills, and on G. Senyum in Pahang, also in Lower Siam.
Some specimens from Borneo agree in having simple veins,
but they have much wider pinnules than Peninsular
specimens and I doubt their identity with this species.
Specimens in Singapore Herbarium :—
PERAK. No loc. Hullett s.n., 1881. Larut, 1,000-
1,500 ft., King’s Collector 2346; 1,500—2,000 ft.
King’s Collector 2214. Temango, Ridley 14211.
G. Bubu, 600-800 ft., King’s Collector 8420.
PAHANG. Tras Valley, 3,800 ft., S.F.N. 21636
(Hoittum). Base of G. Senyum, S.F.N. 22388
(Carr).
SELANGOR. Ginting Simpah, Hume 9110 (F.M.S.
Mus. Herb.).
LOWER SIAM. Khaw Poh Hill, Khsum, 1,200 ft.,
S.F.N. 3829 (Haniff & Nur).

Vol. XI. (1940).

102

19. Diplazium sorzogonense Presl, Tent. Pterid.: 114. 1836
sp dail sorzogonense Presl, Rel. Heenk. 1: 45.
1825.

Stock short erect. Stipes to c. 60 cm. long, scaly at
base with long narrow entire scales, scales dark brown,
sometimes with black edges, to about 1:5 cm. by 1:5 mm.;
rest of stipe and rachis scaly with very narrow dark hair-
like scales. Lamina simply pinnate, to about 80 by 40 em.,
pinne to c. 20—jugate below pinnatifid apex of frond; lower
pinne very shortly stalked, upper sessile, bases truncate or
subtruncate, apices acuminate, margins lobed 34, or more
to costa in lower pinnzx, less deeply in upper ones; lobes
almost at right angles to costa, oblong, ends rounded and
slightly toothed, 4-6 mm. wide above base; texture firm,
drying rather dark; largest pinna to 25 < 5 cm., commonly
less, veins in the lobes to about 10 pairs, all simple, basal
14 usually free of sori; sori reaching almost to margin, the
lamina below the sorus depressed, the depression with a
distinct margin on the dried frond. Sori usually all simple
near the base of a pinna, the acroscopic basal sorus of each
lobe diplazioid towards apex of pinna. Indusia firm,
persistent, rolled back when sorus is mature.

This species is found throughout Malaya, in lowland
and mountain forest to 4,500 ft. altitude. It is a very
widely distributed species, and though somewhat variable
is always recognizable by the fibrillose stipe and rachis and
the peculiar impressed sori, which I have not found in any
other Diplazium of this region (except slightly in D.
subserratum).

D. sorzogonense differs from D. speciosum in having
entire scales, fibrillose rachis, and in the impressed sori. I
think that the two species have sometimes been confused.

Specimens in Singapore Herbarium :—

SINGAPORE. Stagmount, Ridley s.n., 1907; Ridley
11271. Bukit Timah, Ridley s.n. 1897. Mandai
Road, Corner s.n., 20.10.29.

PERAK. Kinta, 1,500-2,000 ft., King’s Collector
7151. Taiping, Scortechini, March 1884. Tapah,
Ridley 14023.

SELANGOR. Kwang, Ridley 13432.

PAHANG. Fraser’s Hill, 4,000 ft., S.F.N. 11198 (Md.
Nur); s.n., Holttum’; 20.3.29; s.n.,. Holitom
31.12.39. Cameron Highlands, 4,500 ft., s.n.,
Holttum 21.5.36. G. Rokam, P. Tioman, 2,000
ft., S.F.N. 18612 (Henderson).

PENANG HILL s.n. Dec. 1881, Hullett. S.F.N. 19302
(Holttum).

KEDAH. G. Lang 2,000 ft., S.F.N. 35012 (Kiah).

Gardens Bulletin, S.S.

103

20. Diplazium speciosum Bl. Enum. 193. 1828.

Stock short erect. Stipes to 60 cm. long, clothed at
the base with shining brown scales with toothed but not
blackened edges, about 1-5 cm. long and 1:5 mm. wide at
base; rest of stipe and rachis glabrescent. Lamina pinnate,
to about 70 * 25 cm., pinne to about 20—jugate, to 20 cm.
long and 3-5 cm. wide, lower pinne on stalks to 4 mm. long,
upper sessile, base truncate (on lower pinne sometimes
narrow), apex acuminate, margins lobed more than 34, way
to costa in lower pinne; lobes 4—5 mm. wide above base,
slightly oblique, oblong, ends rounded, edges shortly toothed,
texture firm, drying rather dark brown; veins to 10 pairs
in each lobe, simple, sori from base of veins over 2/3 their
length, not impressed, often all simple, the indusia thin but
firm, not rolled back when sorus is ripe.

In Malaya this species has only been found on G. Tahan,
specimens from which differ from Kinabalu specimens in
wider lobes of the pinnze; Peninsula specimens agree
however with that from Mt. Dulit, Sarawak, quoted below.

D. speciosum has much the general appearance of D.
sorzogonense, but lacks the fibrillose scales on stipe and
rachis, has toothed scales at the bases of the stipes, and has
sori not impressed. The lobes of the pinne are also rather
more deeply cut and more strongly toothed.

Specimens in Singapore Herbarium :—

PAHANG. Gunong Tahan, Ridley 15991; 5,000 ft-,
Haniff & Nur. S.F.N. 7953; 4,500 ft., S.F.N.
20747 (Holttum).

BORNEO. Dulit, C. Hose, Herb. Bishop Hose. Kina-
balu, 5,000 ft., S.F.N. 25444 (Holttum).

Var. major (Bedd.) Holttum stat. nov.
D. sorzogonense v. major Bedd. Suppl. 40. 1892.

Differs from the typical form of the species in having
fronds to 1:50 m. long, pinne to 6:8 em. wide, the lobes more
acute and more strongly toothed, veins to 13 pairs, mostly
forked.

PERAK. G. Bubu, 5,000 ft., King’s Collector 7403
(type of variety); G. Bubu, 5,000 ft., Wray

3862.

PAHANG. Pine Tree Hill, 4,500 ft., S.F.N. 36508
(Hoittum).

SUMATRA. G. Sibayak, 5,000 ft., S.F.N. 15457
(Holttum).

The type of Beddome’s variety clearly belongs to this
species and not to D. sorzogonense. It is a very fine fern
and might perhaps rank as a species.

Vol. XI. (1940).

104

21. Diplazium subintegrum Holttum, Gard. Bull. S.S. 9:
125. 1937.

Stock short erect. Scales on stipes brown, entire, to
about 1:5 cm. by 2mm. Stipes to 80 cm. long, pale when
dry. Frond simply pinnate, to about 70 cm. long; pinnz to
about 16 pairs, the apical lamina narrowly deltoid, lobed at
the base. Pinne to 24 by 2-5 cm., lowest stalked to 1-5 cm.,
upper 2 or 38 pinne only sessile; bases almost equal, of
lower pinne narrower, cuneate; edges toothed, apices
acuminate. Veins forked near the costa, the upper branch
always simple, the lower branch forked 1-4 times; sori on
upper branch from base 2/3-34 of way towards margin,
diplazioid, usually also on one branch only of the lower
vein, simple.

A mountain species, related to D. montanum and D.
pallidum, differing from the latter in larger brown scales,
longer stalks and narrower bases of pinnez, more branched
veins and normally diplazioid sori; differing from D.
montanum also in the narrowly deltoid lobed apex of the
lamina. This species appears to be confined to the Main
Range of the Malay Peninsula and the Taiping Hills, at
3,000-5,000 ft. The specimens in the Singapore Herbarium
were cited with the original description quoted above; no
more have since been acquired.

22. Diplazium subserratum (Bl.) Moore, Index 338. 1862.
Asplenium subserratum Bl. Enum. 174. 1828.

Stock slender, erect or decumbent. Stipes slender,
clustered, 10-20 cm. long, glabrescent, clothed when very
young with small very dark scales, 2-3 mm. long and less
than 1 mm. wide, the edges very shortly toothed. Lamina
simple, to about 50 cm. long and 4 cm. wide, narrowed
gradually to the base which is suddenly contracted from a
width of a few millimetres, and very gradually to the
acuminate apex; margins entire or irregularly undulate,
toothed towards apex; texture very firm. Midrib grooved
above, prominent beneath and bearing scattered dark brown
ovate—lanceolate scales 1 mm. long or less; scales also on
some of the lateral veins beneath. Veins in small groups at
an angle of about 60° to the midrib, the groups about 5 mm.
apart. Each vein-group formed by forking at the midrib,
the acroscopic branch being simple and the lower forked
2—4 times again; the simple acroscopic branch usually sori-
ferous and often one, more rarely two, branches of the
basiscopic vein also. Sori on acroscopic vein diplazioid,
varying in length, often from the base almost to the edge
of the lamina; other sori simple and shorter; indusia firm,

Gardens Bulletin, SS.

105

persistent, when old rolled back as in D. sorzogonense, and
leaving a faint mark on the lamina to show their size and .
position when young.

This species in Malaya is confined to moist shady
mountain valleys, near streams. It is the only Malayan
species with narrow simple fronds, and so is easy to
recognize. Young plants have fronds which are dissected
to the midrib, indicating a probable origin from a pinnate
ancestor. The resemblance of the indusia to D. sorzogonense
is interesting.

D. subserratum is probably allied to D. lanceum of
India and China, but is larger, and has a shorter rootstock
which is more or less erect with tufted fronds; the scales
also appear to be smaller.

23. Diplazium tomentosum Bl. Enum. 192. 1828.

Stock short, erect. Scales on stipes to about 4 mm.
long and 14 mm. wide, acuminate, dark to medium brown,
the edges irregular but hardly toothed. Stipes to c. 40 cm.
long, frond to about 35 by 15 cm., pinne many, lowest
shortly stalked (to 3 mm.), upper sessile or adnate; lowest
pinne largest, somewhat deflexed and curved, widest 1/3
from base, base narrow, apex acuminate, to 3 cm. wide;
middle and upper pinne narrower, with parallel sides,
almost at right angles to rachis, slightly curved, base
truncate, auricled above, broadly cuneate below. All pinnx

of large fronds deeply lobed (half way or almost to the

costa), lobes c. 3 mm. wide above the base, at an angle
of less than 60° to the costa, their edges slightly toothed at
the tip. Veins pinnate in the lobes, the veinlets usually
forked, up to 7 pairs or more, often all or almost all veins
soriferous, the lowest acroscopic vein with sorus directed
towards the main costa or diplazioid. Texture firm, drying
very dark; rachis and costxz beneath densely covered with
short multicellular hairs with pale cells and dark brown
cross-walls.

Young plants have almost entire pinne ; these approach
D. crenatoserratum in appearance but are distinguished by
their more hairy rachis and costz beneath, usually narrower
more pointed pinne, more deflexed, shorter-stalked lowest
pinne which are broadest near the middle, and darker colour
on drying. Fully developed plants, with their deeply lobed
pinne and deflexed almost elliptical lowest pinne are quite
distinct.

D. tomentosum is closely related to D. velutinum, a
mountain plant with more deeply dissected, though hardly
larger fronds. The tomentum of both is of the same
character.

Vol. XI. (1940).

106

D. tomentosum is a common species in lowland and
mid-mountain forest throughout Malaya; it seems unneces-
sary to enumerate all the many specimens.

24. Diplazium velutinum Holttum Gard. Bull. S.S. 9: 126.
1937;

Caudex short erect. Stipes to 40 cm. long (commonly
about 20 cm.), clothed throughout with scattered multi-
cellular brown hairs, and towards the base with narrow
almost black entire scales, scales to about 1 cm. long and
barely 1 mm. wide at the base. Lamina bipinnate, to about
25 em. long and 15 cm. wide, the rachises and costxz beneath
densely clothed with brown hairs like those on the stipe,
with a few small narrow scales. Free pinne 12-15 pairs,
the sub-basal largest, the upper ones gradually reduced and
grading into the lobed apical lamina. Largest pinne c. 8
cm. long and 3 cm. wide (commonly about 6 by 2 cm.),
subsessile, narrowed from truncate base to acute or
acuminate apex, pinnate towards the base only, for the rest
lobed almost to the costa, the pinnules or segments at right
angles to the pinna-rachis at the base, gradually more
oblique towards the apex. Free pinnules few, the rest more
or less broadly adnate to the pinna-rachis; largest pinnules
about 12 mm. long and 4 mm. wide, base unequally cuneate,
margins serrate or lobed as much as half-way to the costule,
apex rounded, texture very firm. Veins in pinnules or
pinna-lobes 7-10-jugate, mostly forked; sori on basal 14
—2/3 of veinlets; in the largest pinnules several diplazioid
sori, in smaller pinnules only one such sorus; indusia thin,
broad, persistent.

This species, closely related to D. tomentosum but much
more deeply dissected, has hitherto only been found in the
moist shady valleys at Cameron Highlands, 4,800—5,000 ft.
above sea level. No further collections have been made
since my paper above quoted.

25. Diplazium xiphophyllum (Baker) C. Chr., Ind. Fil. 241.
1905.

Asplenium xiphophyllum Baker, J. Bot. 1879: 40.

Stock short, erect. Stipes to 70 cm. long, drying pale,
clothed at the base with dull brown entire scales to 1:5 cm.
long and 2 mm. or more wide at the base. Lamina simply
pinnate; pinnz in adult plants to 12 pairs, terminal pinna
like the rest, texture thin, colour light green when fresh,
light brownish when dry; lower pinne on very short stalks
(2 mm. long), uppermost slightly adnate to rachis, with
buds sometimes in their axils. Largest pinne to about 35
by 6 cm., elliptical, narrowed gradually to the slightly
unequal base and more abruptly to the acuminate-caudate

Gardens Bulletin, S.S.

107

apex; edges slightly and irregularly toothed throughout;
midrib pale, prominent and glabrous beneath, lateral veins
in small groups, at 45° to the midrib, commonly of one
basal pair and one central vein which is forked, one or both
of its branches also forked; veins occasionally anastomosing
near margin but never copiously anastomosing, ending
distinctly within the margin and not united to it. Sori on
basal pair of veins in each group, from near midrib to near
margin, on other veins shorter, commonly 3—4 sori on each
vein-group, the sorus on the anterior basal vein only
diplazioid.

The above description applies to ferns found in the
Malay Peninsula. The type of the species D. xiphophyllum
is from Borneo (Labuan, coll. Burbidge, at Kew), and shows
a frond of a young plant with narrow pinne, and detached
pinne from a mature plant, these being 30 by 3 cm. There
is therefore some doubt whether the Peninsula plant
(matched also by the Kinabalu specimens quoted below) is
really this species, more especially as young plants from
Malaya have quite broad pinne. I do not, however, know
any other species to which our fern could be referred.

D. xiphophyllum, as here interpreted, is a fern of
lowland and mid-mountain valleys, the highest altitude
recorded in Malaya being 3,000 ft. It is fairly constant
‘in character, differing from D. riparium in its much larger
size and in other points mentioned under that species.
There is one aberrant collection, that of Burkill from Ulu
Chineras, which has long narrow pinne (to 3:8 cm. wide),
long acuminate with edges more deeply toothed than
normally, the vein-groups less branched and _ without
anastomosis, and fewer sori in each group.

Specimens in the Singapore Herbarium :—
SIAM. Patani, 160-350 m., Eryl Smith 1922; Khao
Rum, Eryl Smith 618.

PERAK. No loc., Herb. Mus. Perak 234; Jor Camp,
Henderson F.M.S. Mus. 10833; Waterloo, Curtis
1361 p.p.; Larut, 3,000 ft., King’s Collector 2698 ;
Temango, Ridley 14227.

PAHANG. Ulu Chineras, K. Lipis, S.F.N. 15692
(Burkill).

SELANGOR. Ginting Peras, Ridley 7833; Ginting |
Simpah, 2,000 ft., S.F.N. 34310 (Md. Nur),
1,500 ft., Hume 8782, 8777, 8542; Kajang,
Symington 24200 (F.D.).

BORNEO. Kinabalu, 5,000 ft-, Clemens 29406; 3,500
ft., S.F.N. 25258 (Holttum).

Vol. XT. (1940).

108

LIST OF COLLECTORS’ NUMBERS CITED

(The figures in brackets refer to the numbers given

S.F.N.
3829 (18)
7953 (20)
8844 (15)
9770 (13)
9876 (2)
9982 (13)

10018 (2)

11007 (10)
11198 (19)
11473 (2)

11478 (15)
11515(13)
11563 (11)
14103 (13)
15323 (16)
15403 (5)

15457 (20 var.)

15692 (25)
17978 (2)

18301 (16)
18583 (12)
18612 (19)
19302 (19)
19342 (13)
19346 (2)

19498 (17)
19776 (13)
20554 (2)

20747 (20)
20792 (16)
21636 (18)
21645 (15)

22089 (13)

22388 (18)
22422 (17)
23262 (13)
23340 (2)

23361 (13)
23518 (13)
23620 (5)

24382 (17)
24558 (16)
24703 (12)
24704 (17)
24780 (16)
25258 (25)

25444 (20)
31194 (9)

31295 (13)
31334 (5)

32937 (13)
34310 (25)
35012 (19)
35995 (17)
36503 (15)
36504 (10)
36505 (2)

36507 (18)

06508 (20 var.)

37159 (2)
37201 (15)
Ridley
1587 (16)
2167 (12)
2400 (13)
5818 (17)
7833 (25)
7843 (9)
7852 (9)
8139 (12)
8652 (9)
11271(19)
13432 (19)
13975 (13)
14023 (19)
14206 (13)
14209 (12)
14211 (18)
14227 (25)
14229 (2)
15991 (20)

King’s Collector
(Kunstler)

2214 (18)
2346 (18)
2391 (14)
2698 (25)
7146 (14)
7151(19)

7403 (20 var.)

7894 (16)
8420 (18)
8797 (9)

to the species in this paper)

10849 (13)
10959 (12)
Hume (Herb.
F.M.S. Mus.)
8551B(13)
7117(17)
8186 (16)
8542 (25)
8777 (25)
8782 (25)
8997 (14)
9110(18)
9347 (14)
9665 (14)
9677 (14)
Curtis
1352 (8)
1360 (17)
1361 p.p. (2)
1361 p.p. (25)
3369 (14)
10135 (13)
Eryl Smith
618 (25)
1198 (5)
1922 (25)
1927 (14)
1929 (17)
1930 (14)
1948 (12)
Forest
Department
14602 (13)
24200 (25)
Herb. F.M.S.
Mus.
10833 (25)
Wray
3590 (16):
3594 (17)
3616 (2)
3626 (17)
3862 (20 var.)
Haniff
4047 (12)
Alvins
3282 (16)

Gardens Bulletin, S.S.

109

CARPET GRASS, AXONOPUS SPP.

By R. B. JAGOE
Botanist, Department of Agriculture, S.S. & F.M.S.

INTRODUCTION

For many years the apparent difference between
broad-leaved and narrow-leaved forms of Azonopus
compressus, or Carpet grass, has not been generally
regarded as of specific rank, although the names Paspalum
platycaule Poir. for the broad-leaved form, and Paspalum
compressum, for the narrow-leaved form, have been used
by some persons. The original Paspalum compressum,
Swartz., is however, undoubtedly the broad-leaved grass.

These two Carpet grasses have now been recognised
as distinct species, the broad-leaved form retaining the
present name (Axonopus compressus Beauv.), while
Miss Agnes Chase, an American authority, has defined the
narrow-leaved grasS aS a new species, and named it
Axonopus affinis, as follows :—

Axonopus affinis, sp. novo.

Ab Axonopo compresso differt: Culmis et stolonibus
gracilioribus, laminis augustioribus; spiculis brevioribus, 2
mm. longis, obtusis vel subacutis. (Culms and stolons more
slender, leaf-blades narrower ; spikelets shorter, 2 mm. long,
obtuse or sub-acute). |

“The plants are more tufted than in A. compressus,
sometimes forming dense mats with short rhizomes, and
the flowering culms in such colonies are relatively few.

Stolons slender, the internodes short, and the leaf-blades
(usually) not conspicuously shorter than those of the culms,
as in A. compressus.

Culms erect or geniculate ascending, slender, nodes
glabrous (?) sheaths compressed, on the average narrower
than those of A. compressus, leaf-blades flat or folded, 2 to
6 mm. wide.

Peduncles very slender 1 to 3 within uppermost sheath
of flowering culm, finally elongate; racemes 2 to 4; 2 to 10
cm. long, spikelets oblong-elliptic rather more plump than
in A. compressus, 2 mm. long, 0-8 to 0-9 mm. wide, blunt or
abruptly acute. The glume and sterile lemma equal, covering
the fruit or slightly pointed beyond it, 4 nerved, mid-nerves
suppressed, sparsely silky-pilose at base and summit, and
sometimes along the nerves. Fruit pale, 1-7 to 1-8 mm. long,
blunt.” (Agnes Chase).

HISTORY

Axonopus compressus, Beauv., is a native of the West
Indies and tropical America, and has spread to Florida and
Louisiana in the United States. It was introduced to
Singapore about 1895 to 1900, for Ridley refers to Paspalum

Vol. XI. (1940).

110

platycaule in 1903 as “lately established in Singapore’’, since
when it has spread to all parts of Malaya, thriving under
light or medium shade.

Axonopus affinis, Chase, is a native of the southern
United States, and is common in the southern States from
Carolina to Texas. This species was introduced as “Carpet
grass” to the Central Experiment Station, Serdang, in 1921,
seed being sent to the Department of Agriculture by the
late Professor C. V. Piper of the Agronomy Section of the
United States Department of Agriculture. It has been
distributed to many parts of the peninsula, but has not
yet spread widely.

DESCRIPTIONS

The two grasses are readily distinguished, although the
broad-leaved species, A. compressus, in some circumstances,
such as dry, shady conditions, produces long comparatively
narrow leaves, little different superficially from those of
A. affinis.

The distinguishing characters of the two species as
found in Malaya differ slightly from those given by
Miss Chase and detailed descriptions will be given below.

As stated above, the plants of A. affinis Chase, are more
densely tufted than those of A. compressus Beauv. The
former species is normally more free flowering, and the
inflorescence peduncles are usually more elongated or
extended. |

Leaves.—Growing in the same conditions, the leaves
of A. compressus are larger than those of A. affinis, the
blades being much broader, though shorter.

Further distinguishing features are as follows:—

Leaf-blades of A. compressus are wavy-edged and
bright green, with conspicuous hairy fringes along the
edges of the lower three-quarters of the blades. The edges
of the sheaths and the ventral surfaces of the blades are
hairy, with a few hairs also on the backs of the blades. The
blades taper more at the ends and have less blunt-pointed
tips.

The leaf-blades of A. affinis are longer and stiffer, and
are slightly duller green in colour. They are very much less
hairy, sometimes hairless, but usually with comparatively
few, though long, hairs along the edges of the blades and
sheaths, and seldom any on the surfaces of the blades. The
edges of the blades are nearly parallel and the tips are blunt.
The ends of the leaf-blades in both species are minutely
hairy.

The backs of the blades of A. affinis are much more
prominently keeled, and tend to fold in two, especially when
withered or dry.

Gardens Bulletin, S.S.

111

oe An ae eT

as “ANID EY ‘enesoudiues oni oun ey?

oe ee i ee are

Vol. XI. (1940).

112

Ax onopus ayfinis » Chase .

Gardens Bulletin, S.S.

113

A further point of difference between the two species is
that a red coloration in the leaves of A. compressus is very
common especially in dry exposed situations, but it is
extremely rare in those of A. affinis.

Stolons.—Both species are freely stoloniferous; stolons
vary from fairly slender to comparatively stout, and there
is little if any difference between the species in this respect.
Stolons are branching; inter-nodes are short, shorter on the
average in A. affinis; nodes are free-rooting.

In A. compressus the stolon nodes are very white-hairy,
but in A. affinis they are only sparsely hairy.

In A. compressus the leaf-blades of the stolon growth
stages are often conspicuously shorter than those of
flowering culms, while in A. affinis this is not so notable.
| A. compressus is now fairly common under rubber
trees, and there grows most often in tufted fashion,
conditions presumably being not very favourable for normal

stoloniferous growth.

Flowering Culms.—Every stolon node bears a potential
flowering shoot and when circumstances demand or are
suitable, shoots from the nodes form flowering culms. At
times, small or large leafy colonies of flowering culms are
produced on branching rhizomes from the stolon nodes.

The flowering culms proper have only one inter-node,
though at times they appear to have two or three. These
extra internodes are the result of stoloniferous internodes
being formed and assuming a more or less erect attitude,
due to the urge to form floral organs coinciding with good
conditions for vegetative growth.

The flowering culms develop with the elongation of the
basal internode, from the upper node of which and within
the final leaf-sheath is a conic “‘disc’”’ bearing 4 to 6 long
bracts each subtending an inflorescence peduncle.

The blade of the uppermost leaf is short, especially so
in A. affinis.

The nodes of the flowering culm are, in A. compressus,
most usually densely white-hairy, though not always all the
way round, while in A. affinis only the edges of the
compressed nodes are (usually) hairy, but sometimes quite
glabrous.

The full details of the morphological differences of the
flowering culms are however, more conveniently described
by taking each species separately.

Axonopus compressus Beauv. Flowering culms fairly
slender, compressed. Nodes more or less silky hairy, leaf-
blades 5 to 14 mm. wide, 6 to 28 cm. long; uppermost blade
usually short but from 3 to 19 cm. long. From the cone
on upper node of culm, 4 to 6 slender peduncles (1 to 4
conspicuous) 10 to 30 cm. long. Very frequently in A.

Vol. XI. (1940).

114

Axonopus Compressus. Beaue

Gardens Bulletin, S.S.

Axonopus affinis _ Chase

Vol. XI. (1940).

116

compressus, these inflorescence peduncles remain short, and
are not elongated beyond the mouth of the leaf-sheath.
Each peduncle bears 2 or 3, rarely 4 to 6 slender racemes,
the first two of which are at the apex of the peduncle, one
terminal, the second subterminal, with others, when present,
cin una at short intervals below. Racemes 4 to 10 cm.
ong.

Spikelets usually silky hairy, not closely spaced,
alternating in two rows on a slender flattened or triangular
rachis. Spikelets 2:0 to 2:8 mm. long, 0-8 to 1:0 mm. broad.
Glumes II and III membranous, pale green with darker
veins, hairy, conspicuously pointed, and extending beyond
the fruit. Glume II often reddish coloured. Glume IV and
palea chitinous. Stamens three, anthers yellow (when full
of pollen) with narrow purple outer edges. Stigmas white.

Axonopus affinis Chase. Usually more free flowering
than A. compressus. Flowering culms slender, compressed.
Nodes more or less glabrous. Leaf-blades 4 to 8 mm. wide,
6 to 30 cm. long, uppermost blade on culm very short, 2 to
6 cm. long. From cone on upper node of culm 3 to 6 very
slender peduncles (1 to 3 conspicuous), 15 to 30 cm. long.
Each peduncle bears 2 or 3, occasionally 4 to 6, slender
racemes, the first two racemes at the apex of peduncle, one
terminal, the second subterminal, the others, when present,
eae at short intervals below. Racemes 3 to 10 cm.
ong.

Spikelets usually hairy, not very closely spaced,
alternating in two rows along a slender flattened or
triangular rachis, 1:8 to 2:2 mm. long, 0-8 to 1:0 mm. broad
(they are much less pointed than those of A. compressus).
Glumes II and III membraneous, pale green with dark veins
hairy, hardly extending or extending to a short point beyond
the fruit, mid vein very indistinct. Glume II often reddish
coloured. Glume IV and palea chitinous. Stamens three,
anthers purple with translucent dots. Stigmas mauve.

Both species produce seed freely.

UTILITY

A. affinis withstands dry weather better than A.
compressus and is capable of covering the soil better in
dry exposed situations. On poor, unshaded soil, in this
country, the former species frequently grows by densely-
leafy, short internoded stolons, and it is probable that it
will, in time, largely predominate in such conditions.

Under medium shade, however, where both make their
best growth, A. compressus appears to have a distinct
advantage.

| Gardens Bulletin, SS.

117

There is some slight confusion as to palatability and
value of carpet grass as pasture, but this is partly due to
the previous confusion of two species with but one name.

There is no doubt that the true Axonopus compressus
Beauv. is a good pasture grass, being recognised as such in
tropical America and Australia. It appears equally true
that A. affinis has a variable reputation. It is definitely
regarded as inferior in Australia, and chemical analyses
quoted by McLennan clearly show it to be inferior, as fodder,
to A. compressus.

In Malaya a few years ago A. affinis appeared to be
liable to cause purging of cattle, due to a slight increase
in hydrocyanic acid content if cut at too short intervals.
The author has, however, seen A. affinis, under light shade,
grazed by cattle with relish, and obviously in preference to
Paspalum conjugatum.

SUMMARY

A detailed account is given of the two species Axonopus
compressus Beauv., and Axzonopus affinis Chase, and the
differences between them.

The chief distinguishing features may be summarised
as follows :—-- .

A. compressus Beauv. A. affinis Chase.
Savannah Grass. Carpet Grass.
LEAVES.—Broad, hairy frequently Narrower, stiff, folded,
with red colouration. more or less glabrous.
NopEes.—Very hairy. Hardly hairy.
SPIKELETS.—Long and pointed. Blunt or short pointed.
ANTHERS.—Yellow with narrow Purple, with translucent
purple outer edges. dots.
STIGMAS.— White. Mauve.

_ The writer wishes to acknowledge with thanks much
careful work by his Laboratory Assistant, Che Mohd.
Kassim, in the preparation of the illustrations.

BIBLIOGRAPHY

1. AGNES CHASE. “The Carpet Grasses.” Journ. Wash. Science,
Vol. 28, 4. April 1938.

2. C. V. PIPER and L. CARRIER. “Carpet Grass.” U.S. Dept. of
Agriculture, Farmers Bull. 1130, April 1920.

3. A. S. HitcHcock. “Grasses of British Guiana.” Contribution
from U.S. Nat. Herbarium. Vol. 22.6.1922.

4. H. N. Ruiptey. “Turf and fodder grasses.” Agricultural
Bulletin S.S. and F.M.S. Vol. 2, 9. Sept. 19038.

5. H. N. Riputey. “Flora of the Malay Peninsula, 1925.” Vol. V,
page 216.

6. J. N. Mitsum. “Fodder and grazing Grasses in Malaya.”
Malayan Agricultural Journal, Vol. 12, 1924.

7. J. LAMBOURNE. “A _ preliminary report on carpet grass.”

Malayan Agricultural Journal, Vol. 12, 1924.

Vol. XI. (1940).

118

J. LAMBOURNE. “Lawn grasses and their maintenance.”
Malayan Agricultural Journal, Vol. 25, 1937.
C. D. V. GEorGI. “Chemical composition of Carpet grass from

Serdang (A. affinis).” Malayan Agricultural Journal,
Vol. 24, 19386.

K. HEYNE. De Nuttige Planten van Nederlandsch Indie. (The
useful plants of the Neth. Indies.) 1927, page 208.

C. WHITE. “Broad leaved carpet grass.” Queensland Agri-
cultural Journal, Vol. 48, 1935, p. 503.

C. WHITE. “Carpet Grass.” Queensland Agricultural Journal,
Vol. 44, 1985, p. 117.

L. W. McLENNAN. “Carpet Grass.” Agricultural Gazette
N.S.W., Vol. 47, October and November 1936. —

D. O. Cross. “Narrow-leaved Carpet Grass.” Agricultural
Gazette, N.S.W., Vol. 49, Dec. 1938.

WHITNEY, HOSAKA and RIPPERTON. Grasses of the Hawaiian
Ranges. Hawaiian Agricultural Experiment Station,
Bulletin 82, 1939.

Gardens Bulletin, SS.

119

PERIODIC LEAF-CHANGE AND FLOWERING OF
TREES IN SINGAPORE (II)* |

By R. E. HOLTTUM

In my earlier paper on this subject, I published
observations made on a number of trees during the years
1927-1931, and summarized previous observations published
elsewhere on the behaviour of trees in the eastern tropics.
I also gave a short account of the climate of Singapore. The
main item of new information arising from my own
observations was that a number of deciduous trees had
fairly constant leaf-periods, which had no relation to the
yearly calendar; and the conclusion drawn from this was
that the length of leaf-period in such trees is due primarily
to senescence of leaves, the climate of Singapore being so
uniform as to have little or no control.

Subsequent observations, in some cases over a total
period of ten years or more, have confirmed the fact that
many deciduous trees have fairly constant leaf-periods not
related to climate; but they have also shown many curious
irregularities which are not always easy to explain. It is
clear that there is every variety of response from almost
complete independence of climatic change to extreme sus-
ceptibility, and a tree that is regular in behaviour for a time
may later show an unexpected change (e.g. Koompassia).

My earlier generalizations therefore need modification
in some measure, and a fuller discussion of the whole
subject is given below, followed by details of behaviour of
individual trees.

Climate

The temperature of Singapore is so uniform throughout
the year as to have little significance in the present
connection. The only phenomena of temperature-change
which influence plant behaviour are sudden falls of 10 to
15 degrees (Fahrenheit) during storms, and possibly also
cool days following hot weather. These stimulate certain
plants to flower ; chiefly orchids, but also some woody plants,
notably Pterocarpus indicus.

The average rainfall of Singapore is nearly 100 inches
(2540 mm.), uniformly distributed, the driest calendar
month having a mean of about 6 inches (152 mm). The
only constant periodic change in the climate is the wetter
season from October to January, when in addition to the
greater rainfall there is frequent heavy clouding and light
rain, resulting in continuously moist conditions over
considerable periods, in contrast to the relatively short

* For the leh paper on this subject, see Gard. Bull. S.S.
5: 174-206. 1931

Vol. XI. (1940).

120

periods of heavy rain characteristic of most of the rest of
the year. There is often a sudden break at the end of the
wet season, in January or February, and this is usually
the most pronounced climatic change in the year. It varies
in its intensity.

Though the driest month has a mean rainfall of about
six inches, there are drier months than this in almost every
year. A list of the calendar months with less than 2:5
inches of rain (63 mm.) since 1890 shows the following
frequency of occurrence:

February ‘|

March a PA ss 3

April ef a¢, at _ Raases

July OF rs! aL 6

August ‘ba oe he 3

September MM Z
Thus there is an indication that exceptionally a weather
tends to occur either about February—March or July—
August, but it will be noted that in 49 years there were only
22 calendar months with less than 2-5 inches of rain, and
(as shown by the figures at the end of this paper) there is
very great variation in the distribution of rainfall from
year to year. In the period 1927-1937, the following were
the longest periods without rain :—

24. 1.35-14. 2.35 sis eh 22 days
21. 7.35— 9. 8.35 =e 5a 4 | Rare
3. 2.33-19. 2.33 ye os Yaa
8.12.34—23.12.34 be os AG65(T,
13. 7.28-27. 7.28 ee: te Lhe
26. '(.29- 9, 8.29 es <e 15) gy
24, 7.37— 6. 8.37 14

The extremes of dry weather are thus rele slight
as compared with most tropical climates, and are never
such as to necessitate leaf-fall of trees; and though extreme
dry weather may usually occur about February and July,
it is not frequent enough to ensure uniform behaviour of
all trees, and lesser dry periods may occur at almost any
time. Dry periods of from 7 to 13 days occurred as follows
in 1927-19387:

in January
February
March
April
May
June
July
August e
September ..
October
November ..
December

A period of seven days without rain, Lolieriaie a period
of wet weather, is probably enough to stimulate certain

periods

SB | nNoPRNOW OF OTP LO

Gardens Bulletin, S.S.

121

trees to flower; but response varies according to the
condition of the tree, and is in fact so complex that it is:
difficult to analyse. Reference to Fagraea fragrans and
Saraca taipingensis will give some indications on this point.
Most trees (other than those that flower continuously)
probably need more than seven days drought as a stimulus
to flowering.

Atmospheric humidity is high every night, rising above
90% soon after sunset; in the daytime it may fall to about
60% in dry weather, but in the wet season often only
to 85%.

An appendix to this paper gives the daily rainfall, in
on at the Botanic Gardens, Singapore, from 1927 to
1939.

Deciduous trees with annual leaf-renewal

Trees of the following species have been strictly annual

in their leaf-renewal:
Kigelia pinnata (mean date 28 January)
Hymenaea courbaril (early February)
Parkia javanica (mean date 23 February)
Terminalia subspathulata (mean date 6 April)
Canarium rufum (mean date 9 May)
Lecythis sp. (mean date 11 August)
The coefficients of variation of these trees vary from 2:0%
to 75%, and presumably indicate corresponding differences
in susceptibility to climatic variation.

The mean dates of leaf-renewal are interesting. It is
probably significant that the two trees with the lowest
variation (Parkia and Kigelia) renew their leaves in
January—February, following the most regular seasonal
change in the climate, whereas those with mean date of
leaf-renewal in April and May (Terminalia and Canarium),
when there is much variation of climate conditions, show
the greatest variation. Lecythis, changing about August,
is intermediate. Hymenaea is not quite deciduous, and no
exact dates are recorded, but the tree concerned seems to
be quite regular in its behaviour.

There are other trees that appear to have a tendency to
a period of slightly less or slightly more than 12 months.
Such trees as these may keep to very near 12 months for
some years, if they change their leaves at a time when dry
weather is most likely to occur; but sooner or later they
have a short or a prolonged leaf-period, and this may be
the prelude to a series of irregular periods until the change
occurs once more at a time when climatic change is most
regular. For an example of shortening periods see Cassia
nodosa; for lengthening periods Koompassia malaccensis.
The latter had been observed over three periods when I

Vol. XI. (1940).

122

wrote my earlier paper; it then appeared to have a period
of exactly a year, with leaf change in September. It
subsequently deferred its change until November and then
had several periods of considerably more than 12 months.

Two young trees of Kigelia pinnata have indicated that
the date of leaf-renewal may be dependent on the date of
planting, and also that root-stimulus by manuring and
cultivation may induce new growth and result in a change
of some months in the date of leaf-renewal. The relation
of the first leaf-change to the date of planting of a young
tree is one that varies considerably from species to species,
and observations on this point would be of considerable
interest.

Deciduous trees with leaf-period of more than a year
Several deciduous trees have been observed to show a
mean leaf period significantly more than 12 months. These
are as follows:

Period No. of

Species (months ) eat. periods
Cedrela Glaziovii A .. 12-5 ‘70 or 56% 11
Koompassia malaccensis 12-7 1-36 10-7% 12
Hevea braziliensis 520 S83 -86 6-5% 9
Cedrela Glaziovii B Se viay 1-7 121% 10
Caesalpinia ferrea a ah be 2 1-2 8-5% 5
Homalium~ grandiflorum 13-7 1-9 13-9%
Anisoptera megistocarpa 13-8? 1-65? 12:0% 10?
Parishia Maingayi 2. OSS 96 6-2% 6
Heritiera elata -» 20-5 2-6 12:7% 4
Heritiera macrophylla .. 32 1-3 4.1% 4

The two trees of Cedrela Glaziovii provide a remarkable
contrast, as they are growing near together and under
apparently identical conditions. It is remarkable that the
tree showing a period of only 12-5 months, which one might
perhaps think not significantly different from 12:0 months,
has had a very small coefficient of variation over a period
of eleven years. The other tree, which evidently has a
tendency to decidedly longer periods, is probably also more
influenced by climatic change. Climatic influence is more
clearly seen in the Koompassia, which has had its longest
leaf-periods when change was due in the wet season. Such
a tree may show almost exactly 12-month periods for
several years, then a series of longer periods with change
later and later in the wet season, then a series of nearly
12—month periods again. The Koompassia tree in question
would probably be a strictly 12—month tree in a climate
with a pronounced annual dry season. (Another Koom-
passia tree seems to be more irregular).

The Hevea with a very regular period of over 13
months is also interesting, because most Hevea trees in

Gardens Bulletin, S.S.

123

Singapore are very irregular in their behaviour, changing
some or all leaves at any period of dry weather (according
to their previous history and the intensity of the drought).
Schweizer’s observations in a much more seasonal climate
in Java indicate also variation in behaviour of different
trees of Hevea; for further notes on this question, see
below under the heading Deciduous Trees with no regular
leaf-periods.

The behaviour of the Anisoptera tree has _ been
decidedly irregular, and it is not at all obvious how climatic
influence can have caused the observed variations. The
Homalium tree also had one exceptionally long period which
seems unaccountable on climatic grounds.

The two species of Heritiera are the most remarkable
cases in this group, one being native to Malaya and the
other from Burma. It would be very interesting to have
some information about the behaviour of H. macrophylla
in its seasonal native climate. Such a leaf period as 214
years in a deciduous tree must surely be very exceptional,
though individual leaves of lesser evergreen trees and
shrubs in the shade of rain-forest may not infrequently last
as long as this.

Trees with leaf-periods between 6 and 12 months

A number of trees have shown regular periods of less
than a year; these are as follows, with the exception of
those having a six-month period, which are considered
separately below.

Period No. of
Species (months) s.d. periods

Adenanthera pavonina ..— 7-25 ‘74 or 10:2% 16
Cassia fistula Ae Gl 93 10-2% 15
Cassia fistula B 10-5 1-26 12:0% 8
Cassia fistula C 9-6 84 8-7% 14
Cassia nodosa (1928-35) 11-7 4 3:4% 7
Cratoxylon formosum A 9-1 1-09 12:0% 15
Cratoxylon formosum B- 9-1 ‘74, 8-1% 15
Delonix regia Mee ts | ‘94 10-7% it
Ficus variegata A 65 84 12-:9% 20
Ficus variegata B 64 85 13-3% 19
Ficus variegata Ci7)'6-1 -65 10:-7% 9
Ficus variegata THY ys Gek -62 10-1% 13
Lagerstroemia flos-

reginae, north side 9.2 -88 9-6% 13
Lagerstroemia _flos-

reginae, south side 8-9 ‘TT 8-7% 13
Lagerstroemia flos-

reginae old tree 9-4 -99 10-5% 13
Salmalia malabarica A_ 10-5 44 4.2% 13
Salmalia malabarica B 9.3 -98 10-5% 12
Sterculia macrophylla Pg. 6-9 — — 4
Sterculia macrophylla S. 7-0 — — 4
Sterculia sp. vom LOT 1-26 12-5% 15

Vol. XI. (1940).

124

It will be noted that nearly all of these trees have a
coefficient of variation of about 10%. It is notable that
one of the two exceptions, namely the tree of Cassia nodosa
with a coefficient of only 3-4% in the period 1928-35, showed
subsequently a very irregular behaviour. This is discussed
further under the heading of irregular trees.

Apart from the indication that a considerable number

of deciduous trees have regular periods independent of the
climatic cycle, the most interesting feature of the above
table is the comparison of different trees of the same
species. ;
In the case of Salmalia malabarica the tree with the
more regular period is the more vigorous. This species is
interesting as being the only one that has a considerable
obligatory resting period; such trees are often intolerant
of the Singapore climate. The standard deviations of the
three trees of Cassia fistula are also in the order of their
vigour, the most vigorous showing the least deviation. This
may be partly connected with the fact that in this species
no exact date for leaf-renewal can be given, the process
being slow and not always uniform all over the tree; the
observer is therefore liable to greater variation in recording
dates for the less vigorous trees. There is an indication
that some trees of this species in Singapore have a period
of 12 months.

The two trees Cratoxylon formosum are in different
localities, and have never changed simultaneously, but both
have a period of 9:1 months. One tree has a substantially
greater variation than the other, though both are equally
vigorous and of about the same size. It is to be noted that
this species appears to have a period of 12 months in the
more seasonal climate of Penang.

The four trees of Ficus variegata are of considerable
interest, as they come very near the six-month trees. It is
clear however that trees A & B, which have been longest
under observation, have a period significantly longer than
six months, and also a rather high standard deviation,
whereas the three six-month trees appear to have a low
standard deviation. As pointed out below, however,
observation must be carried out over a longer period of
years if a true picture of their behaviour is to be obtained.

The two trees of Lagerstroemia, like those of Cratoxy-
lon, are situated some distance apart, have almost equal
periods, and do not change simultaneously. The tree of
which the two halves have different periods is considerably
younger than the other. a

The Sterculia macrophylla at Penang is interesting as
showing a period of seven months in the much more
seasonal Penang climate. A second Singapore tree seems

Gardens Bulletin, S.S.

125

to have a considerably longer period; but it has not been
observed long enough since its treatment of cultivation and
manuring to allow of a definite statement being made.

Six-month trees

A few species of trees have a regular period of six
months. There are indications that in some cases trees
of Peltophorum have longer periods than this, but in the
cases observed the trees were either young or in poor
condition or affected with insect pests. The list of trees
observed in the six-month class is as follows :—

No. of
s.d. C.U. Mean dates periods
Couroupita
guianensis .. -43 72% 15 Mar. 20 Sept. 16
Peltophorum
pterocarpum .. -43 72% 12 Feb. 10 Aug. 22
Terminalia
catappa .. 64 £10:5% 8 Feb. 11 Aug. 22

It is remarkable that the exotic Couroupita has been
more regular than the local Terminalia in its behaviour.
The dates February—March and August-September for !eaf-
change follow closely the periods of incidence of the driest
weather and are probably controlled by these periods.

Deciduous trees with no regular leaf-periods

There is probably no very sharp distinction between
this class and that of the regular trees, the gap being
bridged by those which have high standard deviation. The
irregular trees here considered are in fact those which are
most affected by small climatic changes, which in Singapore
are of irregular occurrence. The most remarkable of the
irregular trees under observation are Cassia nodosa, Dyera
costulata, Ficus caulocarpa, Hevea braziliensis, Mangifera
indica, Sindora Wallichii, and Sterculia carthagenensis.
Each of these presents some peculiar features, which are
discussed under the appropriate headings below. The most
striking may be mentioned here. Cassia nodosa showed a
very regular period of 11:7 months for several years, and
then behaved very irregularly. Ficus caulocarpa is the
most irregular of all, and I cannot trace any connection
between its behaviour and climatic change. Hevea trees
mostly lose some leaves, and sometimes all, when there is
dry weather, but the single tree with a regular period of
13:3 months is a notable exception. Sterculia carthaginen-
Sis 1S a very interesting case, discussed below in some detail.

Vol. XI. (1940).

126

It has been clearly shown by Schweizer’s experiments
that Hevea leaves have a minimum life of about 5 months;
and it is probable that most other irregular trees have
similarly a minimum period for which the leaves must be
retained, the leaves being liable to fall in any dry weather
after they have attained this minimum. There is probably
always considerable variation from tree to tree in this
class, according to the individual nature of the trees, their
position (soil, exposure etc.) and previous history.

Flowering of Deciduous and Evergreen Trees

The flowering of deciduous trees always takes place
at a definite point in the leaf-cycle; on the old wood before
new leaf growth (Hrythrina indica), simultaneously with
the growth of new leaf shoots (Cratoxylon formosum,
Cassia fistula), or at the end of the new leafy shoots
(Lagerstroemia, Peltophorum). Flowering may be much or
little, or even often absent in some cases (Homalium,
Terminahia subspathulata). The extent of flowering
depends no doubt on the climatic conditions at the time the
flower buds are formed. It is notable that Delonix trees
flowering in Singapore at some seasons are much fuller of
flowers than those flowering at other seasons. In some
species, on the other hand, flowering is very much the same
whatever the season (e.g. Lagerstroemia) and there is
probably in all cases some difference in response according
to innate differences from tree to tree. But whatever
these differences in response, flowering always occurs at a
definite point in the leaf-cycle, and cannot occur at any
other time. From this it follows that, with trees having
a leaf-period of less than 12 months, there is no definite
flowering season, and little response to climatic change, but
a sporadic flowering of individual trees throughout the year.

The flowering of evergreen trees, on the other hand, is
usually not dependent on the leaf-cycle, and bears no
relation to new leaf growth. Flowering is dependent on
dry weather, but the intensity of drought needed to produce
flowering, and the time which elapses between the onset
of dry weather and the opening of flowers, vary greatly
from species to species. Examples are given under Saraca
taipingensis and Fagraea fragrans. Some evergreen trees
however flower at the end of new leaf-growth (e.g. Hugenia
grandis, Cinnamomum iners). 3

Evergreen trees produce new leaf-growth at intervals
of a few months (sometimes as much as a year or more),
the intervals apparently much less regular than the leaf-
periods of most deciduous trees. The old leaves usually
do not fall as soon as the new have grown, but remain

Gardens Bulletin, S.S.

a

127

on the tree for some time afterwards, their total life-period
being often very long. There are however cases rather
intermediate between the normal deciduous and evergreen
types, such as Dyera costulata and Eugenia grandis (see
my 1931 paper, p. 191). Probably in most cases new leaf
growth of evergreen trees is dependent on some climatic
stimulus, often rainy weather; but little information on
this point is available. Some data will be found below
under Mangifera foetida and Saraca taipingensis.

Statistical significance of the records

In all cases the mean leaf-period has been calculated,
and where the number of observations is sufficient, the
standard deviation, which is the usual measure of variabi-
lity. This is calculated by taking the difference of each
leaf-period from the mean, squaring these differences,
dividing their sum by n-1 (n being the number of periods)
and taking the square root of the result. If one date of
a series has been missed, the double period is divided, and
two equal periods entered for calculation of differences from
the mean, but the sum of the squares is divided by n—2
instead of n-—1.

As noted above, it appears that standard deviation,
when expressed as a percentage of the mean period, gives
a measure of the susceptibility of a tree to climatic
influences. It is notable that in two instances (Cassia
fistula and Salmalia malabarica) a more vigorous tree has
a smaller deviation than a less vigorous tree of the same
species, indicating that a more vigorous tree is less
susceptible to climatic influence. It may be that the more
vigorous tree, having a larger and more widely spreading
root-system, has access to a larger amount of soil moisture,
and is correspondingly less affected by a period of dry
weather.

I am grateful to Dr. E. J. MASKELL for advice
concerning the data presented in this paper, and for making
a statistical comparison of some of them, to assess their
Significance. The results of this treatment of a few of the
records, summarized in the following paragraphs, will! give
a general idea of the significance of the records as a whole.

Taking two trees which have leaf-periods of more than
a year, namely Hevea braziliensis and Sterculia macrophylla,
calculation indicates that the probability in either case that
the figures presented show chance behaviour of trees whose
true period is 12 months, is less than -01. That is, the
probability that these trees have a natural period of more
than 12 months is very high.

Vol. XI. (1940).

128

Taking pairs of trees of the same species which have
shown different mean periods, the following results were
found. In the case of Salmalia malabarica, the probability
that the differences observed between the two trees are a
chance result is less than -02; this indicates that the
differences observed are fully significant. In the case of
the two trees of Cedrela Glaziovii, the probability is greater
than -02, but less than -05; this also indicates a significant
difference. In the case of Ficus variegata, however,
comparing trees A and B, A and C, the differences are found
not to be significant. In these trees, the mean periods are
near together, and the deviations rather high, so that a
longer period of observation is needed to indicate significant
differences between them.

It should be noted however that though the differences
between a pair of trees such as the two Salmalias may be
significant, this does not necessarily indicate that the whole
difference is due to the individual natures of the trees. A
tree with good soil and other favourable environmental
factors may be more vigorous than a tree less favoured, and
the two might on that account show some difference of
leaf-period. On this point, we have no evidence. All we
can say is that in an identical climate two trees of the same
species may show a significant difference of behaviour, part
of this being presumably due to innate differences between
the two trees and part to differences of soil conditions and
other topographical factors. In the case of the two Cedrela
trees, the differences of soil and topography appear to be
negligible, but one cannot be entirely sure of this. An
appropriate experiment would be to propagate the two
trees vegetatively, and observe whether the offspring,
planted together in a different place, continued to show
the differences of their parents.

Conclusion

The observations presented in this paper indicate
clearly that the climate of Singapore is so uniform on the
whole that many trees change their leaves at intervals
which have no relation, or little relation, to climatic change.
It is clear however that there is much variation in regularity
of behaviour among such trees, even between different
individuals of the same species. One can find all conditions
from trees with very high constancy of period (the period
bearing no relation to the calendar year) to trees that are
so susceptible that they may renew at least some of their
leaves at any small climatic change.

The constant-period trees appear to be organized on
the principle that their leaves are efficient for a certain

Gardens Bulletin, S.S.

129

period, and are then shed and renewed, whatever the season.
_For these trees, no climatic change experienced in Singapore
is sufficient to make them change their leaves much earlier
or later than their proper period, which is dependent on
the nature of each tree, and perhaps also on its soil
conditions or other local factors. However, trees of these
same species in climates with long dry seasons may change
their leaves at intervals of exactly a year. It is the climatic
uniformity of Singapore which permits such trees to show
their individuality ; in the interplay of internal and external
factors, the internal are here by far the stronger.

At the other extreme are trees which are considerably
affected by such climatic changes as do occur in Singapore.
As such changes are very irregular over a great part of the
year, trees of this kind are irregular in behaviour. They
appear to be organized on the principle that their leaves
have a minimum life; any dry period occurring after the
minimum may cause some leaves to fall, with subsequent
leaf-renewal. There is no sharp distinction between these
trees and those which have constant leaf-periods; the gap
is bridged by trees showing varying degrees of deviation
from regular periods.

_ Another point to observe, deserving of future attention,
is that interference with the roots of trees may completely
change the rhythm of their leafy shoots. It seems likely
that stimulation of roots at a sufficient interval after the
last new leaf-growth will stimulate new leaf-growth,
followed by fall. of the old leaves; but that the reaction
will depend on the age of the last leaf-growth, much in the
same way that Hevea leaves have a minimum life. But
different trees will doubtless behave in different ways, and
much of interest may emerge from watching their reactions.

Evergreen trees differ from deciduous trees in not
losing their old leaves when new ones are produced; the
growth of new leaves is not a consequence of the loss of
the old, and must be due to some other stimulus. Observa-
tions on the behaviour of evergreen trees are few (see
Mangifera foetida, in this paper), but I think they also will
be found to vary in their response to climatic stimulus, It
is clear that some (e.g. Saraca) are stimulated by wet
weather to produce new leaves, and that such production
comes at irregular intervals. The length of such intervals
and their dependence or otherwise on climatic stimulus
probably vary as much from species to species as in the
case of deciduous trees. A further important point of
difference between evergreen and deciduous trees is that
in the former flowering is usually independent of vegetative
growth, whereas in the latter flowering always comes at
a definite point in the leaf-cycle.

Vol. XI. (1940).

130

Records of Individual Trees*

Adenanthera pavonina L.

The tree reported on in 1931 has been under continuous
observation since that time, and has maintained an average
leaf-period of 7-25 months, with a standard deviation of
‘74 month over 16 periods. The longest period was 8-4
months and the shortest 6:2; both these occurred in the
records formerly reported. Leaf-renewal has occurred in
every month of the year except February, June and
November.

Leaf-renewal after the 84 month period occurred
towards the end of a 2-month drought; if this prolonged
dry weather was the cause of deferring leaf-renewal, it is
difficult to see why leaf-renewal after the short period of
6-2 months took place at the end of one month’s dry weather.
In both cases the new leaves developed before any consider-
able rainfall had broken the drought. A subsequent record
of leaf-renewal on 9th July, 1938 gives an average of about
7-5 months for two leaf-periods after 11th April, 1937.

It should be noted that this species is said to be annual
in its behaviour at Peradeniya, losing its leaves towards
the end of the year. On a recent visit which I paid to
Colombo in October, I noticed that all trees were in old
leaf with ripe fruits, and the same appeared to be the case
at Bombay a few days later. See however remarks under
Peltophorum ferrugineum.

The following table gives dates of leaf-renewal, with
the length of each leaf-period, and also the standard
deviation as calculated from these figures.

Months Months

27. 8.27 — 19 Te 4e > 8-2
22. BiZs 6:8 6. 9357 4 1g 7-7
21.10.28 7-0 April 1934 Te
28. 4.29 6-2 1.12.34 Teo+

8.12.29 7-3 15. hies 7-5
21. 8.30 8-4 1.°3.06 8-5
10) 3.58 6-6 10.10.36 7-3

4.10.31 6-8 14) 437 6-0

3.\tuae 7-1

16 Periods.

Mean period:. 7-25 months.
Extremes: 8-4 and 6-2 months.
Standard deviation: -74 month or 10-2%.

Anisoptera megistocarpa van Sl.
A large tree of this species in the Gardens Jungle

has been under observation since 1928; it has shown
considerable variation in behaviour, with long leaf-periods

ij Tn the records which follow, dates of leaf-renewal are the
dates on which leaf-buds were seen to be starting new growth.

Gardens Bulletin, S.S.

131

throughout. The tree is slowly deciduous during many
weeks, new growth appearing when it is almost bare of
leaves. It has had six periods varying between 12 and
16:5 months, and two much longer periods of 23:8 and 282
months. The crown of the tree is clearly visible from my
house, though distant, and I am confident that these two
Jong periods represent intervals between new growth
following bare phases. It is possible however that (as with
the tree of Dyera) new growth may have occurred without
a preceding leaf-fall during these long periods; I might
have missed such new growth, as there is little colour
change between new and old foliage. If we assume that the
long periods are in fact double periods (so far as new
growth is concerned), we find that the standard deviation
is 12%, which is no greater than in a number of other trees;
we them as single periods, the standard deviation is very
igh.

In any event, more observation is needed to understand
the behaviour of this tree. One cannot see any relation
between the dates of leaf-renewal and rainfall. Leaf-
renewal in March, 1931, March, 1932 and April, 1933 took
place after rather dry weather in February (average
115 mm.) which was absent in 1935; but if a dry February
causes leaf-renewal, it is curious that the much drier month
of February, 1936 (43 mm.) had not a more immediate
effect. The dates of new leaf-growth, and the intervals in
months between them, were as follows.

Months Months
Bey BAB), \ 5): aE — Sa + Oo 140
he. ae ves EDS 27. SG iS. chy eS
aa 7 a rn PSUS yeas Beige os"
ee Sete ae porate. | SO wes mee
1, 430 sh LEG

Reckoning the periods 23-8 and 28-2 as double periods
(ie. as 2 & 11:9 and 2 * 14-1 months respectively), we
have the following :—

10 periods; mean period 13-8 months,
extremes 12-0 and 16-5 months.
Standard deviation: 1-65 months or 12-0%.

Caesalpinia ferrea Mart.

A single tree of this south American species, apparently
the only mature tree of its kind in Singapore, grows in the
Botanic Gardens. It has made vigorous growth, and is
very handsome, but flowers poorly and sets very few fruits.
The tree is deciduous, its leaf-fall taking place slowly over
a period of several months. Before all the leaves are gone,
new growth appears on isolated twigs (as in Cassia
nodosa), but general new leaf growth, which is very rapid
when it does occur, is deferred until practically all old

Vol. XI. (1940).

132

leaves are gone. The dates given are for the occurrence
of general new leaf growth, and are therefore less precise
than those given for some other trees.

An isolated observation of this tree was made in 1929;
continuous records date from 1934. The gap in observa-
tions is exactly four times the mean period for 1934-1940.
For the years 1934-1940, the standard deviation is below
10%, so it is fairly clear that this is a tree with a period
of more than 12 months. A few seedlings have now been
obtained from it, and their development will be watched
with interest. The following are the observed dates of
general leaf-renewal.

Months Months
25.8.29 OS dR Ae ee 26.9.37 ~ eS
Oe TA ve 4x 14.2 1°1:360 Bee
19.6356 X Ur Toa Q2AAQ 2 Bz
1972365). 40 12.8

For the years 1934-1940 (5 periods) we have:—
Mean period: 14-2 months.
Standard deviation: 1-2 months or 8-5%.

Canarium rufum Benn.

This local tree has been under observation only since
1935. It appears to have a 12—month period. It is on the
edge of the Gardens jungle, growing under good conditions.
Leaf-fall is rather sudden, and leaf-renewal takes place as
soon as the tree is bare. New leaf buds have opened on the
following dates.

Months
1.5.35 ooo Mean leaf-period: 12-0 months.
20.5.36 12-6 Standard deviation: -65 month
4.6.37 ey oe or 5-4%.
20.4.39 ve a es Mean date: 9 May.
30.4.40 te it oe

Cassia fistula L.

The two trees mentioned in 1931 have been under
observation since that time, and also a third tree. The
two former trees have latterly both shown signs of lack of
vigour, but the third tree is very vigorous and has flowered
heavily on all occasions. As noted previously, it is not
easy to fix an exact date for leaf-renewal of trees of
this species; the dates are therefore subject to a greater
margin of error than most of the other observations
recorded in this paper. The tree B showed such a slow
change that no observations on it were made after 1935.
A fourth tree lately brought under observation near the
others has had one period of 10-6 months.

Two exceptionally large and vigorous trees at Govern-
ment House, Singapore, have renewed their leaves and
flowered in May or June in the following years: 1928, 1931,

Gardens Bulletin, SS.

133

1938, 19389. It seems likely that these trees have a period
very nearly of 12 months. It is evident therefore that
observations on further trees are desirable.

The dates of leaf-renewal of the three trees for which
the longest observations are available, with other data, are
as follows:

A

B C
Months Months Months
1B RY By ae — Sie ys — 0 Bel SA CO —
1d 9.28 9-0 Hoa. 3 10-2 28.10.28 10-6
16:29 8-7 24.11.29 10-4 re aD as a aia 18
1. 3.30 9-0 10. 9.30 9-6 1.12351 9-7
1.12.30 9-0 2050.04 8-5 OT. B32 8-4
29. 8.31 8-9 Boe, 11-5 28. 6.33 10-0
8... D.02 8-3 1 00. leo Mar '955 2 x 10-5
1: 3.38 9-8 10. 3.385 2 «K 10-9 10. 1.36 10-
29. 3.05 3 X 8-3 Mean period: 10-5 4.10.36 8-8
26. 2.36 10-9 months. 28. 6.37 8-7
early Standard deviation: May 1938 10.
Jan. 1937 10-5 1-26 months or 12-:0%. ly a! aS
Oct-ioeT .. 9-5 Mean period: 9-6
5, Aoo 2x 90 months.
Mean period: 9-1 Extremes: 8-4 and
months. 10-6 months.
Extremes: 8-3 and Standard deviation:
10-9 months. ‘84 month or 8-7%.

Standard deviation:
-93 month or 10-2%.

Cassia nodosa Buch.-Ham.

The tree mentioned in 1931 continued its leaf-renewals
after slow and prolonged leaf-fall in the early part of the
year, tending to become earlier each year up to 1935, for
which the date recorded is 10 March. The dates given are
for general new growth of leaf buds, and are usually about
a month later than the first new buds which develop here
and there on small branches in advance of the remainder
of the tree. The standard deviation for this period is
remarkably low.

Then in December, 1935, possibiy as a result of the
rather dry period from July to October, about 34, of the
branches on the tree produced new leaves. The remaining
branches did not bear new leaves until March, 1936. The
branches which had new growth in December, 1935 renewed
their leaves again in July, 1936. In April, 1937 all branches
renewed their leaves.

This tree is said to have been produced from a seed
from Pahang, where the species is common, and leaf-renewal
with flowering is general about May. It appears likely that
the more uniform climate of Singapore has allowed our
tree to show a tendency to instability or to sensitiveness
to climatic changes. A number of young trees of this
species have lately been planted, and their behaviour will

Vol. XLo\(1940).

134

be watched with interest. Data for the nerd 1928-1935
are given below.

Months

28.5.28 iy, y —

15.5.29 oi Seta bs |

15.4.30 i Sa aba Mean period: 11-7 months.
12.4.31 x2 ee gee Extremes: 11-0 and 12-1 months.
16.4.32 » epee bf Standard deviation: -4 month

9.4.33 i MLS or 3-4%.
20.3.34 ahs joes [aa
10.38.35 11-7

Cedrela Glaziovii C. DC.

The trees of this species were wrongly ascribed to
C. toona in my 1931 paper. These two trees have been
under observation since that time and have shown a
remarkable difference in behaviour. Both are tall vigorous
trees growing near together under good soil conditions.
They behave alike in their slow leaf-fall, from the top
downwards, and the appearance of new leaves first on the
top branches after the whole tree is bare. One tree has
had a mean leaf-period of 12:5 months, advancing its time
of leaf-fall from July to December during the period under
review, with a very small standard deviation. For the first
few years it appeared likely that this tree had a period
of exactly 12 months, but it now seems more likely that
there is a significant departure from this figure.

The other tree has shown less regularity (though its
coefficient of variation is only 12%), and has throughout
had leaf-periods decidedly more than 12 months, with a
mean of 14:0. Most of the periods have been 13 to 14
months, but one was 17:7. This long period included the
whole of 1932, the driest year in the period 1927-88. For
other cases of difference of behaviour between trees of the
same species, see Ficus variegata, Hevea and Peltophorum.

The following table shows how one tree had ten leaf-
periods and the other only nine, in the same period of years.

Months Months
8. 7.28 Ji — Aug. 192860 a2 ae ae
4. 8.29 12-9 Sept. 1929... v5 Nee
30. 7.30 11-9 end Sept.1930 ... ree
1. 6.51 12-0 ISIB Sk) Mee Bara: Ej
9. 8.32 12-3 ee es eer a 3
13) 8.338 12-1 July 19384 .. Mee
10. 8.34 11-9 12. se. Joh
10. 8.35 12-0 Oat. 1946" 0"): j\o or paige
12. OG Ds... .. 281. Yond Wee ese oe bus! ee
a he a adie .4 eel se ike. Jeet
17A0.36y eS. ee oa 46. 2.40 55 . st 966
16.12.39) Lily fi) ee 10 periods: mean period 14-0
11 periods: mean period 12-5 months,
months. Extremes: 12-5 and 17-7 months.
Extremes: 11-9and14-0months. Standard deviation: 1-7 months
Standard deviation: -70 month — or 12-1%.
or 56%.

Gardens Bulletin, S.S.

Ceiba pentandra (L.) var. caribaea (DC) Bakh.

A tree of this variety (the West Indian Kapok),
received as a stump from Buitenzorg and planted in October
1933, has shown a six-month leaf-period since 1936. The
new leaves appear before the old have quite fallen; the
change was very gradual in September, 1938 and March,
1939. Observed dates (incomplete) of new leaf production
were:

8.4.36 — Sept. 1938
27.10.36 March 1939
25. 4.37 March 1940
15.10.37 Sept. 1940

Another tree of the same origin, planted at another place
but at the same time, renewed its leaves on 28th December,
1936, 1st June, 1937, 9th February, 1939 and 26th July,
1939. It has thus shown a similar six-month period, but
has changed its leaves at different times from the first tree.
The reason is probably that this tree was cultivated and
manured some time after planting (the first tree has never
had such treatment), but the date of this treatment was not
recorded. Fora similar case, see Kigelia.

A seedling raised from seed from the Gold Coast has
also shown a six-month period. Ceiba pentandra in Java
has a twelve-month period. Possibly these young trees will
lengthen their period as they grow older.

Couroupita guianensis Aubl.

An old tree of this species was reported as having
three six-month periods in 1929-1931. A record over 16
periods is here presented, the average being still exactly six
months and the standard deviation identical with that of
the tree of the local species Peltophorum pterocarpum.
Leaf-fall is very rapid, and also leaf-renewal; the dates
given are therefore closely comparable. The tree is only
bare for a day or two.

A young tree, planted in 1925 in a thicket among other
trees which restricted its growth, did not behave at all
regularly at first, but it appears to be settling down to a
more regular rhythm as it grows older; it now has its
crown clear of other trees and is growing more vigorously,
but has not yet begun to flower. It does not change its
leaves at the same time as the old tree.

In De Tropische Natuur, XXIII (1934), p. 244, it is
recorded that the Couroupita trees at Buitenzorg change
their leaves in March-April and September—October, thus

Vol. XI. (1940).

136

agreeing with the old Singapore tree. Mr. Ahmadi,
Superintendent, Municipal Gardens, Bombay, informs me
that the Couroupita trees in the Victoria Gardens, Bombay,
change their leaves regularly in March and October. This
regularity of behaviour in such different climates is
remarkable.

The observed dates of leaf-renewal of the young tree
in Singapore were as follows:

Months Months
19 A BP ay sh — 9, T.36 7-9
Te Assan cies ye: 8-5 2). 1.37 6-5
yay mas ese Nee oe Cat 2). Vo oe ea ease st 5-9
2m. DBD) es a x F-9 7 1b B39: nee 4 << 62-2
I3At 30+. ‘3 5-7 22. AO. 2. o 6-2
18. 840° °.. a 5-9
The data for the old Singapore tree are as follows.
Months Months
15. 9.29 — 8. 9.33 5-6
23. 3.30 6-3 19. 3.34 6-4
16. 9.30 5:8 1. 9.34 5-6
22... ocan 6-1 12. 3.35 6-2
18. 9.31 5-9 15. 9.35 6-1
5, 3.32 5-6 25. 3.36 6-3
10, 9.82 6-2 5.10.86 6-4
19. 3.33 6-3 17. 4.37 6-4
12. 9.37 4-8
16 periods.

Mean period: 6-0 months.

Extremes: 4-8 and 6-4 months.

Standard deviation: -43 month or 7-2%.
Mean dates: 15 March and 20 September.

Cratoxylon formosum (Jack) Dyer.

Two trees of this species, in different parts of the
Botanic Gardens, Singapore, have been kept under observa-
tion from 1927 onwards. As shown below, their periods
have been closely similar, and on the whole very regular,
averaging about 9 months, but the trees are quite apart in
their dates of leaf-renewal. The behaviour of these trees
is in my opinion one of the most remarkable examples in
the series of observations recorded in this paper.

A tree of this species in the Waterfall Gardens, Penang,
has been noted by me during visits in January and February
1931, 1932, 1933, 1935, 1936, 1987 and 1939. In each of
these years the tree has renewed its leaves about the end
of January (2.e. in the dry season). It is possible of course
that this species is one that varies considerably in its leaf-
periods from tree to tree, but the difference of behaviour
in the seasonal climate of Penang seems to me likely to be
significant.

Gardens Bulletin, SS.

137

The dates of leaf renewal of the Singapore trees have
been as follows.

A . B
Months Months

1, 9.27 — 6. 2.28 _

10. 6.28 9.3 21.10.28 8-5
WT, aun 9-2 30. 6.29 8-3
23. 2.30 11-2 7. 4.30 9-2
27. 9.50 7-2 17.08.51 9-2

1 8:31 10-1 11.10.31 9-0

1. 4.32 8-0 3. 8.32 9-7
11.12.32 8-3 16. 4.33 8-4
18. 9.33 9-1 1. 2.34 9-5
end June 1934 9.4 30.12.34 10-9
25. 3.35 9-0 12. 9.35 Galle Ore
16.11.35 7-7 26. 5.36 Lh6, Bb
20. 9.36 10-1 8. 3.37 Hie nese
a Pf 10-0 2.10.38 ee Se

2 Re | gia yang 3, | gs ae OES as

Mean period: 9-1 months. Mean period: 9-1 months.

Extremes: 7-2 and 11-2 months. Extremes: 8-3 and 10-9 months.
Standard deviation: 1:09 month Standard deviation: -74 month
or 12-0%. or 81%.

Delonix regia (Boj.) Raf. (Poinciana regia Boj.).

An old tree of this species was under observation
between 1928 and 1936, over eleven leaf-periods. As noted
in my earlier paper, leaf-renewal is slow, and it is
impossible to give accurate dates, or dates which are strictly
comparable from one leaf-period to another. It is however
evident that the mean period of the tree is in the neighbour-
hood of 9 months. After a lapse of 3 years without
observations the tree was noted as in new leaf in June,
1939, which agrees with the last observations of July, 1936
(4 periods of 8-8 months occupying 2 years 11 months).

Trees of Delonix regia are commonly planted in
Singapore, but even those planted simultaneously do not
always. keep in step together. It takes them a few years
to develop a regular leaf rhythm, and this may perhaps
change gradually with the age of the tree. Observations
on this point, made on a group of young trees planted
together, would be of great interest. It would be necessary
however to treat all trees alike as regards cultivation (see

Kigelia).

As noted in my earlier paper, some Delonix trees may
be seen in flower at any time of the year in Singapore, but
it appears that at some seasons flowering is more abundant
than at others. Possibly the climatic conditions at the time
when flower buds are produced may be responsible for this;
i.e. it may be that trees which are bare during wet weather
produce fewer flower-buds than those which are bare during
dry weather. There is however no proof of this. At

Vol. XI. (1940).

138

Penang a Delonix tree renewed its leaves in January, 1931,
1932 and 1933. Further Penang records are desirable.
The Singapore tree has renewed its leaves as follows.

Months Months
Bay ie. |), ve — Jan. 1933 10
DU; ue) (Fs sit aby De Aug. 1933... ae
ye Wy + er oo.) Of Mar Ap. igs) «22 ate
LO AACAD S42 no LOG Jam. 1985. 2, Or ee
VC ey ea -. 77 Sept.-Oct. 1935 8
March 1932 8- July 1936 9

Mean period: 8.8. month.
Standard deviation: -94 month or 10-7%.

Dyera costulata Hook. fil.

As reported in 1931, two trees of this species have
been under observation. Leaf-renewal tends to be at
rather irregular intervals, somewhat as in Hevea. On the
big tree, which has two main trunks, leaves are sometimes
renewed on a few branches only, or on one main branch
system, between general renewals. New leaf growth
occurring in the wet season (October, November) has in
all cases preceded fall of the old leaves; at other times the
old leaves have fallen before the new leaves began to
unfold.

The big tree has had complete leaf-changes at intervals
of from 3 to 11 months, sometimes with partial changes in
between. Out of the 18 such periods, 9 have been about 6
months, 4 considerably less than 6, and 5 considerably more.
A comparison with rainfall records indicates that there is
some tendency for leaf-fall to follow dry weather; but this
does not invariably occur. As above remarked, new leaves
sometimes appear before the old leaves have fallen; in four
such cases (all those noted) the period since the preceding
leaf-fall was almost exactly six months. In other cases,
under apparently similar conditions, no leaf-renewal has
been recorded, and the old leaves appear to have persisted
for 10 or 11 months (1931-382, 1934-35 and 1938-389).
Where new leaves appear before leaf-fall, the old leaves are
always shed within about two weeks following.

The smaller tree under observation (with one trunk)
has changed its leaves rather regularly each year about
March—June. In 8 years (1928, 1929, 19382) it has also
changed again about six months later, but so far as my
observations go, this has not occurred in other years.
Taking only the changes in March to June, the mean period
is 123 months, and the standard deviation -°84 month
(63%). This tree also occasionally produces new leaves
on a few branches without a general renewal. When new
leaf growth was general in November, 1932 it was not
preceded by leaf fall.

Gardens Bulletin, SS.

139

Dyera trees are fairly frequent in the Gardens Jungle,
and they tend to change their leaves at the same time,
though they do not always behave alike. The two trees
under observation have changed their leaves simultaneously
(or approximately so) on the majority of occasions, but
on others have been curiously diverse in behaviour. The
following are dates of general changes only.

Large Tree Small Tree
Months Months

6. 9.27 a 17. 3.28 —

17. 3.28 6-4 11. 9.28 5-7

14, 6.28 2-9 14, 4.29 7-1 12-9
6. 4.29 9.7 3.11.29 6-7
20.10.29 6-5 25. 5.30 6-7 13-3
25. 5.30 7-2 1. 6.31 12-2
1.10.30 4.2 22. Doe 11-7

2. 6.31 8-0 3.11.32 5-4

12. 4.32 10-3 21. 4.33 5-6 11-0
20.10.32 6-3 21. 5.34 13-0

7. 8.33 9-5 25. 4.35 11-1

21. 5.34 9-5 4. 5.36 12-3

3. 5.35 11-4 14. 5.37 12-3
9.11.35 6-2 23. 6.38 13-3

31. 5.36 6-7

3.12.36 6-1

3. 6.37 6-0

25. 9.37 3-7

23. 6.38 8-9

Erythrina lithosperma Miq.

Trees of this species in Singapore are remarkably
constant in flowering in December and January, towards
the end of the wet season. The trees are not deciduous,
but this flowering appears to be connected with new
vegetative growth which takes place early in the wet season,
the onset of which is rather regular about October. I have
no full record of the behaviour of one of these trees
throughout the year, and only enter this observation here
as the flowering of Erythrina lithosperma is one of the most
regular seasonal phenomena in Singapore. The trees are
not very showy nor abundant.

Fagraea fragrans Roxb.

As reported in this Bulletin, Vol. IX pp. 73-78 (1935),
trees of this species, which are abundant in Singapore,
flower gregariously about May every year. The trees are
evergreen, and their vegetative growth is practically
continuous, but their flowering is such a regular pheno-
menon that it should be mentioned here. The height of
the flowering period, which lasts about 2-3 weeks, occurs

Vol. XI. (1940).

140

approximately four months after the end of the wet season,
when the first dry period occurs. A week of dry weather,
with no more than about -1 inch of rain, following much
wetter weather, appears to stimulate the formation of
flower buds. Fruits ripen about 314 months after flowering.
There tends to be a small flowering about October—
November, but this is irregular, and a few flowers may
occur on isolated trees also at other times, but no general
flowering of all trees.

Ficus caulocarpa Mia.

The tree reported upon in 1931 (under the name
F’. glabella) has been under observation since that time and
has renewed its leaves three times each year, except 1935
and perhaps 1938, in which it changed four times. The
causes of the length of the individual periods are doubtless
complex, but it is fairly clear in a few instances that
distribution of rainfall has had an influence. 1931 was a
very wet year, and the long leaf-period was longer than
usual; this was probably the cause of an unusually late
first change in 1932. In 1934 December was unusually dry,
and this probably accelerated the first change in 1935; but
why the next period should have been so very short is not
at all clear. In several years there is a definite succession
of periods, approximately to 3, 4 and 5 months, but in
other years this succession is much less distinct. The mean
dates of leaf renewal (1928-37), are 12 February, 9 June
and 8 November (giving intervals of 96, 117 and 152 days),
the standard deviations being 19, 22 and 17 days; these
deviations are distinctly high as compared with those shown
by most of the deciduous trees here reported.

The amount of rain falling during a leaf-period bears
no relation to the length of the period; I have also compared
the rainfall from 14 days before the end of one period
to 14 days before the end of the next, but this also shows
no correlation with length of period.

Another younger tree, about a mile away, was under
observation for a time, but probably not all changes were
observed. Leaf-renewal dates were noted for this tree on
7 occasions between November, 1934 and June, 1937. The
differences in time between the nearest leaf-renewals of the
Gardens tree were 13, 23, 33, 32, 15, 138 and 53 days.

A young plant of the same species, which had started
epiphytic growth on an old tree in the Gardens, changed
its leaves once simultaneously with the large tree, but on
two other occasions at times 40 and 41 days different.

It thus appears probable that the factors controlling
leaf-fall and leaf-renewal in this species are complex and

Gardens Bulletin, S.S.

141

vary from tree to tree. They are certainly not simply
related to climatic conditions. It is possible that fruiting
may influence the length of leaf-period ; it seems to be rather
infrequent, and the small figs are not easily observed on
the high branches of the Gardens tree. Unfortunately I
have no records of fruiting.

The dates of leaf-renewal observed are as follows :—

1927 bo Oct.
ASZS os as qree) oan: 3 June 4 Nov.
(iy! ao 15 eee DED: 17 May 20 Oct.
(0S) ae eee. can. 8 June 11 Nov.
Pook bse -«» LO Feb. 28 June 2 Dec.
£952 —(i« <2 jioomaren, 1 July 9 Nov.
1s ee .. 16 Feb. 3 July 29 Oct.
1OS4. ee. oc 6 a eb: 19 May 2 Nov.
1 Se a» 20 dan, 7 April 14 July 5 Dec.
c/s ewe .. 24 March 10 July 30 Nov.
5 ae «kd Heb. 27 May 20 Oct.
Eas bes. .. unrecorded 20 Aug. 31 Dec.
LCy ee s«oi24 April, 27 June 2 Sept.

LS a (nae a) 13 June

Ficus variegata Bl.

This is a common species in Singapore, and seedlings
are abundant in the Botanic Gardens. A tree was men-
tioned in my 1931 paper under the name F. polysyce.

| This tree appeared at first to have a period of six
months, but later proved to have a rather longer period,
showing an average of 6:5 months, with standard deviation
of -84 or 12:9%. Its most peculiar behaviour was in 19382,
when, having renewed its leaves in March, it persisted until
November with the next crop of leaves, and then changed
again in June 19838, the only leaf-renewal in that year.

A second tree, observed (with two gaps) from 1929
to 1939 was rather less regular, as shown below. Two other
trees, observed only from 1932 and 1933, have shown
greater regularity, with a less tendency so far to exceed
a six months average. Mr. Corner has pointed out to me
that the first mentioned tree is a male tree (figs contain
male and gall flowers only) and one of the last two trees
is female.

It was mentioned in 1931 that the male tree bore six
crops of fruits ina year. The females trees under observa-
tion bear less frequent fruit crops (3 to 4 in a year), but
in neither case could any relation between fruiting and leaf
change be noted. The trees are cauliflorous, belonging to
the subgenus Neomorphe. (For further information on
this species, see Corner’s paper in Journ. Malayan Branch

Vol. XI. (1940).

142

R. Asiatic Soc. XI: 48-52. 1933). The dates of leaf-renewal
of the four trees and their leaf-periods in months are as
follows :—

A B
Months Months
12. 8.28 — 25. 8.29 ig —
6. 1.29 4-8 es eS we 6-2
14. 7.29 6-3 7. 431 a i Ge
7. 250 6-7 Oct. 1931 ef 6:
5. 8.30 6-0 22. 5.82 ae 7-
py Ae S| 6-1 23.10.32 ; 5-0
20. 8.31 6-4 Feb. 1933 ae 4.
6.43.32 6:5 Zi; Boe te 6:
Nov. 1932 8: AS a. ae 6-2
V1. G35 7: 18. 9.34 - 6-5
26. 1.34 7-5 Das Wes 5 a 6-4
6. 9.34 7-3 Nov. 1935 ey 7:
15. 3.35 6:3 June 1936 ox 7-
Oct. 1935 7: 20.11.36 : 5:
April 1936 6- 1 6.37 ory 6-4
30.10.36 6: 13.12.50 3 ee
TA SST Ue ee 6-4 June 1939 ; 6-
Pe EO serer iit’: 3.) eT
Ge PMS Ne 12 re des 7-0
C D
Months Months
ALUN es a ee ae — 9. Beers: Ts —
DOM Dae hike OK One 14,.43,355.0 3 as 5-2
LBs soaker ye 6:5 Feb: 1934" 324 2° x 6S
22.4535 5-2 24. BB4,.¥o4 bh 6:
19: ao 6-0 18. 2.35 6:
Tl, 2:56 6-8 6. 3.5) 5-6
Sept. 1936 7. pa ae es 7-5
March 1937 6- Sept. 1936 6-
end Aug. 1937 5. 80.) Sol 6-
end Sept. 1937 __—.... sane a
Py: eee (rake 2 oe ee
2 a Pa, 5-4

Mean periods: A, 6-5; B, 6-4; C, 6-1; D, 6-1 months.

Standard deviations: A, 84; B, 85; C, 65; D, -62 month.

Coefficients of variation: A, 12. 9% : B, 13.3% ; C, 10-7%;
D, 101%.

Heritiera elata Ridl.

A record of a leaf-period of nearly two years for
a tree of this species was reported in my former paper.
Subsequently there was a gap of four years in observations,
and then three more periods, successively of 17, 22 and
20 months were recorded. The tree is usually not bare
before new leaf-growth occurs; the new growth is conspicu-
ous because of its pale colour. Flowering occurs on the
old wood, usually during the period of new growth, but
may perhaps (as with H. macrophylla) occur at other times.
The tree is a very tall one, remaining from primitive forest.
A young tree, about 100 yards away, was observed to make
new growth at the same time as the old tree in 1940.

Gardens Bulletin, S.S.

143

The periods seem rather unequal in length, but the
standard deviation is not excessive as compared with other ©
trees. All recorded cases of leaf-renewal have occurred
after unusually dry periods. This tree is probably one
with a naturally long leaf-period (though the period is less
than that of H. macrophylla) which is rather much
influenced by climatic change occurring towards the end
of its normal leaf-period. Observations of leaf-renewal are
as follows.

Months Months
ry pees a. in — March 19385 .... te rau
eR Of aan ae pay) Sees Auge. 1986'S, wide ae

Pewee, 3% ory eee
Se Pao i) L oe

Mean period: 20-5 months.
Standard deviation: 2-6 months or 12-:7%.

Heritiera macrophylla Wall.

A small but quite vigorous tree of this species, which
is native in Burma has been under observation since March
1929. It has an exceptionally long leaf-period, more than
2144 years on four successive occasions. Leaf-fall is
prolonged over perhaps two months or more, at first slow,
then progressively more rapid. Leaf-renewal is quite rapid,
and is complete in a few weeks. In one case leaf-renewal
took place on one branch after another over a period of six
months instead of on all branches together.

It is notable that the Malayan species H. elata also has
a long leaf-period. It would be interesting to have records
of the widely distributed coastal tree H. littoralis; also to
know something about the behaviour of H. macrophylla in
its native climate.

The dates of leaf-renewal of the tree of H. macrophylla,
and lengths of leaf-periods, are as follows.

Months
5 (Re i Me << —
tee Ae “f 32
14. 7.384 to Jan. 1935 ee 31 (to 37 months)
Per SST). \ ss nt a? 31
cs Oa oo oe 33-4

Hevea braziliensis Muell.-Arg.

The four trees mentioned in my earlier paper have
been kept under observation since that time. Three of
them illustrate the usual behaviour of Hevea trees in
‘Singapore, namely response to dry weather whenever it
comes, and consequent irregularity of leaf-renewal. The
two larger trees have sometimes changed their leaves
completely at the same time; sometimes a complete change

Vol. XI. (1940).

144

by one tree has been accompanied by a partial change in
the other. The summary, printed below, though incomplete,
gives a sufficient record to illustrate their behaviour.

One of the young budded trees, recorded in the previous
paper as never undergoing a complete leaf change, but
always changing leaves on a branch (or a few branches)
at a time, continued to behave in the same way for some
time afterwards. It was later once observed to change all
its leaves simultaneously, but no full records have been kept.
It may show a greater tendency to complete change as it
grows older.

The other young budded tree, reported in 1931 as
having renewed its leaves completely three times, has
consistently continued this habit, with a period which
appears to be significantly more than 12 months, the mean
for nine periods being 13:3 months, with standard deviation
of only 65%. The leaves of this tree are small, and it may
be a morphologically distinct variety.

It is interesting to compare these observations with
those of Schweizer (Mitt. der Naturforsch. Ges. Bern,
1932, pp. 1-8) at Djambi in Java, where there is a regular
dry season. Leaf change of the Hevea trees at Djambi is
related to the onset of the dry season, but some trees are
always earlier or later than others in leaf-renewal.
Characters of earliness or lateness are shared by all trees
of the same clone. One tree was found to change its leaves
frequently, a branch at a time; after manuring with
Ammonium sulphate, the leaves all fell together, rather late,
at the next dry season, but after two years the tree resumed
its irregular habit.

Schweizer defoliated certain trees (the behaviour of
which had been previously recorded) at various times in
relation to the season of leaf-change. If the trees were
defoliated more than 5 months before normal leaf-change,
they produced new growth and the next change occurred at
the usual time. If defoliation was carried out only 2-3
months before the date of normal leaf-fall, the subsequent
leaves remained on the tree up to 3 months after the usual
date of change. That is, the minimum life of Hevea leaves
appears to be 4—5 months.

In Singapore, dry weather is irregular in its occurrence
(see introductory remarks in this paper) and it seems
probable that any dry period will affect some Hevea trees
which have leaves more than 4—5 months old. If the dry
weather is severe, all the leaves on the tree may fall; if
not severe, only the oldest leaves. As there are few budded

Gardens Bulletin, S.S.

145

trees in Singapore, the behaviour of the trees is very
irregular, and after every dry period some trees may be
seen with autumn colours or bare branches.

For other examples of irregular leaf-change, see
Sterculia carthaginensis and Mangifera indica.

The single regular tree has shown the following
succession of leaf-periods.

Months Months
ed Oy: a — Pe eek) sc -. 150
(a. eo Pe Bes 7) Re > fe) ees
fo +2) koe April 1936... ioe i *
mr 923i": ig. ee 6.38? Va otf Z-
2TA1:32 14-0 hy os eee 347148

Mean period: 13-3 months.
Extremes: 12-2 and 15 months.
Standard deviation: -86 month or 6-5%.

Partial or complete (full) changes of leaves of the
two large irregular trees were observed as follows:

" Tree M Tree W
— ie x) Mow: - part Aug. full
126. “ af March part
74 .. Aug. full Aug. part
| S| a .. April part March full
Sept. part
1930 iF $F March full March full
Us) oe Ls May spart July full
Aug. part
iL sy Ae .. March full unrecorded
Oct. part
1933... <n ee Oe Jan. full
Aug. full
ioe? .. “y unrecorded unrecorded
$9g5.)? \. Le ees * > Sal
Oct. full Oct. full
ee .. Aug. part unrecorded
re -- March full March full

Homalium grandiflorum Benth.

A tree of this species, which is native from Tenasserim
southwards, has been under observation from 1932 to 1939,
ail leaf-changes except one having been observed. It will
be seen that all periods were more than a year, the shortest
being 12-5 months; one period of 17-5 months was however
much longer than the others. It is not obvious why this
very long period should have occurred. 1934 was a wet
year (115 inches), but April, May and June, when leaf-fall
might have been expected, were the driest months. Leaf-
renewal took place after only 12-6 months on 30 December,
1936, though October and November had been wet months.

This tree flowers very rarely, with the new leaves.

Vol. XI. (1940).

146

Only one flowering was observed in the period under review.
The observed dates of leaf-renewal are as follows :—

Months Months
By ees ge 3 ny — Siuliose.. ae aot.) koe
DB, ae ee Bh bes | 18: ‘Zee 2 ea
20.' 9.340). Ao, ID Si Sae <a 0 Se
a A ee Seles

Mean period: 13-7 months.
Standard deviation: 1-9 months, or 13-9%.

Hymenaea courbaril L.

A fine tree of this West Indian species has been under
observation since 1927 and was reported upon in my earlier
paper. It is never bare of leaves, but it has produced new
leaves regularly every year in January or February,
simultaneously with the fall of the old leaves. Flowering
follows, at the end of the new growth, in May or early June,
and the fruits ripen and fall towards the end of the year.
Exact dates are difficult to record with this tree, as new
growth often begins on one branch and gradually spreads;
but the tree is undoubtedly very regular in behaviour.
There are indications that early or late ending of the wet
season influence the time of leaf-renewal. Thus in 1928
wet weather persisted until February and new leaf-growth
was not observed until 19th February. In 1930 new growth
was observed on 12th January, following a dry December.
In 1931 January was wet throughout, and the first new
leaves were seen on 22nd February.

It is evident however that all trees of this species do
not behave alike. In 1931 I reported on a tree that had
flushes of new leaves on various branches in turn through-
out the year, though occasionally renewing its leaves almost
entirely at about the same time as the first tree. A third
tree has since been noticed, which renewed its leaves in
September 1935 and 1936. This tree renewed its leaves at
the same time as the first, in February, 1940.

Kigelia pinnata DC.

A tree of this African species, reported upon in 1931,
has continued very regular in its behaviour, renewing its
leaves rapidly in January or February each year, the mean
date being 22nd January. As will be seen below, the
standard deviation is very small; this is particularly signifi-
cant, as the time of leaf-renewal is more sharply defined
than in some other trees, and so the dates are more strictly
comparable. Leaf-fall is complete within a fortnight, and

Gardens Bulletin, SS.

147

new growth usually begins just before the tree is bare. .
Flowering takes place on the old wood, sometimes just before
leaf-fall, but more usually within two months afterwards.
The flowers (which are open at night) are regularly
visited by the local bats, but fruits are not set very freely.

A very interesting comparison of behaviour is provided
by some young trees, raised from seeds obtained from
Africa. These trees were planted out, as young seedlings,
on 19th October, 1931. They dropped their leaves and made
new growth, in November, 1932, November, 1933 and
November, 1934. Then in March, 1935 they were dug and
manured, and they made new growth in April, dropping the
old leaves soon after. Leaf-fall and new growth subse-
quently took place in April, 1936 and April, 1937. This
appears to indicate that the species tends to keep very
accurately to a 12—month period, but that stimulus to root
activity may induce premature new growth. Schweizer
found that manuring of Hevea trees affected their subse-
quent leaf-fall and leaf renewal (see Hevea braziliensis).

Though the first Kigelia tree has been very regular
in its time of leaf-renewal, it has varied a little with the
time of onset of drier weather after the wet season. Thus
it was very early in January, 1935 after the unusually dry
December, 1934.

Months Months
1; 229°)... bs — oF ay Se" ee 2° 120
Ly ee Ae | COI If 2 aS a Jowag & by
| hee 2 Jo) ae te, eo | eee
7 es i a ee Sy ey ae ae |
14. 1.338 Parte 1 2: 2141

Mean period: 11-9 months.

Extremes: 11-4 and 13-3 months.
Standard deviation: -57 month or 4.8%.
Mean date: 28 January.

Koompassia malaccensis Benth.

An old tree of this species, remaining from former
primitive forest, was reported in 1931 to have renewed
its leaves in three successive years in the month of
September. This regularity subsequently changed, leaf-
renewal being deferred until 8 November in 1930. In that
year August was exceptionally dry, but there was a good
deal of rain from September onwards, and possibly this wet
weather was the cause of deferred leaf-fall. The next year
leaf-renewal took place in December; then in 1933 there
was no leaf-change at all, the next being in March, 1934.

Vol. XI. (1940).

148

After that there are two gaps in observations, dates noted
being 17 May, 1936, 8 July, 1938, 15 June, 1939 and 12
May, 1940.

The mean period for 12 periods has been 12-7 months
and the standard deviation not abnormally large; but I
think that the incidence of dry weather or rain has consider-
able influence. Thus, once leaf-fall became deferred from
September into the wet season, it became later and !ater,
extending to March in 1934 when January and February
were both very wet. In 1936 February was dry, but March
very wet. New leaf-change is likely to continue about June
and July, when fairly dry weather is frequent, for some
years. If it does so, the mean period of 12-7 months will
be seen to be too high. The rather early change in 1940
was probably due to unusual dry weather early in the year.

There is another tree of this species, beside a road
outside the Gardens, doubtless also a relic of the former
forest, which appears to have shorter and less regular
periods, but I have not adequate observations on this tree.

The dates of leaf renewal, and the intervening periods
in months, are as follows:—

Months Months

17. Oeil kt 2 — 18. 3.34 14-8
44 Oe e. ~~ AES 27D 26. ¢ ot ae eee

3 See Os eat Asie ht $.F.38: |" .2> > ee ee
ra | oe & e 15. 6.29. ..<% ~. eee

Bits UT -. 245 12540" ¢- .. 10-9
Bode 62h. pee
12 periods.

Mean period: 12-7 months.
Extremes: 10-9 and 14-8 months.
Standard deviation: 1-36 months or 10-:7%.

Lagerstremia flos-regine Retz.

The same two trees mentioned in my earlier paper
were kept under observation until 1937. The older tree
continued to renew its leaves first on the lower branches.
The early stages of growth of new shoots are very gradual
and it is not easy to fix exactly comparable dates, even for
the same branch. Part of the variation shown therefore
may be due to this cause; but still it is not larger than in a
number of others.

The smaller tree continued as before to change the
north side at a different time from the south, though the
matter was later somewhat complicated by the development
of a small central part which was more or less intermediate.
The origin of this behaviour can probably be traced to the
fact that when the tree was first planted a larger tree stood
near it on the north side, shading that side considerably.
The north side of the tree, as shown in the table below,

Gardens Bulletin, SS.

149

had a somewhat longer average period and a somewhat
greater standard deviation. The two sides were near -
together in 1927 (only a month apart) but the north side
gradually lagged behind, at the end of 10 years being six
- months later than the south side.

A number of young trees of this species were planted
out by one of the roads in Singapore about 10 years ago.
At first most of these trees flowered about the same time,
but later flowerings have shown increasing differences from
tree to tree. It seems probable that each tree, according to
its nature and surroundings, develops a fairly constant
period of its own, and that there are small significant
differences of period from tree to tree.

A tree in Penang renewed its leaves in 3 successive
years in January or February. It is likely therefore that
this species is sufficiently influenced by the regularity of
climate at Penang to change its leaves at the dry season
every year. Further observations on this point are now
in progress.

The behaviour of the two Singapore trees is indicated
by the following data:

Smaller tree Smaller tree
north side south side Old tree
Months Months Months
20; 8.27 — 20-727 — ae ey 5 —
10. 6.28 9-7 12. 5:28 9-8 27. S28 8-8
Zt ADS 10-4 10. 3.29 10-0 10. 3.29 9-5
i ae 10-1 fh es Re 8-2 2 Of a 8-2
BoL.o0 9. 1. 9.30 9-4 29.10.30 11-4
6. 9.31 9-4 os EEE 8-1 15, 6:51 7-6
netGoe 9-0 12 252 9-3 ol kee : 9-4
i o.0e 9-2 20.11.82 9-3 Ce ae 8-7
(gS) 8-6 eur toes: 8-3 24.10.33 * S02
i, 4.35.2) 18 30.12.34 2 «.8-5 i 6.35. 2) 5C 38
20. 8.37 3-X< 9-4 Silties:. |. 9-3 24. 2.36 3 8:3
Mean period: 9-2 9. 7.36 S. Dec. V9GG6 | | Ob
months. ma. Ber S 7. - Oet. 19ST... -10-
Standard deviation: Mean period: 8-9 Mean period: 9-4
‘88 months or 9-6%. months. months.
Standard deviation: Extremes: 7-6 and

‘77 months or 8-7%. 11-4 months.
Standard deviation:
-99 months or 10-5%.

Lecythis sp.

The tree of this species (probably L. ollaria L.) briefly
reported upon in my previous paper has now completed 11
periods in almost exactly 11 years, with a standard deviation
of only :71 month; it is thus one of the most regular trees

Vol. XI. (1940).

150

under observation. The fluctuations from year to year are
not large, the longest period being 13-3 months and the
shortest 11 months. I can see no correlation between
rainfall records and these late or early leaf-changes.

The leaves of this tree turn rather bright yellow quite
suddenly, and are all fallen in 2-3 weeks. New buds begin
to open as soon as the tree is bare, or sometimes before all
the old leaves have gone. The dates given are therefore
closely comparable. This tree has never flowered, though
it is fully mature.

A small tree, probably of the same species, planted
near the former tree on 24th February, 1928, renewed its
leaves about February 1931, 1932 and January, 1933.
Subsequently it was dug and manured (no exact records
kept) on more than one occasion and subsequent changes
were irregular. It will now again be brought under
observation.

The dates of leaf change of the older tree are as
follows :—

Months Months
8.7.28 Ne ae — 16.9.34 i wins ae
18.8.29 est 74 alee 21,8.35 it oo 123
8.8.30 Ne sey eee 28.8.36 ai . «eee
30.8.31 g22 Peripe 29.7.37 J so ee
16.9.32 “ eee 20.7.38 Me ow eee
7.9.33 11-7 30.6.39 re 22. ie

Mean period: 12.0 ‘months.
Mean date: 11 August.
Standard deviation: -71 month or 5-9%.

Mangifera foetida Lour.

A tree of this species grows beside my house and has
been under observation since 1927 except during absences
in 1930, 1934 and 1938. It is not deciduous but has
presented a rather regular behaviour as regards leaf-
renewal and is worth recording. When new growth occurs,
each twig produces either new leaves or flowers, or both.
In one case only new leaves were produced and then a
general flowering occurred three months later; in another
case a general growth of new leaves followed a month
after a general flowering. The usual procedure however
is for leaves and flowers to be produced simultaneously,
and the records below are for such (except that of 20th
April, 1935 when no flowers were produced). It will be
seen that the tree is less regular than some of the deciduous

Gardens Bulletin, S.S.

151

trees, but has a standard deviation of only 11:6%, thus
indicating the probability of a leaf-renewal not altogether
dependent on climatic factors. The old leaves do not fall
immediately the new begin to develop, as in Mangifera
indica; an inspection of the tree shows that on some twigs
leaves persist for at least 214 years, while on others less
than 18 months. The actual dates of general leaf-renewal
are as follows:

Months Months
AN. Se: a — 2H (2.35) bY. ae 9-7
25. G28. .: _ 9-0 ot i 5 a
= ge tes: ES it 7-4 ROH AAS. . 2s, cae
i kg oS a} 6-5 7 A A oc ee
i is | alae Lf 7-2 rece. AS 7-1
Soar) ok w BTS iw 2 ae Ae 8-0
5. 5.32 = 8-0 23. 9.38 rn

Assuming 16 periods: mean period is 8.25 months.
Standard deviation: -96 month or 11-6%.

Mangifera indica L.

The two trees by the lake in the Botanic Gardens,
mentioned in my previous paper, were kept under observa-
tion until 1937. After 1930 the trees rarely changed their
leaves completely, but usually changed on a group of
branches at a time. Change, whether partial or complete,
usually took place simultaneously on the two trees, but
sometimes one tree had a partial change, while the other
did not. Leaf-fall always accompanies leaf-renewal, and
never precedes it, so that the trees are never bare.

Flowering is rare, having occurred three times only in
10 years. There may have been a few flowers at other
times, but no general flowering. As in other countries,
flowering appears to be quite independent of leaf change.
Flowering follows dry weather, but not all dry periods
produce flowering.

It appears that the leaves usually last 9 to 12 months,
but may exceptionally last only 4 to 5 months on part of
a tree. The case is evidently a complex one, and needs
more careful observation. Though the behaviour of these
trees may seem to be rather like that of Hevea, in the
Mangifera leaf-renewal precedes leaf-fall, whereas the fall
comes first in Hevea. This may mean a quite different
mechanism of response to climatic change, but careful
analysis of the behaviour of the trees is needed before we
can be sure of this. Mangifera trees would be interesting
subjects for experiment on the lines indicated by Schweizer.

Vol. XI. (1940).

152

The following table summarizes the behaviour of the
two trees from 1928 to 1937. “Full” means a complete
leaf-change, “part” means leaf-renewal on part of the tree
only.

North tree South tree Flowering
1928 ...... daly, full July, full
1929 .. June, full June, full 3.2.29
1930 .. March, full May, full
1931 .. end May, part end March, full
July, part July, part
1982 .. April, part April, part
Aug., part Aug., part
19838... Febs part Feb., part 26.2.33
Aug., nearly full June, part
Aug., part
1934 .. no record Nov., full 25.2.34
1985 .. March, full Aug., part
end Oct. part Nov., part
1986 .. end March, almost full end March, almost full
June, rest of tree
Sept., part
1937. .2, Feb, part April, part
April, part

Parishia Maingayi Hook. fil.

A tree of this species has been observed at leaf-change
on five occasions, but not all consecutively. It appears
probable however that the period is well over a year, in
the neighbourhood of 15 or 16 months. I do not think it
possible that the period is so short as half this, as the tree
is in a fairly conspicuous place beside the potting shed at
the Gardens, and though it was not on my regular list of
trees, I could hardly fail to have observed it if it changed
more frequently. It loses its leaves completely and stands
bare for two weeks or more before leaf-renewal.

Herbarium specimens taken from the same tree prior
to my observation of leaf-change indicate that the tree was
flowering on the following dates: 25th April, 1923, 17th
June, 1926 and 29th April, 1930. These specimens were
probably taken about a month after the opening of the new
leaf-buds. With this assumption, the dates in 1926 and
1930 agree with a leaf-period of about 15 months, but
that of 1923 does not. The observed dates of leaf-renewal,
and the estimated leaf-periods, are as follows :—

Months Months
20.10.32 s45 ie — aD. Dead). 'S: 2. fe-2
21. eo. uae 2 x AGD 100 B40" °k fy ae
10.11.36 "2 ry 15%

Probable mean period: 15-6 months.
Standard deviation: -96 month or 6-2%.

Gardens Bulletin, S.S.

153

Parkia javanica (Lam.) Merr.

The tree of this species mentioned (as P. Roxburghii) ©
in my earlier paper has behaved in a more regular manner
than any other of which records have been kept. Its
inflorescences appear in September or October, after which
leaf-fall begins slowly, being completed about the beginning
of February, the fruits ripening at the same time. New
leaf-growth begins after the tree has been bare for about a
fortnight. The mean date of leaf-renewal is 23rd February,
and the standard deviation only :24 month or 7 days.

A very fine old tree of the same species in another part
of Singapore has been observed in 1931, 1932, 1935, 1936
and 1937 to renew its leaves in March, about a month later
than the tree in the Gardens. The data for the Gardens
tree are as follows:

Months Months
Dleeeee oe . x. — (ae) ae cxeeric ss
72 er as aa geet Z-0 Al ty wits) eae Sr LB
Ueno Py yet ay elem Zed see ban ah) ia
je aa ora He cds ae at vie ial |
yi eee) ilaee Wl 5 ae ae Be VAD
Zon uae 12-1

Mean period: 12-0 months.
‘Mean date: 23 February.
Standard deviation: -24 month or 2-0%.

Peltophorum pterocarpum (DC.) Backer.

Two old trees of this species, mentioned in my earlier
paper (as P. ferruginewm), have been regularly observed
for 10 years, and have been very regular in their behaviour,
the leaf-period averaging exactly six months, with standard
deviation of less than half a month. The mean dates of
12th February and 10th August compare with 8th February
and llth August for Terminalia catappa over the same
period.

The tree E mentioned in my earlier paper was evidently
not in a vigorous state. Its leaf-changes were at rather
long intervals (up to 9-5 months) and not at all regular.
Observation of this tree was not continued.

The young tree F behaved much more regularly after
the two short periods of 4-4 and 4-7 months in 1929-30. If
the observations from 15th January, 1930 onwards are taken,
the standard deviation is only -75 month or 11:0%, which
is not excessive; I think the short periods were due to
attack by caterpillars. Possibly the tree will approximate
to 6 months as it grows older. Compare the behaviour of
the young tree of Couroupita.

As noted in 1931, this species flowers at the end of the
vegetative growth, the fruits ripening about the time of

Vol. XI. (1940).

154

leaf-fall or somewhat earlier. Flowering is sometimes
absent, but this does not appear to affect the length of the
leaf-period appreciably.

The behaviour of this species in Ceylon and Java is
evidently irregular (see my 1931 paper, p. 201, references
to Wright and Coster). When I was in Bombay in October
1937, I noticed that all the trees bore fruits, the result of
flowering earlier in the year, and that most bore also some
flowers, though not a full flowering; but there was no sign
of new leaves. The Superintendent of the Victoria Gardens
at Bombay kindly informs me that the trees lose their leaves
gradually from December to February and produce new
leaves in March. They thus renew their leaves once a
year, in contrast to Couroupita guianensis, which changes
twice a year both in Singapore and Bombay. I noted that
the Peltophorum trees of Bombay were much taller than
those I have seen in Malaya, though I could otherwise see
no difference. It is possible that the Malayan trees belong
to a distinct race with the peculiarity of a tendency to a
short leaf-period.

Two old trees Tree F
Months Months

8. 1.28 rf — d, EPR ey —
24. 6.28 5-5 11.11.28 7-3
a. Bee 7-3 14. 4.29 4-1
vA Ry ea 5-5 25. 8.29 4-4
30. 1.30 6-2 15. 1.30 4-7
8. 8.30 6-2 22. 7.30 6-2
4 Za 5-9 10: 2a0 7-5
y= ae eo | 6-4 Nov. 1931 8-
28. 2.32 6-2 26. 6.32 7-5
11--9.22 6-4 Feb. 1933 7-4
) See 5-8 U7:: Oe bese 56 7-
26. 8.33 5-7 Ya ee 2 er ee
25. 2.34 6-0 2h: Fae. es >,
25. 8.34 6-0 1. 3.36 7-4
nis 2a 6-1 2.10.36 7-0
22.,8.85 5-8 20. AST 2 oe : ft ae
21. 2.36 6-0 Mean period: 6-8 months.
28. 8.36 6-2 Standard deviation: 1-27 months
6. 237 5-3 or 18-7%.
19. “RST 5s Suey Gal Mean period from 15.1.30 on-
in... Fisk oh 2 -X» 5-6 wards: 7-3 months.
23. Za +4: Be 7-2 Standard deviation: -75 month

Mean period: 6-0 months. or 11-0%.

Standard deviation: -48 month
or 7-2%.

Mean dates: 12 Feb. and 10
Aug.

Pterocarpus indicus Willd.

There is a note about this species in my former paper
(p. 202). I have little further to add concerning the
periodicity of leaf-change of trees in Singapore. They are

Gardens Bulletin, S.S.

155

certainly less regular and less free-flowering than in the
seasonal climate of Penang; and trees in Singapore tend .
to change their leaves a branch or more at a time, often
taking weeks or months to renew their leaves on all
branches. Flowering follows new leaf-growth, and the
extent of flower-bud production is probably influenced by
climatic factors. Trees of this species in any one locality
show a remarkable gregarious flowering which has been
little noted in botanical literature.

The inflorescences develop up to a stage when the
flower buds are well grown, and then rest for a time; then
all buds which have reached this stage will suddenly flower,
often making a most spectacular display, lasting only a day.
The next lot of buds will go through the same process, each
inflorescence flowering usually in two or three stages. It
is a remarkable sight in Penang to see many trees clothed
almost all over with gold; the flowers are small, and in a
full display do not show individually.

The stimulus causing all buds at the receptive stage
to develop is undoubtedly a fall of temperature, as in other
cases of gregarious flowering (notably of the common orchid
Dendrobium crumenatum). It appears that three davs only
are needed from the stimulus to the opening of the flowers.
The stimulus may be a sudden fall of about 10° F., such as
often accompanies a storm during the heat of the day;
or a cool day, in which the temperature does not rise
appreciably above 80° F.

All cases of flowering recorded by me in Singapore
in the year 1940 (21 occasions) have followed three days
after one or other of these temperature conditions. The
sudden fall of temperature has ranged from 7° to 15° F.
- The largest flowering of all followed three days after a day
of very uniform cool temperature.

As an indication of the slow rate of leaf-change, and
the dispersal of flowering of one tree over considerable
periods, the following case is of interest. A tree near my
house bore flowers on 21 separate occasions between 9th
April and 20th July, 1940. Most of these were only small
flowerings, only two involving any large part of the crown
of the tree. In the previous year (1939) the same tree
flowered on 8 occasions in August and September. The
biggest general flowering of trees in Singapore in 1939 was
in June, and in 1940 on 30th May.

Salmalia malabarica (DC.) Schott & Endl.

The two trees reported on in 1931 (under the name
Bombax malabaricum) have been kept under observation,
and have continued to behave in a similar manner, the upper
tree with a consistently longer period than the lower, as

Vol. XI. (1940).

156

shown below. The standard deviation of the tree with
longer period (which is also the more vigorous of the two)
is only -44 month; of the other tree -98 month. The trees
always have a bare stage of 2-3 months irrespective of
the time of year. Leaf-renewal has occurred in almost
every month of the year.

It should be noted that these two trees are not growing
under very favourable soil conditions and that their
growth is poor. They will form interesting material for
cultivation and manurial experiments (see remarks under
Kigelia), which have now begun.

Upper Tree Lower Tree
Months Months

30.23.28) 9 0 £3 — 30. .d.8 1, / 2% re —
1, 22 Me -2 , 106 16.12.28 ~<A
18-2229) 23 ~~ a8 6.10.29 erin
LTEROS oe ~. S04 10. 6.30 le
AAD SU iE. VeOVAte 3. 531 on
i ee wal? 1 SO 6. 3.32 judd (es
yp i i + ded 1.12.32 ot ee
[ e - aae a). 6. 8.33 = st eats
a. Se ae peed tee 20. 4.34 rd ee
1G) Zoey: LZ aes 1$/'3.35 .. 10-9
2A, B28 Do Nase sas 6.10.36 2 xX .93
Mean period: 10-5 months. 15. 7.37 9-3

Extremes: 9-9 and 11-2 months. Mean period: 9-3 months.
Standard deviation: (excluding Extremes: 8-2 and 10-9 months.
1956-39) -44 month or 42%. Standard ltr -98 month
or 10-5%.

Saraca taipingensis Cantley.

A fine tree of this handsome Malayan species has been
under observation for some years. The tree is evergreen,
flowering on the old wood, independently of the production
of new leaves. New leaves are produced in flushes at
irregular intervais, but the old leaves do not immediately
fall.

Though I cannot trace a definite connection between
flowering and dry weather, it is fairly clear that flowering
follows about a month or so after the beginning of a period
of dry weather. The extent of the flowering naturally
depends on how recently the tree flowered last. There are
usually two fairly full flowerings in a year, and often small
flowerings between. The flowerings recorded between 1927
and 1937 have been in the following months :—

January .. Once July -- Once —
February .. Three times August .. Six times
March .. Six times September .. Four times
April .. Once October .. Twice
June .. Once

This indicates that the flowering tends to be in
February-March and August-September, following the
principal dry periods.

Gardens Bulletin, S.S.

157

New leaves come in flushes at irregular intervals. As
the old leaves do not fall when the new develop, the new
growth is not dependent on the age of the old leaves. It
seems to be most frequent soon after wet weather which
follows a dry period, but this is not invariably the case,
and the response is probably a rather complex one,
dependent on the condition of the tree as well as on the
weather.

Sindora Wallichii Benth.

The two trees of this species mentioned in 1931 have
continued to behave irregularly as regards leaf-change, but
records are incomplete. The tendency seems to be for a
minimum leaf-period of about 12 months, with possible
retention of leaves up to about 18 months. There is some
indication that continued wet weather about 12 months
from the last leaf-change prolongs the life of the old leaves;
but the two trees do not always behave alike in this matter.
Thus, they both renewed their leaves completely in February
1931; one tree renewed them again in February, 1932,
though January, 1932 was not a very dry month, but the
other delayed until August 1932 before renewal. The first
tree had previously renewed its leaves in September, 1929,
but failed to do so in September, 1930, though July and
August were both unusually dry, deferring renewal until
February, 1931.

The trees are slowly deciduous over a rather long
period, and sometimes leaf-renewal occurs on part of the
tree only, on other parts later. The response is evidently
a complicated one.

Sterculia carthaginensis Cav.

The tree of this species mentioned in 1931 has been
under observation since that time and has proved to be a
very interesting case, comparable with that of Hevea, but
less sensitive to climatic change. Its behaviour is worth
considering in some detail.

The first leaf-renewal observed was on 11th December,
1927. The period from 12th November to 10th December was
exceptionally dry, rainfall amounting to only 2-60 inches,
instead of an average of about 10 inches. The next leaf-
renewal was observed on 22 July, 1928, following a dry June.
The next renewal occurred on 10th March, 1929; the rainfall
in February was 10-97 inches, but there was a rather dry
period from the 4th to 15th inclusive (12 days) with 1-00
inch, which may have caused leaf-fall to begin.

The weather was wet from August to November, 1929,
December having only 4:49 inches. The tree had new leaves
on a few branches about 10th December, but did not renew

Vol. XI. (1940).

158

all leaves until 17th February, 1930; the weather was dry
from 12th January onwards (1—11th January, 3:17 inches;
12-31st January, 0-72 inch). It is strange that the two
unusually dry months of July (1:89 inches) and August
(1:18 inches) did not cause another general leaf-change,
but there were small developments of new leaves at various
times between August and November, 1930, (16th to 29th
September and October 7th to 16th, were also dry periods).
The next general leaf-renewal took place on 21st February,
1931; January was very wet up to the 19th, but from the
20th to 16th February the rainfall was only 1-80 inches.
There was no more dry weather during the rest of the year
(not more than six consecutive days without rain at any
time) but a few new leaves were produced by the tree in
December.

In 1932 the early part of January was dry (1-17th,
‘56 inch), but leaf-renewal was deferred until 20th
February. Following dry weather in July and August, there
was occasional small leaf-renewal on various branches
between July and October, but no general renewal until
16th January, 1983. There was hardly any wet season in
the latter part of 1932, and only 1:19 inches rain from
3rd to 19th December; this presumably induced an early
leaf-renewal in 19338.

In 1934 and 1935 there was again leaf-renewal in
sanuary, that of 1935 being early (12th) following a very
dry December. There was an unusually large partial leaf-
renewal on 18th September, 1935, about half the tree being
involved; this followed a very dry July and August (2-73
and 3:03 inches). November and December, 1935, were
wet, and also January, 1936; dry weather began on
February lst and a general leaf-renewal (involving the fall
of those leaves produced in September, 1935) began on
3rd March, 1936.

The wet season at the end of 1936 was normal with
a short break at the beginning of January, 1937, the rest
of January being wet. Leaf-renewal occurred on 13th
February. No further change occurred up to October, 1937.
I have no record of the behaviour of the tree at the
beginning of 1938. There was a dry fortnight in January
and leaf-renewal probably took place in February. July
was very dry with 2:4 inches and there was another dry
period in October (2:7 inches in 24 days). -It was pre-
sumably this. which caused a complete leaf-renewal on 30
November. Though there was a good deal of dry weather
in February and March, 1939, leaf-renewal did not occur
until 8th May, a period of 5:3 months.

This tree is clearly unstable.. It appears to have a
minimum period of five months (55 months from the

Gardens Bulletin, S.S.

159

half-renewal of 18th September, 1935 to 2nd March, 1936, and
5:3 months from 30th November, 1938 to 8th May, 1939) and -
a maximum of about 12 months (prolonged for some leaves
to 13:7 months, on one occasion, following the large partial
change of September, 1935). It appears that after about
5 months any pronounced dry weather may cause at least a
partial leaf-fall, followed by leaf-renewal on the branches
affected; leaf-renewal occurs a month or more after the
onset of the dry weather (in a complete change, leaf-fall
is complete in 2—4 weeks after the first signs of brown
leaves). But it should be noted that where new leaves are
produced on only a few small branches, these fall at the
next general leaf-fall, even though this occurs only 2 or
3 months later.

The following table shows dates of complete changes
only, with the intervening periods in months. For the
years 1929-1937, the mean period was 12:1 months, with
standard deviation -88 month or 73%.

~ Months Months
22 i — yi LESS 7: Se Pak DIT Wye 4
4 Oe he 7-4 Pee By hs eas Aes’
eae ee ny 7-6 VARS et MONO Py 2 Pe!
Dh ah pans tigi as ees te hts BSP P|
co ie A Ge Reo ebe 25 0 S0AL38 .. 2x 108
Zn Boyes be be Le a eet CR xy 53
cA Rae Kid be EOD BED AO es pi: 9-5

Sterculia macrophylla Vent.

I have short series of records of three trees of this
species, one in Penang and two in Singapore. Two of the
trees have shown a fairly regular period of 64 to 82
months; the third tree was in poor condition, and much
more irreguiar, in 1927-31. This third tree was subse-
quently cultivated and became more vigorous. It showed
a period of 9-2 months in 1938-39. The records indicate
this species as a short-period tree, both at Penang and
Singapore. Its natural distribution is said to be confined |
to Malaysia. Records of its behaviour in the more strongly
seasonal climate of East Java would be interesting. The
following are data available for the three trees above
mentioned.

Penang Singapore Singapore
Months Months Months

eb he aes = 26." de os ler: Day Gla oe,
4s DAU cy: 8-2 So. BiBey Wk. 6-4 5. 8.28 10-9
June 19382 .. 6- BO; OBO Si): 6-8 1. 6.29 9-9
Sy, Aceon set « 6-5 20,113.30... 7-7 20. 6.30 12-6
AUP. 1955... 7 2.) Oooh 7-2 2. eal T-4
8.10.38 —

Le: 7:39 9.2

13. 3.40 8-0

Vol. XI. (1940).

160

Sterculia sp.

This tree, the species of which has not been identified,
has been under observation for 12 years and has had a
mean leaf-period of 10-1 months, with a good deal of
variation. The standard deviation however is little more
than 10% of the mean period. The data are as follows.

Months Months

16. 9.27 — Dec. 1933... 2X 9-5
12. 8.28 10-9 25.3.0 >i 2 6"

1. 6.29 9-7 285 3.36". t+ x 9-1
14. 4.30 10-4 March 1937... a4 9-5
& Ast 11-7 ee a ee |
10. 3.32 11-2 21 ae Ae 8-0
13: B23" 12-1

15 periods.

Mean period: 10-1 months.
Standard deviation: 1-26 month or 12-5%.

Tamarindus indica L.

The old tree mentioned in my paper of 1931 died not
long afterwards. The young tree was kept under observa-
tion, but being planted in a clay soil was not in a very
vigorous condition. It showed a tendency to have periods
of less than a year for three years, but then had a period
of 17-7- months, the subsequent leaf-growth being poor and
slow in developing. In September, 1938, three months after
a leaf-renewal, the tree was well dug and manured, new
leaf-growth occurring two months later. The future de-
velopment will be watched with interest. The dates of new
leaf-growth are as follows.

Months Months
ib. Zoe is Pe! — 15.12:54.2 5% .. 10-
> ee 4.123577 3: in ae
(See oo Ae eS - Peer i?
15. 2.34 10-7 1, tae es -s.. Soe

Manured 18.9.38. a
November 1938, more new leaves.

Terminalia catappa L..

Records have been continued of the behaviour of the
tree mentioned in 1931. It has subsequently shown more
irregularity than was indicated in the period 1927-31. This
irregularity was chiefly in 1932-34, and is probably to be
traced to the dry weather of August to December, 1932,
which caused a leaf-renewal on 25th December, 1932, after a
period of only 45 months. This was followed by three
rather long periods, which brought the month of leaf-change
to August once more. The long period of 1936 is surprising,

Gardens Bulletin, S.S.

161

as the months of June, July and August were all fairly dry.
New growth of leaf-buds is rapid and simultaneous all over
the tree, so that dates are closely comparable.

As noted previously, this species changes its leaves
twice a year in the strongly seasonal climate of east Java,
in Ceylon, and in West Africa. In East Java it changes
its leaves in the latter part of the wet season, and again
in the latter part of the dry season. On the analogy of
the behaviour of the Singapore tree, one would expect the
leaf-periods in East Java ending February—March, during
which the weather is much wetter than in the other period,
to be longer, and the drier leaf-period, ending August-—
September, to be shorter. Records on this point would be
interesting. It is clear that the species tends to have a
leaf-period of about six months, and that the present tree
is more sensitive to external influence than the tree of
Couroupita, but less so than the trees of Ficus variegata
mentioned in this paper.

At Peradeniya there are two dry seasons (one longer
than the other) and the change occurs during the dry
seasons. In Singapore the mean dates of leaf-change of
the tree observed (8th February and 11th August) also
correspond with the periods when dry weather is most
frequently experienced.

Recorded dates of leaf-renewal of the Singapore tree
are as follows.

Months Months
8. 1.28 — 1t.. 6.33 he 6:
a T2656 6-0 21. 1.34 : 7-3
3. 2.29 6-8 20. 8.34 5 7-0
11. 8.29 6-3 2a. Ssh ne 6-2
9. 2.30 6-0 10. 8.35 - 5-5
8. 8.30 6-0 18. 2.36 ai 6-3
6. 2.31 6-0 20. 9.36 A 7-1
mo Bol. 6-5 23 ued ny 5-6
62:32 5-5 9. 9.37 Se 5-9
es Py 6-1 1.9.38 ya eR
yp) Wa 4. 15.°3.39 sf 6-5

Mean period: 6-1 months.

Extremes: 7-3 and 4-5 months.
Standard deviation: -64 month or 10-5%.
Mean dates: 8th Feb. and 11th Aug.

‘Terminalia subspathulata King.

This local species is represented by a fine tree on the
edge of the Gardens Jungle in Singapore. This tree has
been very regular in its behaviour since 1928. It begins
to lose its leaves about October and loses them slowly all
through the wetter season, more rapidly when the drier
weather begins in January or February, and is nearly bare

Vol. XI. (1940).

162

for perhaps a month before the new buds begin to open.
The opening of the buds is quite rapid and simultaneous
all over the tree, so that dates are closely comparable.
Flowering does not occur every year.

The factors which influence the date of leaf-renewal
are probably rather complex, as the period of leaf-fall
extends over several months. There is little doubt that the
tree has a natural tendency to a 12-month leaf-period, and
also that the occurrence of the bare period immediately
following the break at the end of the wet season is signi-
ficant; but the slow loss of leaves through the wet season
is curious. The data for this tree are as follows.

Months Months
19.4.28 teh oe — 10.5.34 Are a iiated
abs fe ane 4) Oe Fi 8.30 Ne . 2 eS
16.3.30 ie Dates 45 |) 10.4.36 Ab: . pees
Blips d Ms ssdasite eA ae 29.4.37 i, ae
10.4.32 p AWA 12.4.39 a, Zin he ee
2.4.30 : 1-7 12.3.40 ape ee

Mean period : 11.9 months.
Standard deviation: -88 month or 7-:0%.
Mean date: 6th April.

Gardens Bulletin, SiS.

163

‘SOYOUL FZ-CTT “aVoA oY} AOJ [eqOT,

ceil 86-6 LO-L 29-9 Lé-¢ 86°€ 8o-7 VI-6 €8-11 08-91 L6-01 Gé‘81
G0: 6P- —< ray < G1-G v9'l
= = Ge: SL ae — ae = 89- i £0:
GG: F0- £0: = ee <= = 9L éL: 1 ee G9:
Lg: Hig 10: G0: FO: ioe = ae 90- 60: OF-G 89-1
rc: 10- OL 9€- 08: = v6: IG: a = G6: so
L9°€ él: PE: Ge'l == oe en al OL: aes T0- 5a
10: éT- GG: 90: F0- ass Le: 0: 68° 90: — 10:
69-6 66 oe vg - as ie ap v8: v0: T0- LET
oF: LO a= LO: [é-1 oe ST: — ie L8-1 = 90:
v0: <— 9T SY 9F°1 = v0: &I- Le GO- 0G: aa
GO: Si: aa £0 OL: mR 8é- &G: re 90- G8: 80:
80: F0- a = G6: Or GI: él LVI 10-6 09- IG:
as GO: = 60 se oe £0: rs 68-1 68°G = 1 aa |
10: 10: ies =: == LO: r= 9G- re IG: es <=
gsc éI- = vP =e IF: £0: 8G- 80: ee rm, 8T-
ct: ae — ra G0: 6G: =a LO: Seer — 661 90:
— LG: Be ee =e — ae bi coe i €0° GEG
LO: T0- oe == TO: =— =a a — <= 6I- 10-1
IG: 60: 10 £0: a a x — ¥9-1 o aT: ney
6S: — — VI: LG: £0: bis G9: — a GI: Le
6ST 81° — are ca a am =r = ain I€- €0-
ele 7 =e nee ee 68: 66-1 > 0&: G0: 10> 66°
= IL 9&- ae ae F0- de 6-1 &T- [el a PI:
10 en 9g: =o es éI- ap €0- 90- &- = G0:
pe 80: LT: a LO: 10: oe LG: = LG: a a
60: [€: Oe: ae [ele OL: ca arc = 66° LO: T0-
T0- OG: —_ T&T es GG: 9I- ae LO- 9L°& LO- Sy)
90: == T6- ea es 90- GG: 99° £0: 66° GO: 90-
80: 0G: 0G: 80: €1 Té-T 10: 09-1 10- ae Gol PG:
FG: ott OF: CO: oa G0: a OL: EG: = OF: 16:
60: 28 GPG an = we Ts 10: OL: = | LO-G F6°&
‘00d | “AON | I9Q049O | ‘ydog | qsnsny Ayne | eune | ARTA pudy Yoleyl | Arenaiqoq) Avenue pe |
|

LGOL .

mA OD SH Id 6 rE CO

. XI. (1940).

164

86-01 89-L1 80:8 08°6 PLOL €é-L OL-E £9°§ L8-T1 GL‘8 Grrl
&P: aya 6§: GEG 3 5
—_ ae — T9- = es 8I- OT: = =
6G: bP: de — cP = = 80: ra i ats
Laan | = ri. =“ — él: = Le: pass a GO'T
a a — G0: ot oa 5 oe LO 10-6 a
LI: ae [9- ie GL: ya ee Pil oe 10: 4S
10: 61: ae = Gol 2 bad — =e OI: 9T-
= él 1 PI GL: ae <= €1 eS re, G0°€
GG: ~e: I&: = zit as £9 — a2 = If:
L¥: re == 69° mg oF ae = ae OTT igs
£0: is “As SF: RG ee oa Mae ope Og: =e
a0 6%: F0'1 —; O&: a = - Gel ited og:
ee 8I- eI: as LiL rr I€ oe o:T 1 Ma | 10:
ze 96° le = = == = 0G Fae 66°T G0:
IL: 66°F OFS =3 Pe ey r fea = GG: 96:
Gol [v1 8I- 8I- 69: eet ed | 6c G0: Lg:
9G° Sr eT: GE: ak aes re ail Piel 60 0g
LT 0G: £0: 6g: 60: aoe ie Gl == es =e ise
Iv: 6E°1 — 99-6 19: faa 80- G0: SO a: “
ed £0: <> 90: es Or == 6: 90 pose 90°T
see = 90 at G9 F0°% = = = rs GP:
GE: S9-1 = aa ee 68: GE: = 86: GS-1 oar
Lert 00°G 9 ee: So eo 9T- Gol ae- aa =
g9- Ic: = | OF: =a = ace G6: 89-1 st =
06: — 8& 96: F0- I — — £6: <a 60°G
a 661 ax ae 90: or 80: 6¢ 80: = ss
8L: £0: VG = —o GG: = — G6: =e ail
LO: i, ae ae! = —e GE: = == ae 90-
ae 9€°1 OT: €G'1 LET 60- - sje os rs 9I-
GE L9- £0: ea =7 “e age = =— ire LO-1
= 16 call P8: al 1G: a 3 “e OL ae ee
7 7 |
‘09d AON 19Q 0990 ‘qdog qsnsny Ajne | ounr : ACW | Judy youvyy =| Arenaqo,y

Le SY ee eee, es ee ee en em

8661

HUOdVONIS ‘SNACUVS OINV.LOG AHL LV ‘SHHONI NI “TIVANIVY

Arenue pe

mn O10 HID Or DD

Gardens Bulletin,

165

*SOYOUT CQ-9g -IBeA CYR OF [BIOT,

60-9 19-11 T€-8 $0-6 | L6-L GILG IL-8 L9-8 1G-G 90-6 66-8 GGG
9I- €¢- 86°T ax FG'1 86-1 rae Ié
ES: 80: SI: cr: €6°T ore see — os ee aa O€
=e 96:1 GG: 08-G LO: see G0: eI: 6G: ati = 6G
G6: — 9¢-T C1 —— 27 GGG GG: pa F0- — == 8S
= — OT: — 60: es — SZ: — LG: L¥- = LG
c0-: 1g = GO: : oe — G8: G6: FI 96°T LG: = 9%
ee <a 90°G oe FE: €9: GLI = = 18: aoe = GE
cae so veo 3 L&: =e CT: LO: PE: = LQ: ore FG
=e 90- 90: G0: aoe = = nen = = = eal €Z
IT: — 61: i as = = es = ais = €Z GG
€Z: OF — ss ae GI: ee = = a ¥6 a IG
OG: = FG: GO: oat 60- — 9Z°1 OL: SI — = 0Z
a &&: GG: ae Sr 80: A a 60: 89: = = 61
=. ¢9-T — "ES ae ag = = — 60: FI- sts 8ST
g¢- aa — 1G: atte as ae = GO: 10: Af OT: LT
6G: 10: — OG: — $9- GG mee ae = — 9T 9T
— FI- €0- fest LO- I€- — = 10 IZ: 16: = cI
= 9G°% =< IT: saa 20% ae F§- 16 0Z- — Ot al
= L@- 80: = = 8: a LZ: ee oe IT: GZ: — €1
oat FI- =— OL 161 = G0: g¢- G9: €I- C0: GI ral
= 90: L9- — 19-1 — —- | = = OZ: L9-T — II
= 0€-Z 61: CQ: 0G — = ae c6°1 nt = OL
7a oT: O€ a — — a oie a ad a == 6
IL: TO: es Sa at — 9T =a se 80: re 8
80- = GT:. C0: = a — — — —- — i
10: 6F: 0€-T II: — oo IG: a aa 62: O€: 9
LG: $55 ee $9: — ee G8°G = sr €&: SEG G
F0- = Or LO: Pe = = 90: “=< FI- = 17
EZ: = a oe a = 9T es at Fike a €
GFT: == == = a= =< = LL: = OL-T 89: G
FI: OF: = OL: Se — a 90- aoe Gg: 96: I
| iS
‘00d | “AON 18q 0300 ‘gdog qsnsny | Ajne 3 oun? me judy yore yl Areniqe,q, Avenue e

a A RR 2 RR

6c61 |

l. XI. (1940).

166

| |
80-01 é8-OL 10-6 I@-L | €I-1 Loe €9-61 éG-O1 0-1 [L-G SEP 68°€
Th: ¥9: ser xT es sae == Ig
LO: 66° €G: eI: ma LO: SS = oe = — O€
OI: ot: éI O9-% = ri =e = = ets = 6G
G0: €E°G "Sor ae = £9- Sie =s, — = 6¢ = 8G
60: 9¢- = =< an = Laas eas es — — 1é LG
10: = 0G: | SG: ie a: ste rang as = OL: = 9G
68: | ar g9-T | ¥G I€- iad = LI: as = = = GG
om 8I- GL: a = O&: = 68-1 LE-& “ies = aa FG
oe Se a 1G: pe = 9€- = a 9G: = ae = 6G
6g 6: OF: = ar a T€- fis LG: €I = ae GS
i 60: GG as — = = es F8- a 5 = IG
0g £0- ae nee 10: a a LQ: 9T- = Free LO- 0G
$I 80: LO: | — ha ie CG" ee PG: Ene fa = 61.
08 g9-1 LG-T | as = <= 10-1 _ <e €I- i == 8I
~— 8E-G €é-1 | os GE al 8P'g L9- =r Té-T €0- a LT
Té: G0: see Zee = = IT: 55 oe See G0: O&: 91
L¥'I = oa 9g ae a= re a 60: 6L°T oS OL: ST
&P- £0 GG: Par 90: <3 8é-1 =~ FET ST = F0- lal
= ag =e | 66-1 ae OT: = = << = ae — €I
gg: G0: oa G0: a 10: — GG: = Se: Ses = él
LT: eS = gO: VG a 6L°E 66'1 OL-T GE: — CF: a!
= ae =e | = eae ae == LO: G0: 98 [6-1 €I-1 OT
= = os =e = a ge 61° mage oe r8:1 a 6
88: 10 ais sed —4 3 = 16: ed OF see as 8
£0: = = G61 = ats soe £0 €I- Ss == 90: L
= 8: €I-1 i 81 8G él ai COG <— wie = 9
68: 9g: 80: 61 5a am —- Pee SFG ee = GS: g
LG: 0G: 60: si ee | = ee LG-G Ga" 10: ee: GI-l ¥
60: FO: oa as ae eae oF ES SN = = 60- €
aaa = 90: Bg on =r: gd CFI salty way = = G
PPI | aca | poses 60: os =. ae VG: 9T- ay = | <a) I
— —— a aa — = — a —— = —— —
"00d | “AON 19q019Q | ‘ydog qsnsny Aine oun fr | ACW qudy yoreyy =| Arenaqog) Arenuve | —
O€6I

WUOdVONIS ‘SNAGUV) OINV.LOd AHL LV ‘SHHONI NI “TTIVANIVY

Gardens Bulletin,

167

"SOYOUL 8Z-GOT aVok og AOF [!eIOT,

| |
PSII 6G ET 0€:8 | 89-6 16:9 ¥S:°8 | S9-L O1-6 99-L PG-L GLP FL-<l
LO aes 10 IP 61 L&T =
Lg 24 8z-1 CP: = = = sexe eo Sees Gg =
<—¥ 06 ears — G8: VE oe PL-G 1é 8G a
€1 9L éé1 | 6I €6: LI ae 8ST: FG = fear —
9€ = = IG: Sg ee LO: 6G sd =< are
08 80°T I€ — 68-1 66°C —= 66: 1G = tee Se
6L cs: ir os = 1g = = ae se =e
99°¢ IL: — 60-T = L¥- 6€-T suis GE — 60-1 9€
GG 90 61 90- Bae a OF: 60 6€ ane as LS
OT ete [L 60° LO = Gr: 60-1 = GS-1 oe OL
ie = at +S-T = LET 66° == 1é 10 =< 80
a r +9 90: = = == = <= VE So na ese
oF 80-1 —. | 69-1 — | ¢0 — — = 120 — | 9b
= F0- Ss =e = 7 £0- <i T&T — G0: Tél
9T OTT 90 tal = aa ere SF nr =o EE: 61
Il = sO 10-1 = aoe — a == eae 60 Ig
9¢ <= EE aa FI: = = = 80- ee él vP-l
vg: = =a 88 6: = 80- act G0: VE <= Ga!
IT: IT = £0 £0: = == FI: | a 2 es 06-1
ol: 9G O& Or 19-T eas = eri ee F0-1 nae L9-T
él: 6L cre &¢ = as L9 60: #9 OF 1s 0€
10- <2 6F a a = = OF: = 18 = FI
SF 8L LUT 61: eee Ss x= seat e: =e 2 6
60: ae = ee VG ae = IZ ae Fe pe: 8
= 69-6 aie = = G8: ae = rE | OT 64 L
19-2 SS; = c0 =< ca ars = ae 90 3 9
LT: ELT <3 = L9- Fan Ba So = 06-1 6: oa ¢
£0: QI: eF: ne ev ss) 20 | S13 | eee fel = = 7
=n G0: $e gO: Si €9- €é-1 GE: $8: IG: 10-1 €
. 60: 86-1 = = 60: = = OI: FO: = G
89- gs: LO: =e aes €F-1 10: F0- a == a iF | I
‘00q | “AON | 1940990 ‘ydog | qsnsny | Ayn ep oun : AeW judy your | Arenaqoq| Avenue e seu | —
| !
Té61

rraret Tey ATTN «(6 NAwereraTreTsThR) OAT Aer mr ope se eee Pe ee Ae ee

ol. XI. (1940).

168

= <2 Se eee eee = = a es
69°9: 61-2 66°€ FG-8 €8-1 66°E1 £0-¢ 69'FI ¥0-9 OL-9 06°F | L9-9
<i F0 = 06:1 ae mad | ae Ig
= aT: a) G0: as aie <oy 861 re; £0 €I 0€
0G: el 9€ Fo GI: = = — at €é = sid 66
0G ‘oe = 60 IP: LO: = LT = adios as ce: 8G
LT: = FO-1 xy sig = 80- ars &I GI: 7 60-1 LG
63: €1 = = 6I: ad 68-1 ah = ZO-1 be: | PDT 9%
60: 8G =a él = = I&: FO =e = T0- GG
= = €& S8:T =< Ss a P6'E ar 9g I€ £0 FG
&G: ine ae 0G = oF ng L9O-G 8E- 09- ie: €: £6
on: z= ie c0 S2 OF OL we = P6'1 él Piel GG
ira oe 60 = OL ae = VG LY GL: 60-1 cist 1G
[6 £0: 9 = 80: a oT: gO: = 90: 09: Le: 0G
eee G0: LO: 7 G0: = al ay cag as Gel ‘ue 61.
=n Ge: it: sara 66: = £0: Fg: Bie = ny ig: 8I
as ae as GG LO: cs 0G: = Or co = = LI
nga F0- 60: = 60: Sax sie £0: €0- me = = 9T
oa G9: ae GG: F0 =F 90: = 90 a €8 a SI
91- a pre f= as <2 G0: a oa 40: = €I- ial
Sst = 1: = ag 60: ae = e=. 90:1 10: ai an €I
10 VE 09: =< gee i = 18-1 = = ea = él
cad IL ae Go-€ = FI a == = = a3 == II
3 9L-1 os ome ox OF-G VI: as GL e 1é = Ol
Ig a | a LT £0: = =< = cO os x al 6
Ee €1 = I€ =, 88°G 6E- e- 9&- = ie = 8
LO mig a ail ee OT-G See = ae OL: = 6G él L
== = ny oe == Sol 88: = = = = I€: 9
cO- BG: = ce: = G0: ae ass OF: a £0: = g
[&- £0: =a L8: = T0- 8ST: £6: = = = oy P
80: a <= = GG: =r = = OF: = =, a €
OL-G 9T- GG: Ol: fant 69-1 = 10: 96: = = = G
G0: L&- is 32 eae 99- = 68: F0- LG-l es = I
‘00 "AON 410Q 0900 ‘qdog qsnsny Aine oun ABTA | judy | yoreyy =| Arenaqeg| Acenue pe

——S— snes

GE6l
HUOdVONIS ‘SNACUVY OINV.LOG AHL LV ‘SHHONI NI “VIVANIVY

Gardens Bulletin, &

‘Seqoul Q[-16 Teoh 9Y4 JOJ [V4OT,

169

99°6 OL'8 ¥9-9 Ge-Il €9-¢ | 6-9 09-01 | GL-¥ LS-6
ae PE et 88-1 = Pole a
61 &e'l ST-€ = a Lé-T on aa | ae
80°G IG: Ge: VG: €¢ as ST-T a a
ae G0: re Se oT = ae LO | GG: a
FG Te: ea ane 10: eu et gas 99- am
ae €0°1 | 80 ee ae 6L: Lg: v1 €8 F0
ee FP FO = re ae 89: 3 gO GL:
O€ VG aap =o as F0- G0: OL: ee 8G
male OT'l ae = SF G0: = F0 ae —
ae GP se ee =s OL =e Pie; = 60
—— GE 7) ae a3 gO: — ed = VE
= a GE Or £0: Gc: él Ee OF BLT
“2 8G = m= 68-1 = Es LO aE ¥9
see Er &G = 60:1 IG: If rae —s L8
ae = 6g = 80 91-1 &G 61 a =
£0 OF ea 6&- =e res él oe = =
oe Si ae £6: —s 90: 1G: Se | ae a
5 = OI-T 8é ¢0-T I€ GG £0 coo ag
él ray Or &¢ GG: £0 9€ patios = =
0G ar Wee FE LG: IT — > eR == IL
ur = ae Se 6: a 9 GL Sty 6a
cP aa a Se: LO: ates oe 6L°1 ae =
bar 86° ay €L a 90 9G A | | = 2
OVD OL: 0G: 90°G a ee €0- L8:1 | — 60:
6T: can wae 1G:G x Te F0- 66: cae 0G:
ales ay i ee aia OTT 90: =. SFT:
6€- << Sie ae a nak a 8c: sa ci"
= a = 8F'1 ie a — s: ae £6-
se == ee 80: = as = ae as E8°G
Fig 5 a es LVI a = G8: ay £6: age
B= £0: a = pee Sa SA 8L: = ris
| ait
19q099Q | ‘qdog | qsnany | Ajne | oun | AVN Tudy youreyy =| Avensqoy) Arenuee

€e6l

. XAT. (1940).

170

‘S23 L6-FI 8E:8 G0°8 GL°L 89-11 L0-9 O19 TS:9 L8°€1 Té-L 96S
0G = 99°€ xs ed 9G ad Ig
id 69: a & FO: sai uaip oe i F0- — ane O&
1G: OF: G0: — aa ta es 91 96° =r Bogs 66
Gs: PP: 9T°G me a — ae = = = Ss are 8G
= FG ee = re 8: = es Lal LV: en == LG
ease boas = == bE [8-1 60: = — LG: ae ea 96
= ie = ee GP OF: 09-1 fica mat ee: Og = GG
90 60-P sean =< a ee: GI: = ae GO- — pe FG
or = = £0: Sok LO: oT = ors = OF peed &G
ie G0: Or = &G° | $61 ae = = | = cate GE
S= LO: cat 66 £0: — IL = v8 GG: <7 i 1é
ad Si. = < =a 96° a) = oe aa ar, OS-G 0G
= 8G: GI: aes F0-T GE: a: = =e 90: = T0- 61
ae Lé- GO: | 2 ¥0: 6: UL as = = 8: 60: 8I
10: 9G- GO-E 0G-1 TL'T QE: F0: €L: Zoe 8g: ee GS: LT
ee IF: = 5 ig Rea 90: GPG 0G IG: 68-1 GPP 9T
nea ¥6° ST- as a? 16:1 = LG: i GG: ae 9I- ST
= c6°1 OL: =e Be GE°G ee a O1-G GO- = OL: al
acs TAL TO: = 90 ae a) a G6- v0: = GS: §I
ae — TZ: = = 6: 66 as eo: gs" 00-1 OT él
aa T9- Pel ae 80 Say 0G: G8: O.-T = or IT: IT
= 90: ae <7 <z eae 97: 90: G0: Ce 09: Or OL
ies 9I- 0€- = = 80: = = = 33 £0: = 6
ee GP: 1 oe a Erg xa ie = ce €8-°P api S'S 8
96: al == aS i= aE — —s G0: 68°1 60°1 06-9 L
6: FG aes == SI: ae = = a 00-1 v0: OLS 9
£0: a oss L8- = ae as Le: 60: 60: oh £0- g
Tg: £0: iam 60: 90: = = OT: eae OL: 61 GP: 14
99 = = FP: co ee G0: TO:T ee 2c: a OT: €
— 96: Sai LG-G bes aes 69° 80: ras CG" an =s G
90: 88-1 =e 68° en = =e GO: ea O€: ae ia I
‘00 AON 19qQ049O ‘ydog qsnsny Ajue eune | Ae judy yore, =| Areniqoq) Arwnuee
Eel

WUOdVONIS ‘SNAGUVD) OINV.LOG AHL LV ‘SHHONI NI “TIVANIVY

Gardens Bulletin,

171

"SOWPUL EF-68 avoOA 9Y4 IOJ [BIOT,

au | LO-E GO'E ELS SE Bim Ce 6L:P SL-or €8°F | Lael
= = ee c— ae =a EG
— = oe aa LO: as sI- ce == 0€
Ss GO: IZ: = ar ae 90- = = 66
=o el: €€- a5 = a= = GOT we: — 8E
Ig: 66 GT = GO- — oe O€- a =— LG
89-1 a = a FI: = = I9-T = — 9G
9G LG: G0: son LG-T = es = = —— GS
Se oe ae =a — CT: G0: GE — = FG
OG: = Ié: LO:T Seed ES a = mre cg: €3
9T- Sag = ay 6¢ («OFT = — = OT-’ GG
FI- == ag pane: — | 10: = hen 9G: i in
6¢ = ee = ae | = SL rem I II: 9T 0G
SS = €0: == GI-1 | es no et ae GT- = 61
a= —— 9G = a mall FS ea FE ra 8I
<i see rae == G0: | — 90- 06°G G9-T om LT
GG: ae Pr gs = 80- —6880°T G0: G9: LIT E> 9T
IG: ice =e = 80: 86: GP: al GG: ria GT
a= €9°1 GT: —< oo | PI 09-1 = — G6: FI
ee Ig: 9<- IT = . a GP: a4 cas = €1
nS ame &¢6 — eas a €I- er == = GL
—— — =a — aes OLE GS: 6€- ~ aie Pex II
= Saar LO-T 2S FET ae | = 90: ex ee OL
F0- G6: = — == =< €I €0- = ! —— 6
—s = oe LG: in — Sond ¥0- saa | LET 8
a eo as aes so sp ha | Pa LO- ms | 19°§ L
6G a= a aie €¢ es G0: 18: a | Sg: 9
as ao aoe = a ae 90- i = _ 66: G
= Sie ae FI co: OF: aaa = ae 90- P
= = a os 18: —_ 99 98: se CFI €
G0: = a= 06 O€: 9T- ae oh Ss sd G
= = ame = G8'l 10-1 coe FP- ae oa I
|
19Q019Q | *ydog qsnsny Ayu eun er AVI Judy Yoleyl Areniqe,q| Avenue p

lL XI. (1940).

172

LG-Ol r8-FI LE-Ol IL-l GO-F GOP 86°E 8G-L 99-9 68-el 69-1 LO-L
= +) = and aa 98°% an Ig
£0°6 | 90: 98:1 F0- 80: aia | = se L8: ae 8&: O€
= | 6ST = = wig St rai Pies | — wee = 16 66
10: «8%: 10- = GO: far i ald | a3 0G: Se = 8G
LO-T ot a1" ete cO- = oe es a) = i eerie 3 tae LG
= Sls T9: ae — — = 8g: | oe a i a 9G
<a «TE: 691 = — <i FS: al | 69: a = = GG
5 «61 ae a Gg: i <a = | 66 £0 ma = VG
me | sag aS g9 =— - 0G'1 ace e nei 80 = €G
oe | OT 1G: wis GG: ea ra LO ag: 91-& at 90 GG
&€ : = 8&- ee G0: fs PE: = éI = 60 te IZ
ras G0 = OL F0- =e = 00-1 re pe = 10 0G
= 6E 0G: ae: &P- = or Ol: | oF = a re 61
8E-E GL‘G €L: = 80: en | ae == <= L9 = re 8I
= =~ aie 2 GL ae g3 Fel a ee 60 | £0 LT
OL = poe FI i as | = 6G: GE: <= i ar ae 91
€T- = LT = a €8 = 6€- | GE 91 aa IG: ST
rks g0-T oa 90 et = = = ee tg GHG 90 ia Lal
== F0 oe os ri ats —_ 9P | LO- GE +0 — eI
=e = = is a PL 1G ne PP: v9 — PL él
ra a 19 eS we = €€ €1 LO: £6 —, PPG [T
GG een 9& = PI: wal — 68 = = seid = Ol
IZ $8: eel eS a PLS | = 60 eet = ites me 6
od 8G TE: a, IL: | al 68: aes = F0- = 8
= IG: "we — $0: =e | == sats LO-T =< ar LO: L
8G: IZ: Ze = ET: set = VG: 6P: Ge: a OL: 9
[9- = 2S 20 er 90: 7" = VG: GP: — 66° g
T0- a 60:1 rs yg | gs ae, 06: = “i | — G0: P
LAU 8é-1 = - “Ss = LO: 8%: = copy Wee er LN! 60: €
Gol 18: ted ile 60- 60: | = acs £6: os = 10: G
= way we ee are i! oe ris 66: = med PG: I
"20(T "AON 104Q0}9Q ‘qgdoq qsnsany | Ajne | ounp | Ae | judy qoreyy =| Arenaqo | Avenue pe

ee
9€61
WUIOdVONIS ‘SNACUVY OINV.LOG AHL LV ‘SHHONT NI “TIVANIVYL

Gardens Bulletin,

173

‘SoyoUl FR-L8 Bok oYyy AOF [BO],

Le61

TATA TWONTS SONATA WD) OTINYV TOU WHT, LY ‘SHHONI NI “TIVANIVU

9S:E1.. 69: OL 9L-9 GG-L 99-Ff | 19°G G0°& 9L-8 6G'6 LG:¥ 16:9 OL-OT
ae — = is PE: — — ie
80: a 60: SS pee Gey a Tel oe = Ig: a O€
ip ce: TO:T a Eee — a = 67° ei “aa 6%
== a aa are = eee en a $9: ae a aed 83
— atl abe 70: we | — 0 9-€ —} | 0-8 8Z £0: LZ
GF-l $91 GL: 89- FE: oe = apse: Ste. = ——- | &&: 9G
at ae 90- I¢-1 = | sayy [0: 80: F6: G0: €0 | 66° cS
os GG: OF: €0- — igs 89- = [G-G al iced | er VG
<ee 98: ot LT: = | 60- = — €s- $0: = | is 63
9L- 90: GO: F0- —— | 90: = ae | 86 G0: — «OCT GG
Fé: IZ: LT: 08: a ia ee = ee See = SLI a ike
OZ: T aa cs: LO: OG = ma — reece = = = 0
GG: €8°1 €¢ = Sit are “= 90- 93° IP: ara = 61
G6: 19: ae << LT°1 = = OZ: 1 9€- ae! GI: = 8I
L¥: cia ge ee er Ex as 0G: T0- ad €9- ia LL
89-1 OF: —, LET 1é-T =z | ¥I- FG a = €Z- OL‘T 9T
08-1 me £0: 60: EL ae LO: a: sea 5 GL: 99: GI
aa snd id - saad ae = | OT: 6: = 89 = FI
90: Iz l ers GO: = = = | LS: + aye ata’ oe €1
G6: 55 O- T&: 0% 8g a= cee ery Gol sad 68° ral
ae = ee €¢- == oar 80: | = 10-1 GO0- 18 69: EL
sia 90- ap = as 90- ae = me ad == zee OI
a LG: a =< es €8: == = —— — = ot oe 6
GP: ce ae es GS: — = Ee: a rig a raed 8
aaa 80-1 Hl fies a GI: 90- 00:6 ane 90-1 ie =F a L
G9: — €é-1 Race = LS: nme: 8L G0: oo = Ea 9
EG: 9€-T so — aA ae == 7 G0: =e acd ee ¢
¢cO0- —— aa SL: +s as == OL LG: == $0: a 17
90- ‘ase ¢c0- GP: a aed = sae —s GG: pe GE €
69- = wa 90- si 9¢ = CO: — 10- 8-1 a G
98: Lé- oS 1 PE: aia ce = eo =, cad 98- aes: IT
| s ce ee |
"00, AON 18q019O | ‘qydoq | “qsnsny Aine oun er ACW judy Youre Areniqe,7| Avenue pe |
LR

Tol. XI. (1940).

174

Go-9 €0-L PES €6°9 VLG G¢-8 g9-9 b9-L 08:8 88°8 GG-OL
at SLT its — 8é 6g
ss as om = ie 16 =e A 86:1 LO
OT: ore LST — = are T9- LO a 61
60-1 £0 = ee — = ner — = 8é- 80
et GG a: — cs, 19 ss 16: F0- OL 6L°G
= LEG “7 = an ae — Gol LI OF OL
at Gg a) a 8 — sate = me = LL
80 10-1 og LG = a EO7 oF 60 — Gg:
= = ne — ane LL: a = ee _ O1-
Le: = GL OL — <= 90 10-1 60'1 =o oe
au) a 81 3 OT: a a GG: Gé-1 — a
= 0G ees 9T- eK =— rie — oa = 60:
a9 oe = iat = a a 60: — = 60-
GE ae 69% —— cae see Tot = 61 = ==
£0 = as II as = 6: — 6¢ a, =
G& 9T ae ae =? —— = ne cs res SI
— — — — — IL: -_— — — — —
rane nes ae oF £9 as LI: ve 61 = a
= awa 9G 6: 66 IT-l = = 1 aa | 91 =
GG: L@:l 9L- vO IL gO <3 = awe = =
LO: oH ae Se ig — = — 90-1 = =
Ss es = = 8: aes VG 66: == 0G: —
= = =r sa ae = 60 — — OL = 6
GO: = = ee = = = a 0G: a | 8Z- 8
OF: 90: IT: Sp ras = os = == 1&-Z al L
96° oS = <= = | 68° 61 6: GG 69-°€ 7s 9
—s L9 ee = rs 91°G at I@: ik =a 60 g
is a 66° — = = 9T-T OTT = <a = r
_— ok oa ni 6T- | ames €I Le: = FO: 6€ §
= Le: or LG: = a fy Ka a0): nals 80-1 el G
— me 1a GL-G me 10-T = GG &G se si I
"AON 19qQ019Q ‘ydog qsnsny | Ane ouner ACW judy yore, =| Arenaqo yg) Arenurve
8E6l

GUOdVONIS ‘SNHCUV) OINV.LOG AHL LV ‘SHHONT NI “TVIVANIVY

Gardens Bulletin, |

175

‘SoyoUl £.G-9G IVoA OYA AOZ 1BIOI,

SL-E1 &L°6 99°8 99°6 06-ST 66'V 0¢-S Vay 69-4 ST al a Lam
80: So 10-6 LL == = GGG [€
= ar eg: = 8é = = ee oe: 8L == 0€
ae ata! = ae ore 06 a ee a a: ore 66
g00- rm Sos = 1G-@ se G0: es <a = ae so 8G
a Lvl VE: Es OL-T eae a =% at aa ELT F0- LG
F0- = Le: si re ee =< a= G6: G6: €G ar VI: 9G
5 = ace oz 60: GE: rea oe as ot se 80- GG
ie 6el &G 68°€ é = os os ies rap £0: ITT VG
a Ska = = rae LO: ee se see es aa PP: &G
es a 6I- “=F ae =< sod = os +. ry" GI: GG
€t 6T 61 €0- = — =~ see es a LG as 1é
=¥5 om = GG" = =—S == “ee = es er 1g- 0G
ae = €0- eae Fe ae oe ma — 0G a oe: 61
as SS GG: OL: SII Tél = See aa es as ae 8I
=z SD: = 8G: eel oe == = —= ea oe él LT
LG-% io = eae LO- ee 90: c8: G8: a ae a 9T
89- we Sg: cy: G8: cot 8&- = = ae ae = ST
£9-L £0 ee LO: =e see = oa" 6G: =e ses os ial
—= Pe: Le: 9L°T si == a whe SI: = iF: ¥6 él
Il él el 68: = GG: = SY Sh = = as él
Se i 10-1 a OT ae Se ee es = =e Se: IT
oes ea 89-1 = sa a, 99°T = ace “as ae a: Or
Le: =< re =e 10: re ee 00:1 Soe LG: == se 6
as 60: — on = sn = 6V: 86°G 80- Li = 8
c: 9L°T 9€ a GL-S = 66 OG: xi = 67: Gv: L
File as “ie he GL 9S: a me Gg: ae pile = 9
IG: — ee a = <i = a 89: LL 8I-1 G9: g
ae se 8E-1 = = = er ee 10- cme = es v
See vO LT gO: as =e car == = Ve" La: = €
oe = ay = iF ig 60: HE Oia ae ee Le G
88-1 Iv 80: 68: LO- ay ae = ras 8G: = G0: f
00] “AON 18q0V0O “‘qydog qsnsny Ajne oun fr Ae judy yoreyy =| Arenaqoq| Arenuee
6€61

. XT. (1940).

“wae! By
a dace
mee a
ets,
‘ei
_

4

‘

RL) eee

e
Ll
7’ ay
‘
*
om
s

ol

r
yee
oo
Pe as

a
a.

eh a

PUBLICATIONS OF THE BOTANIC GARDENS,
SINGAPORE

Ps Pe bP
, Aw .. "
Pte) oa eee bs
oe ay Rea
Re A ne
. i

=

1. The Agricultural Bulletin of the Malay Peninsula (ord ae
series] nos. 1—9, 1891-1900 (out of. print). =e SP
2. The Agricultural Bulletin of the Straits and FMS. .
[Second. Series, monthly issues]. Vols. 1-10, 1901= R
1911. Most numbers are AV aan, price 50 See cag
each or $5 per volume. — 7 Rah ore
3. The Gardens Bulletin, Straits Satilowan te: FE IE Sa ; vs
Vol. 1 nos. 1-5, January to May 1912 yer Apres
tural Bulletin of the Straits and F.M.S., Third
Series). nos. 6-12, December 1918-—Mareh
1917 (as The Gardens Bulletin, S.S- ee a

Vol. 2 nos. 1-12, July 1918—August 1921: PS
Vol. 83 nos. 1-12, August 1923-—March 1925.
Vol. 4 nos. 1-12, June 1926—January. 1929,
Vol. 5 nos. 1-12, August 1929—June 1932. 3
Vol, 6 nos. 1-15, (issued as parts. 1-3) December

1929—October 193025 Bs
Vol. 7 parts 1-3, September 1982—June 1934.
Vol. 8 parts 1-4, October 1934—October 1935. -
Vol. 9. parts 1-4, December. 1935—-March 1938. . <
Vol. 10 parts 1 and 2, January, 1939—August, 1939.

The above former issues. of the Gardens Bulletin may.
be purchased from the Botanic Gardens, Singapore, at $5 ue
per volume or. 50 cents per number for vols. 1-5 (Vol.:6,; of —
15 numbers, $7.50) ; in some cases two or. more numbers ee
were published together, Vols. 7-10 are published at $8. -
per volume;-the prices of parts vary according to their size.
The parts are published at irregular ery as elaatende
is available, : Porch ise. Pee

‘Ue * )
Aro

¢ eran s Bteem ae relivicy

afi i
ay

Notes ‘on ae Aeaiomaty Soak ‘Bannon of Malayan
* Phanerogams, 1V: Ixora by B. J. H. Corner - - = — 127

esicae IX: Two new-Coryphaceous Genera in
lay by C. X. Furtado = 936

a macrorrhiza and its Varieties bo Cx. Picts: =. 244
nification < of Calophyllum calaba 1 bp CX. Fartado

ae

Prunteo 1s tHe Gov RNMENT PRINTING. tore ick: Sixourone,
us eo ee WT: Cunny, GoveRNMENT PRINTER, es

te
ee % =

“9a” a bet eae wee oe
ey - apo Wess ob aReDRE;
migaa oh gli REGrIveD

De oak ee Oe he ee

THE

GARDENS’ BULLETIN

STRAITS SETTLEMENTS

Vol. XI 30th August, 1941 Part 3

“NOTES ON THE SYSTEMATY AND DISTRIBUTION

OF MALAYAN PHANEROGAMS, IV: IXORA

By E. J. H. CORNER,
Botanic Gardens, Singapore

After studying for several years the living plants of
Ixora in Malaya, I find that the two most valuable guides to
the affinity, and therefore to the identification, of the
species are the colour of the corolla and the fragrance or
scentlessness of the flower. After these one must have
recourse to the size of the bracts and sepals and the shape
and venation of the leaf. The size and hairiness of the
corolla and even the shape of the petals I find to be most
variable and, for the majority of the Malayan species, by
no means specific. But the most important character of
the genus is found in the structure of the inflorescence,
whether its branches are articulate or not, because this
criterion distinguishes J. grandifolia, with its inarticulate
inflorescence falling off in one piece, from all other Malayan
Species. It is a primitive mark that connects Ixora with
such genera as Pavetta and Tarenna.

It seems that the extraordinary variability in the size
of the leaf and corolla is not appreciated. From the most
recent works on the genus (Craib, Bremekamp) the impres-
sion 1s gained that many new species can be blocked out
from the old and that minute differences in hairiness, leaf-
size, or shape, and length of corolla-tube suffice to define
them. But I am certain that this view is mistaken and,
if followed, must lead to the making of so many “species”
that the classification of the genus will become impossible.
Such splitting, based on relatively few herbarium specimens,
merely obscures the issue which is to have names for the
major, well-defined groups of individuals. Moreover,
through limitation to the herbarium, many equally variable
and important characters such as the colour of the flowers
and the shininess of the leaves must be omitted, though
from such features the living plants are most easily
recognised. If it is permissible to make species on the

177

178

size of the leaf, so it is on the colour of the corolla e.g. I.
candida and I. lutea. But the more one studies the wild
plants the more one will encounter intermediates between
the microspecies of the “splitters”. For example, a good
case might be put forward for uniting I. chinensis, I.
javanica, I. congesta, I. Lobbii and I. Robinsonii because
there are a few plants which cannot be classified as one
or the other, but can be referred equally well to two of the
species. To the field botanist in Malaya, Ixora is clearly
in a ‘state of evolution’ comparable with that of Rubus,
Rosa, or Oenothera and, systematically, as complicated. We
must, therefore, compromise between the needs of the field-
botanist and the scruples of the herbarium to procure a
satisfactory classification.

To prove my point, I have defined the Malayan species
in the widest possible sense and, under the red-flowered
species, have indicated to some extent their variation by
numbering the recognisable forms. It will be seen that
these show such gradation that none can be separated as a
species e.g. I. javanica with the corolla varying from 2—5 cm.
long, 1-2-5 cm. wide, and from a basic yellow through pink
to deep red. That these continuous variations should have
been overlooked by previous authors must throw grave
doubts on the value of their classifications. Certainly the
keys of King, Ridley and Pitard are impracticable. I believe
that by combining all such variable characters as the size,
shape, texture, colour, shininess and venation of the leaf,
the length of the stipule, petiole and peduncle, the variation
in size of the corymbs and flowers and the colour of the
petals, no less than a hundred ‘forms’ of J. javanica could
be detailed for Malaya alone. And how many of these
characters and combinations of characters would be
hereditary? I think that the main species of Ixora are
known. It is now time for the field-botanist and the
geneticist to explore their variation.

I have not attempted to dissect Bremekamp’s latest
paper on the Malaysian species (Bull. Jard. Bot. Buit. ser. 3,
vol. 14, 1937). I doubt if anyone will be prepared to
struggle through so much detail before the name of a species
can be found. . The classificatory value of the details I
consider so dubious and trivial that I would regard most of
Bremekamp’s species as forms of subspecies and his series
as species, did our study of these forest plants usefully
extend to such fractionising. If so many “species” can be
recognised in the herbarium, what is to become of the genus
when we shall have reckoned also with all the characters
that disappear from the living plants on drying? If the
length of the rudimentary sepals is so important or the

Gardens Bulletin, S.S.

179

excursion of a few epidermal cells on the style into
macroscopic vision, then so are the sliminess of the root-cap,
the rate of mitosis, the latent period of tropism, the glitter
or convexity of the epidermal cells and the exact tint of
the young leaf. I disagree so profoundly with unargued
“splitting’’ that I leave it to others to consult this new
language.

KEY TO IXORA IN MALAYA

Corolla white or greenish white: fragrant : Group A
Corolla, with pink or reddish tube and white or

pinkish petals, not changing colour: flowers

often fragrant: wild .. Group B
Corolla wholly yellow, red or, occasionally, pink,

mostly with yellow or orange petals changing

to red: mostly scentless (all cultivated,

coloured Ixoras) re a 5% Group C

GROUP A

Sepals more or less leafy, at least twice as long
as the ovary
Dwarf shrub with ovate red sepals: wild .. I. clerodendron
Medium to tall shrubs or small trees: sepals
green or pinkish
Sepals 2-3-5 mm. long: leaves rather
narrow and blunt, with 6-9 pairs
of side-veins: cult. os I. finlaysoniana
Sepals 3-12 mm. long: leaves medium
to broad, pointed, with 10-20 pairs
of side-veins

Sepals 3-6 mm. long: wild oF I. umbellata
Sepals 6-12 mm. long: cult.
(? wild) var. multibracteata

Sepals linear, triangular or toothlike, not leafy,
shorter than the ovary or not more than twice
as long (except J. Brunonis)
Stems, infloresc. and undersides of leaves
hairy: veins 10-20 pairs
Dwarf shrub: leaves narrowed to the
distinct stalk: corolla glabrous... I. pumila
Medium-sized shrub: leaves more or
less cordate, sessile or indistinctly
stalked: corolla hairy .. + I. Brunonis
Glabrous: leaves drying blackish
Blade tapered to the base: corolla-tube

8-1-3 cm. long © I. nigricans
Blade generally broadly rounded at the
base: corolla-tube 2-3 cm. long .. var. ovalis
GrRouP B
Inflorescence upside down, the branches jointed:
flowers not fragrant .. , I. pendula

Inflorescence erect or not. bent. ‘down, the
branches not jointed: often fragrant
Blade elliptic, ovate or obovate with distinct

veins I. grandifolia
Blade lanceolate, rather thick, with indis-
tinct veins ae id i var. lancifolia

Mol. XE.) (1941).

-

180

GROUP C

Corolla-tube less than 4-5 cm. long: fragrant:

wild
Flower-clusters 2—5 em. wide: corolla-tube
8-12 mm. long: lowlands :

Clusters 6-20 cm. wide: tube 3-9 mm. long:

mountains
Corolla-tube more ‘hee: 15) em. “ae ae
fragrant

Sepals 2-5 mm. long, longer than the ovary,
rather leafy
Flowers yellow-red: leaves not cordate:
wild she

Flowers pink: leaves often cordate:
cult.

Sepals -5-1-5 mm. long, (aerate! footh:
like, shorter than the ovary
Leaf more or less sessile, the base
narrowly to widely cordate, at least
on the vegetative shoots
Leaf 5-18 cm. wide: infloresc.
branches red: petals. blunt:
wild
Dwarf shrub % m. high:
leaves crowded xz
Spindly shrub 1-24 m. high:
leaves well-spaced ;
Leaf 1-7 cm. wide: infloresc.
branches green: petals pointed:
cult.
Flower red or orange turning
red
Leaves acute: petals
lanceolate-elliptic

Leaves often blunt: petals
ovate-elliptic

Flowers yellow, eas

Petals yellow fading to pale
pinkish white ie

Petals yellow then streaked
with rose-pink

Petals pink, PEAS deep digae
pink

Leaf stalked, the base never aordanek
infloresce. branches mostly red
Leaf-stalk 1-3 mm. long:
cult.
Leaves small, narrow,
-7-2-5 em. wide: petals
pointed be

Leaves rather blunt, 2-2-
6-5 cm. wide: petals
blunt

Corolla 1-4-1-6 cm.
wide, tube . 3-3-5
cm, long

I. concinna

I. micrantha

a. mcsidai se"

I. coccinea var. rosea

I. Scortechinii

var. caulescens

I. coccinea

var. bandhuca
var. lutea

var. decolorans
var. aureo-rosea

var. rosea

I. “Sunkist”
p. 285

I. chinensis

te Gardens Bulletin, SS.

= we A Oy a
AO rn ns in es
Ls = -
. bend

181

Corolla? 1-1-2... em:
wide, tube 1-8 cm.
long

Leaf-stalk 3-20 mm. long:
wild or cult.

Riverside shrubs’ with
lanceolate leaves -7-38
em. wide

Leaves glossy: side-
veins 8-15 pairs,
distinctly inarching
near the margin ..

Leaves rather dull:
side-veins 6-10
pairs, not inarching

Not riverside: leaves 3-15
cm. wide

Mountain shrub of
the Tahan neigh-
bourhood (Ulu Pa-
hang, Trengganu,
Kelantan): leaves
leathery, very gra-
dually tapered into
the stalk: petals
pointed ae

Leaves scarcely lea-
thery: stalk always
distinct from the
blade

Leaves with 14-26
pairs of _ side-
veins, distinctly
inarching
Leaf rather nar-
rowly elliptic to
narrowly — ob-
ovate, acumi-
nate, up to 8-5
cm. wide: side-

veins 14-26
pairs: petals
acute ec
Leaf rather

broadly elliptic,
up to 15-5 cm.
wide: side-veins

13-19 pairs:
petals mostly
blunt

Leaves with 6-14
pairs of _ side-
veins, up to 11-5
cm. wide
Side-veins 6-12

pairs, distinctly
inarching

Vol. XI. (1941).

var. dixiana

I. Lobbiu var.
stenophylla

I. javanica

I. Robinsonii

I. Lobbu

I. congesta

I. javanica

var. retinervia

182

Side-veins not
inarching
Side-veins 9-
14 pairs .. var. multinervia
Side-veins 6-9
(-10) pairs var. paucinervia
Flowers buff-
yellow or’
pale orange
buff mid “Yellow Javanica”
Flowers yel-
low then
clear pink,
never red “Pink Javanica”
Flowers yel-
low turn-
ing reddish
pink or red “Red Javanica”

Il. amoena Don
Gen. Syst. 3, 1824, p. 571.

Typus:— .

Wallich 6121 D, E, F, (Penang): fide Hooker.
Interpretations :-—

Hooker, F.B.1L 3, p. 146.

Pitard, Fl. Gen. Indoch., 3, p. 328.

Craib, Fl. Siam. En., 2, p..149.
Misinterpretations :-—

King and Gamble, Mat. Fl. Mal. Pen. 4, p. 154 as

I. stricta Roxb.

Ridley, Fl. Mal. Pen. 2, p. 94, as I. stricta Roxb.
Synonyms :—

same I acacia var. amoena Ridley, Fl. Mal. Pen. 2, 1922,

p. 94.

From the published descriptions, it is impossible to
see how I. amoena differs from I. javanica. The latest
opinion is that of Craib who states that I. javanica has
longer stipules, but, as he gives not the length of stipule in
either, one cannot appreciate the difference. In I. javanica
I find that the stipular point varies from 1—8 mm. even on
the same plant but that it is generally long in plants growing
in shady places, especially by riversides, and short in plants
in the full open. Hooker says that I. amoena has a laxer
habit and “longer, lanceolate, more membranous leaves”
with more distinct petioles than I. javanica. Pitard copies
Hooker. King and Gamble remark that only a slight
difference in reticulation of the leaves distinguishes /.
amoena from I. javanica (=I. stricta in their sense).
Ridley distinguishes his variety amoena by the more
coriaceous leaves in contradiction to Hooker, though both
cite Wallich’s collection. In the following list of identified
specimens, a still greater variety of treatment will be
discovered. Hence I regard I. amoena as a synonym of I.
javanica.

Gardens Bulletin, S.S. .

185

IDENTIFIED SPECIMENS OF [. AMOENA IN THE
SINGAPORE HERBARIUM

Craib’s identifications :—

Keith 129 (=I. javanica form 1): Kiah 24360 (= IJ.
javanica form 2). Haniff and Nur 3910 (=I. javanica
form 9): Curtis 2607, Ridley 8354 (= I. javanica form
10): Ridley 14998 (— I. javanica form 23): (evidently
Craib has not considered the variation in flower-size).

British Museum’s identification :—

Robinson and Kloss, Korinchi Exp. 19.3.14 (= I. javanica
form 29).

Merrill’s identifications :—

Moulton 6703/15 (= TJ. Lobbu form 3): (almost identical
specimens from the Anamba and Natuna Islands, v.
Steenis 897 and Henderson 20185, are determined by
v. Steenis as I. javanica).

Winkler 2154 (=I. javanica var. retinervia, probably).

I. Brunonis Don

Gen. Syst. 3, 1834, p. 573.

A shrub up to 3 m. high: twigs, corymbs and undersides
of leaves rather closely hispid-hairy, uppersides of leaves soon
glabrous.

Leaj-blade 14-34 * 4-105 cm., lanceolate-obovate, or
nearly oblong, cuspidate-acuminate, the base narrowly to rather
broadly cordate, membranous: side-veitns 11-17 pairs, more or
less inarching near the margin: petiole 0-5 mm. long, very
short, hispid: stipular point 9-20 mm. long, filiform.

Corymb 6-12 cm. wide, generally dense and sessile, or
with a peduncle up to 2 cm. long, the bracts, sepals, branches
and corolla hispid hairy: bracts 5-10 mm. long, filiform:
flowers white, fragrant: sepals 2-5-5 mm. long, filiform: corolla-
tube 3-4 cm. long, the outside of the tube and the mouth hairy:
petals 6-9 mm. long x 2-3 mm. wide, oblong, acute, hairy on
the outside.

Distribution:—Burma, Siam, North Malaya (Penang,
Upper Perak, Kedah).

Type:—Wallich 6136 (Penang).

Very little is known about this species which appears
to be rather rare. The oblong-obovate, cuspidate, hairy
leaf with almost sessile, cordate base is distinctive.

I. chinensis Lam. Chinese Ixora

Encycl. 3, 1789, p. 344.
Synonyms :—
I. stricta Roxb. Hort. Beng. 1814, p. 10: Fl. Ind. 1, 1820,
p. 888: Fl. Gen. Indoch. III, p. 326: haud sensu Hooker,
King, Gamble, Ridley.
Flamma silvarum peregrina Rumph. Herb. Amb. 4, p. 107,
t. 47.
Icones :—
Rumphius loc. cit.
Curtis, Bot. Mag. t. 169 (as' I. coccinea).
Wight, Ic. t. 184 (as I. stricta).
Bot. Reg. t. 782 (as I. crocata).
Typus:—
Sonnerat, spec. sinense, in Herb. Lam.

Vol. XT. (1941).

184

A glabrous shrub, ? flowering all the year in Malayan
countries.

Leaf-blade 4-15-5 & 2-2-6-5 cm., elliptic, lanceolate-elliptic
or narrowly obovate, subacute or blunt, tapering gradually to
the suddenly rounded narrow base or even narrowly cordate,
not amplexicaul, subsessile, more or less coriaceous, rather dull
green often yellowish green, with 6-10 pairs of side-veins not
or indistinctly inarching: petiole -1—-3 cm. long: stipular point
-1—-3 cm. long.

Corymb 5-10 cm. wide, dense, sessile or with a peduncle
up to 1-5 cm. long, with crimson, articulate branches; sepals
blunt to subacute, shorter than the ovary: corolla-tube 3-3-5
em. long, the limb 1-4-1-6 cm. wide, the petals broadly elliptic
or orbicular, very blunt with the edges not or scarcely recurved,
apricot yellow turning brick red.

Distribution:—S. China, Indochina: elsewhere introduced.

I follow Merrill in his interpretation of this species
(Interpr. Rumph. Herb. Amb. p. 487) though he did not
give botanical evidence or a description of the species that he
had in mind. But the small, subsessile, blunt, few-veined
leaves and extraordinarily blunt, broad petals, which make
the flower look very full, readily distinguish it and by this
means one can certainly unite the plants of Rumphius,
Lamarck and Roxburgh. Pitard’s description (Fl. Gen.
Indoch.) corresponds exactly.

Whether I. chinensis is specifically distinct from I.
javanica I do not know. The very short petiole of
I. chinensis appears to be the only diagnostic character.
A critical study of the wild species in Siam and Indochina
(cf. I. amoena sensu Pitard) is much needed. I. crocata
Lindley, which has always been referred to I. stricta sensu
Hooker (=I. javanica), seems to me unquestionably I.
chinensis. Whether the illustration of J. javanica, Bot.
ee t. 4586, represents I. javanica or I. chinensis I cannot

ecide.

On the other hand, J. chinensis has been often mistaken
for I. coccinea and it seems that it still is. The more or
less amplexicaul leaf-base and pointed petals distinguish /.
coccinea, but fragmentary herbarium specimens may not
be easy to identify because the leaves on short flowering
stems may be reduced and have a base more like that of
I. chinensis. Moreover, the garden variety called J. dixiana
is referred in the Index Kewensis to J. coccinea though it
is certainly a dwarf form of I. chinensis. I. dixiana, in
fact, differs from J. chinensis only in its dwarf habit
(14-1 m. high) and small flowers (corolla-tube 1-8 cm. long,
limb 1-1-2 cm. wide). I am inclined to think that the
plant called by the garden name “Sunkist” in this paper
is a hybrid between I. chinensis and I. javanica or I. Lobbw..
I have seen no published description that would fit it.

I. stricta Roxb. sensu Hooker, King and Ridley is /.
javanica. |

Gardens Bulletin, SS.

185 |

{. clerodendron Ridley

Trans. Linn. Soc. 3, 1898, p. 311.

A small, little branched shrub; young parts pubescent.

Leaf-blade 23 & 9 em., lanceolate or obovate, narrowed
to both ends, glabrous above, pubescent beneath: stipules 1-2
em. long, triangular subulate.

Corymb 2-5 em. wide, dense, pubescent, many-flowered:
flower white: sepals 6 mm. long, ovate, pubescent, longer than
the ovary, red; corolla-tube 2-5 cm. long: petals 3 mm. long,
oblong-ovate, subacute, white.

Fruit 6.mm. long, grey with red-streaks, crowned with
the red sepals.

Distribution:—Pahang (Malaya).

Type:—Ridley’s specimen from Kuala Tembeling.

This description is taken from Ridley’s. There is no
specimen in the Singapore Herbarium. The species must
come very close to I. pumila which differs in the smaller,
lanceolate sepals and relatively broader leaves. I suspect

that they may be identical. Both are from the same district
of Pahang.

’

I. coccinea L. . Indian Ixora
Sp: Pl.ard: Ed., 2764, p.. 159. .(exel. synon.).
Synonyms :—
I. grandiflora Br., Bot. Reg. t. 154.
Icones :—

Bot. Reg. t. 154 (as I. grandiflora).
Wight lc. t. 1538 (as I. coccinea).

A shrub or treelet to 4 m. high, flowering throughout the
year, glabrous except the puberulous branches of the corymb.

Leaf-blade 3-5-1383 * 1-7 em., elliptic, varying blunt to
acute, the base more or less widely cordate and amplexicaul,
sessile or shortly stalked, (often narrowly cordate or broadly
rounded and shortly stalked on short flowering twigs), rather
dark shiny green, scarcely coriaceous: side-veins 9-14 pairs,
inarching near the edge: petiole 0-7 mm. long: stipular point
2-7 mm. long.

Corymb 5-10 cm. wide, sessile or with a short peduncle
up to 2-5 em. long, the branches puberulous, articulate, green,
the flowers in trichasia the central one generally sessile, the
laterals shortly stalked: sepals broad, subacute, reddish, shorter
than the green ovary; corolla-tube 2-3-5 em. long, the limb
1-3-3 cm. wide, the petals -2—8 cm. wide, lanceolate-elliptic,
acute, pinkish yellow or reddish pink with orange base turning
blood red or crimson.

Distribution:—N.E. India, China, Indochina; elsewhere
introduced.

Type:—Osbeck’s specimen of the Flora Zeylanica.

var. bandhuca Ixora Bandhuca

Syn. I. bandhuca Roxb. FI. Ind. I,:1820, p. 386.

Wight Ie. t. 149 (copy of Roxburgh’s drawing of I.
bandhuca).

Bot. Reg. t. 513. |

MOLEX, C194 ),

186

A bushier, more spreading shrub: leaves obtuse or sub-
acute: petals 8-10 mm. wide, ovate-elliptic, rather suddenly
pointed, not lanceolate-acute, hence the flower fuller: corolla-
tube 3-5-4 cm. long, limb 2-5-3 cm. wide.

Distribution :——-Hindustan.

var. lutea Yellow Coccinea
Syn. I. lutea Hutchinson Bot. Mag. t. 8439.

Petals buff-yellow, unchanging or fading pale yellow, never
reddening, ovate-elliptic: sepals green: corolla-tube whitish.
Origin:—Royal Botanic Gardens, Peradeniya, Ceylon.

var. decolorans var. nov. Pallid Coccinea

Lamina semper minore, 2-5-7-5 x 1-5-3-8 cm., elliptica,
acuta, basi anguste cordata, (laminis ramulorum floriferorum
saepe vix cordatis vel cuneatis), costis lateralibus utrinsecus
7-12: petiolo 1-2 mm. longo.

Floribus luteis dein roseo-pallescentibus vel albido-roseis;
corollae tubo albido, 2-7—3-1 ecm. longo; corollae limbo 2-2-3 cm.
lato, petalis -6—7 cm. latis, ovato-ellipticis, acutis.

Origin :—horti?

Typus:—frutex in hort. bot. Singap. ‘

/ var. aureo-rosea var. nov. Gold Coccinea

Petalis luteis dein marginem versus roseis vel roseo-striatis,
semper aureo-roseis, nunquam coccineis, late ellipticis, acutis,
7-8 mm. latis: corollae tubo 4-4-5 cm. longo, pallide roseo-luteo,
limbo 2-7-3 cm. lato.

Origin :—horti?

Typus:—frutex in hort. bot. Singap.

‘var. roSea var. nov. ~ Pink Coccinea

2? Pavetta incarnata Bl. Bijdr., 1825, p. 950.

Floribus roseis, primo pallide roseis, nec rubris nec luteis:
sepalis 2-2-5 mm. longis, late lanceolatis, foliaceis, roseo-
marginatis, ovario longioribus: corollae tubo 1-5—2-5 em. longo,
pallide roseis, limbo 1-2-2-3 em. lato: petalis -2—6 cm. latis,
oblongo-ellipticis, acutis vel subacutis, proxime marginibus
recurvatis: ovario 1-1-5 mm. longo, virido: ramulis inflores-
centia viridis.

Origin :—? : cult. in Malaya.

Typus:—Corner 34514 (Woodleigh Nursery, Singapore,
9.11.37 = Form 8).

I. coccinea was, and perhaps still is, confused with I.
chinensis, as I have mentioned under that species. But the
description of Linnaeus indicates how I. coccinea must be
interpretéd, that is “foliis ovalibus semiamplexicaulibus”,
and in this wide sense it is a useful species. I have often
observed that the leaves on flowering twigs incline to the
shape of those of J. chinensis so that such a twig by itself
may be difficult to distinguish from one of J. chinensis
except for the pointed petals. However, amplexicaul
cordate leaves can always be found on sterile shoots of /.
coccinea and never on those of J. chinensis, or indeed, of any

Gardens Bulletin, S.S.

187

other red-flowered species in the sense in which I under-
stand them. I believe also, that one may find that I.
coccinea has always green and I. chinensis red corymb-
branches. |

I. bandhuca seems not to be recognised nowadays.
From dried specimens it is no doubt difficult, but with the
living plants it is easy. The difference from J. coccinea is
well shown in Wight’s two plates which are copies of
Roxburgh’s drawings. The distinctions are, however,
merely varietal. The var. bandhuca is the commonest
variety of I. coccinea in Malaya.

I. lutea is clearly only a colour-variety of typical /
coccinea, in which the red colour does not develop. Hence
I have reduced it to varietal rank. The amplexicaul leaves
distinguish it from the yellow variety of I. javanica.

The var. decolorans is proposed for a bush growing in
the Singapore Botanic Gardens and which has _ been
mistakenly called J. bandhuca for many years. I have not
found a description of it. The washed-out colour of the
flower is not beautiful.

The var. aureo-rosea I propose for another plant
growing in the Singapore Botanic Gardens, and believed
to have been introduced from Kew. I can find no description
of it either. It is a magnificent, free-flowering plant of
considerable vigour for which a name has long been needed.

The variety rosea is probably Pavetta incarnata Bi.
of which there is no modern or ample description. It has
the subsessile amplexicaul, pointed leaves typical of J.
coccinea on the vegetative shoot, but those on the flowering
shoots have generally a simple, rounded base suggesting
that of I. javanica. The sepals offer the only discrepancy
in identifying the variety with I. coccinea, the sepals of
which do not exceed the ovary in length, but a .specific
value can hardly be attached to this feature when the over-
whelming similarity with J. coccinea in other respects is
borne in mind. I do not find that a pink variety of [.
coccinea has been described and as this is a frequent plant
in Malayan gardens it is in need of a name. If J. rosea
Wallich (Bot. Mag. 2428, Bot. Reg. 540) is correctly
identified with I. stricta sensu Hooker (= I. javanica) then
it cannot be my var. rosea, though t. 2428 is very suggestive
of it, excepting the pubescence.

What I have called J. coccinea form 4 has been grown
for several years in the Singapore Botanic Gardens as I.
oculata, which name I cannot find in botanical literature.

As I have remarked under I. chinensis, the garden
species I. dixiana is a variety of it and not of I. coccinea
as stated in the Index Kewensis.

Vol. XT, (1941).

188

KEY TO THE FORMS AND VARIETIES OF I. COCCINEA

Petals red
Petals lanceolate a ie attenuate, -2—--8 cm.
wide 4 ” I. coccinea
Corolla-tube 2-2-5 ‘cm. long
Limb 1-3-1-6 cm. wide: uta -2—3 em.
wide — fe Form 1
Corolla-tube 3-3-5 em. long
Limb 1-5-1-7 cm. wide, petals -4—5 cm.

wide es a Form 2
Limb 2-2-3 em. wide: petals -5—6 cm.
wide ty ag Form 3
Limb 2-7-3 em. wide: petals -6—-8 cm.
wide Form 4
Petals ovate elliptic, rather suddenly pointed,
‘8-1 cm. wide : var. bandhuca
Petals yellow, never reddening var. lutea
Corolla-tube 3-2-3-7 em. long: limb 2-4-2-7 em.
wide: petals -6—-8 cm. wide Form 5
Corolla-tube 4-4-5 cm. long: limb 2-7-3 em.
wide: petals -8--9 cm. wide we a Form 6
Petals yellow, fading pale pinkish white i var. decolorans

Petals yellow then streaked with rose, rather broad var. aureo-rosea
Petals pale pink turning deep pink, never red or

yellow, oblong elliptic .. : -c var. rosea
Corolla-tube 1-5-1-8 em. long: limb 2-2-3 em.
wide: petals -5—-6 cm. wide, subacute i Form 7

Corolla-tube 2-2-5 cm. Jong: limb 1-2-1-5 em.
wide: petals -2--3 cm. wide, acute with a
minute mucro 7m ap Ke Form 8

I. concinna Hook. f. Trogon Ixora
EBL 3, 2880; 6: TAT:

A small to medium-sized tree up to 20 m. high: wholly
glabrous.

Leaf-blade 6-5-18 x 2-7 cm., elliptic, tapered to each
end, more or less acuminate, membranous or subcoriaceous:
side-veins 8-13 pairs, distinctly inarching: petiole -2-1-2 cm.
long.

Corymb 2-5-5 em. wide, sessile or on a short peduncle up
to 1-2 em. long, the branches slender 1 mm. wide or less,
articulate, glabrous (? red or green): sepals -5-2-5 mm. long,
lanceolate-triangular, variable in size, shorter than the ovary,
as long: as or longer than in different collections, never leafy:
corolla-tube 8-12 mm. long, short, yellowish pink to reddish
pink: corolla-limb -7—1 cm. wide, the petals acute, yellow turning
pinkish and -finally duli-red, or partly red: ovary 1-1-5 mm.
long: flowers fragrant.

Fruit -7—1 em. wide.

Distribution:—Malaya, Sumatra, Borneo.

Type:—Wallich 6149 (Singapore).

This is one of the less variable Ixoras. It is distinguish-
ed by its small leaves and flower-clusters and its fragrant
flowers. The variable features are the texture of the leaf
and the length of the sepals. The leaf may be membranous
or distinctly coriaceous, perhaps dependent on the size of
the tree and its situation. The sepals may be shorter than

Gardens Bulletin, SS.

189

the ovary or considerably longer or of intermediate length,
their length being fairly constant in each collection.
Although this variation seems to have escaped the attention
of botanists, so that we have been spared from numerous
micro-species which would have made the identification of
the trees impossible until one had seen the sepals, the range
of variation is greater than that of any other species, being
5 times the minimum length of the sepals.

I. concinna has been mistaken in the Singapore
Herbarium and in Ridley’s Flora for my I. javanica var.
retinervia from which it differs in its smaller, fragrant
flowers, less pointed leaves with the veins making a more
acute angle with the midrib and not so distinctly inarching,
its laxer and more slender inflorescence branches (2 mm.
wide in I. javanica var. retinervia) and smaller fruits (1-1:3
cm. wide in var. retinervia).

I. congesta Roxb. Malayan Ixora
Fl. ind. +1: 138205 ps 39T.
Synonyms :—
I. Griffithii Hook. Bot. Mag. 1847, t. 4825.
I. fulgens Roxb. sensu Hook. fil. F.B.I. 3, 1880, p.. 146.
An evergreen shrub or small tree like J. javanica but:—
Leaf-blade generally larger or broader, 13-31 5-15-5 ecm.
wide, with more numerous, stronger lateral veins (13-19 pairs)
generally with distinctly inarching marginal loops.
Sepals blunt to acute, generally acute.
Petals yellow or pinkish yellow turning red, generally blunt
to subacute.
Distribution:—W. Malaysia to the Moluccas.
Bremekamp has recently substituted the name I.
Grifithu for Malayan plants which have been referred to
I. congesta (Journ. Bot. 75, 1937, p. 169). Roxburgh
described I. congesta from a plant growing in the Calcutta
Botanic Gardens and said to have been derived from the
Moluccas. There is no type and the description is brief.
Hence Bremekamp writes of it “species moluccana dicta,
parum descripta et absentia typi haud certe noscenda,
dubiosissima”’. Why then did Hooker reduce I.. Griffithii
to I congesta, in which he was followed by Kurz, King,
Gamble, and Craib? I think that Bremekamp has over-
looked the traditions of Indian Botany which were handed
from Roxburgh to Hooker through the succession of
Superintendents of the Calcutta Botanic Gardens and
through Roxburgh’s own circle of botanists, of which
Wight and Arnott have given us the picture. As no reason
is advanced for rejecting this long-standing interpretation
of J. congesta, one must continue to employ it in accordance
with the Rules of Nomenclature. J. fulgens offers a parallel
case where the full description of Rumphius should enable

~ Vol. XT. (1941).

190
one to identify Roxburgh’s species, but Bremekamp denies
the Moluccan origin of J. fulgens.

Merrill has identified J. congesta with I. fulgens but,
as I have shown under that species, nobody has yet identified
I. fulgens (Enum. Philipp. Fl. 3, p. 548). Rumphius’ plate,
which is the type of I. fulgens, shows an Ixora with pointed
petals. The only form of J. congesta with pointed petals
is my form 9 and that most botanists would probably not
admit as I. congesta:

Although it is generally easy to distinguish J. congesta,
I. Lobbi and I. javanica, there are collections in the
Singapore herbarium which connect them as I have
mentioned under those species. What I have called J.
congesta form 9 is especially problematic.

KEY TO THE FORMS OF I. CONGESTA

Corolla-tube 3-7-4.5 cm. long
Limb 1-5-1-8 cm. wide: petals -4—-5 cm. wide, blunt to
subacute ye Form 1
Limb 1-8-2-1 em. wide: petals 5—6 em. wide, ‘subacute - Form 2
Corolla-tube 2-5-3:8 em. long
Limb 1-1-3 em. wide

Petals -3—-4 cm. wide, subacute BR i Form 3

Petals -5--7 cm. wide, blunt 2 Form 4
Limb 1-4-1-8 em. wide: petals blunt to subacute

Petals -3--5 em. wide ’ ae Form 5

Petals -5--7 cm. wide iM “a yf: Form 6
Limb 1-8—-2:1 em. wide

Petals -3--5..ecm. wide, subacute ; de Form 7
Limb 2-2-8 cm. wide: petals -5—-7 cm. wide

Petals blunt to subacute a 4 oe Form 8

Petals acute re Form 9

Limb 2.5-3-8 em. wide: " petals blunt to subacute
Petals -5--7 em. wide, orange yellow then red: limb

2-5-3-8 em. wide: veins of leaf 9-14 pairs .. Form 10
Petals -7--8 cm. wide, pale yellow then salmon pink:
limb 2-5-3 cm. wide: veins of leaf 14-19 pairs .. Form 11

Form 1
Corolla-tube 3-75-4-5 em. long, limb 1-5-1-75 em. wide:
petals -35 cm. wide, blunt to subacute.

Malayan Collections

Perak; Derry 10716, Henderson 10202.
Johore; Corner 39296, s.n. 1.9.35 (Kota Tinggi-Mawai gow
Singapore; Cantley s.n., Ridley 4168 (det. J. Lobbii),
9.11.99 (Chua Chu Kang).
Form 2 :
Corolla-tube, 3-75-4-25 em. long, limb 1-75-2-1 em. wide:
petals -5-—-6 cm. wide, subacute.

Malayan Collections
Malacca; Goodenough 1907.
Form 2
Corolla-tube 3-3-75 cm. long, limb 1-1-25 em. wide: petals
-35—45 em. wide, blunt to subacute.

Gardens Bulletin, SS.

yas

. Form

Form

191

Malayan Collections

Penang; Curtis 2975.

Perak; Cantley s.n., Kunstler 675, Wolferston s.n. April 1900,
Wray 702.

Pahang; Henderson 24432.

Selangor; Ridley 13409.

Johore; Corner 25968, Ridley s.n. 1900 (Batu Pahat), Singapore
F.827.

Singapore; Cantley 2668, 2672, 3005, Goodenough s.n. 1892
(Pulau Ubin) Ridley 5675, 6911, s.n. 1890 (Tg. Penjuru).

4
as Form 3 but petals -5—-65 cm. wide.

Malayan Collections
Singapore; Cantley/s.n. (det. J. stricta).

2
Corolla-tube 3—3-75 cm. long, limb 1-5-1-75 em. wide: petals
-35—-5 em. wide, blunt to subacute.

Malayan Collections

Penang; Curtis 667, 1729, 2257 (det. J. Lobbi by King &
Gamble).

Perak; Curtis s.n. Feb. 1900 (Lumut), Henderson 102038, Ridley
sn. Feb. 1892 (Lumut).

Pahang; Nur 18786.

Selangor; Goodenough 10481, Hume 7162, 7607.

Negri Sembilan; Holttum 9691, 9768, Nur 11570.

Malacca; Alvins s.n., Burkill 788 (det. I. fulgens), Derry 8,
Goodenough 1692.

Johore; Corner 29766, Kiah 32326, Ridley s.n. 1894 (S. Bau).

Singapore; Hullett 118, Ridley s.n. Reservoir Woods, s.n. Oct.
1889 (Chan Chu Kang).

The flowers are pinkish yellow on opening, turning to

light reddish orange. It seems that there may be states

with

Form

Form

persistently pink flowers.

6
as Form 5 but petals -5—-7 cm. wide.

Malayan Collections
Penang; Curtis 2265 (see also Form 8).

é
as Form 5 but corolla-limb 1-75-2-1 cm. wide, petals
subacute.

Malayan Collections

Perak; Henderson 10116.
Malacca; Goodenough 1692.

Form 8

Vol.

as Form 5 but corolla-limb 2—2-5 cm. wide, petals -5--65 cm.
wide, blunt to subacute.

) Malayan Collections

Penang; Curtis 2265 (see also Form 6) ee I. fulgens by
King).
Selangor; Hume 7577.

MLE CL9LL

192

Form 9 Ixora Singapore

Leaves as in I. congesta but rather narrow 4-28 5-9 cm.,
with 12-18 pairs of lateral veins.

Sepals rather blunt.

Corolla-tube 2-5-3-5 em. long, the limb 2-3-2-8 em. wide:
petals -5—-6 cm. wide, ovate-lanceolate, acute, opening pinkish
buff with only the base clear orange yellow, turning deep
carmine-red or crimson.

Malayan Collections
Singapore (cult. in Hort. Bot. Sing.); Corner 30794, 33699,
Nur 22, s.n. 26.9.18, Ridley s.n. 1891.

It seems that this Ixora was received from Kew many
years ago as I. macrothyrsa. But the plants in the
Singapore Gardens have lost their labels and there is no
certainty whence they came. Instead, the name I.
macrothyrsa has for many years been given to the Common
Red Ixora of our Gardens which I call J. javanica form 7.

Whether 7. congesta form 9 is I. macrothyrsa I have no
means of ascertaining. According to description (Bot.
Mag. t. 6853) the flowers of I. macrothyrsa are persistently
red, but the description is not sufficiently critical.

This form 9 has the leaf-shape and size of I. javanica,
the pointed petals of some forms of I. javanica as well as
of J. Lobbii, but the numerous veins of J. congesta. Its
Systematic position is problematical yet demonstrative:
because of the veins I classify it under J. congesta. It
corresponds remarkably well with I. fulgens as that species
must be interpreted from Rumphius’ plate and description.

Unlike the common I. javanica form 7, this form 9 is
strictly seasonal in flowering like the wild plants of J.
congesta, I. Lobbii and I. javanica.

Form 10 Ixora Kemaman
Leaves with only 9-14 pairs of side-veins.

Corolla-tube 3-25-4 cm. long, the limb 2-5-3-8 em. wide:
petals -5—65 cm. wide, oblong, blunt.

Malayan Collections
Trengganu; Corner 30002 (Kemaman).
Pahang; Symington 28825 (Gunong Tapis), 28761 (Sungei
Lembing).

This magnificent Ixora with large, brilliant flowers
appears to be limited to the watersheds of the Kuantan and
Kemaman rivers. It should be classified with J. javanica
multinervia according to my criterion of the number of
veins but the dark brown colour of the dried leaves and
their coarse, conspicuous blackish venation is identical with
that of I. congesta so that I am compelled to regard it as an
aberrant form of this species. From the large flowered
forms (7, 8) of J. Lobbii it differs in the elliptic, simply
pointed leaves (not lanceolate-obovate and acuminate) the
fewer, coarser veins and the blunt petals.

Gardens Bulletin, S.S.

Pi 193

Form 11
Corolla-tube 2-5-3-8 em. long, limb 2-5-3 cm. wide, petals
-6—-75 em. wide, blunt to subacute, pale yellow turning light
pinkish red.

Malayan Collections
Singapore; Corner s.n. 25.1.37 (Reservoir Jungle).

I have noticed in Singapore that plants of this form
generally flower after those of others forms of J. congesta
at each seasonal flowering.

Sterile or Fruiting Specimens (not flowering)

Malayan Collections

Trengganu; Moysey 33308.

Perak; Cantley s.n.

Pahang; For. Dept. 15139:

Malacca; Alwins s.n., Goodenough s.n. 1892 (Tebau).

Johore; Holttum 9221, Lake & Kelsall s.n. 20.10.92, (K.
Sembrong) Ridley 11164, s.n. 1900 (Minyak Buku).
Singapore; Cantley s.n., Burkill 709, 714 (det. J. fulgens),
Hullett s.n. 1893 (Changi Road), Ridley 8410, s.n. 1889
(Selitar), s.n. 9.11.99, (Chua Chu Kang), ganz L90L >CY¥o

Chu Kang).

I. finlaysoniana Don Siamese White Ixora

Gen. Syst. 3, 1834, p. 572.

A shrub, commonly becoming a small tree 5-6 m. high with

_ stout trunk as thick as a man’s leg and massive old branches:
glabrous except for the very sparsely puberulous branches of
the cyme.

Leaf-blade 6-16 2-65 em., rather small, lanceolate
elliptic to lanceolate obovate or almostly spathulate, varying
blunt to simply acute, the base very gradually attenuate, rather
leathery: petiole 4-8 mm. long, short: stipular point 1-2-5 mm.
long, short.

Corymb 6-11 cm. wide, sessile or with a peduncle up to
1-5 cm. long, the branches articulate, pale green, dense, and
sparsely puberulous, flowers white, scented: bracts 6-12
1-5-2-5 mm. linear-lanceolate, pale green: bracteoles 2-5
-O-1 mm. linear: sepals 3-5 * 1-5-2 mm., oblong, pointed or
rather blunt, glabrous, green-veined, longer than the ovary
(1 mm. long): corolla-tube 2-3-5 em. long, glabrous, the mouth
also glabrous, pale greenish: petals 5-7 > 2-5-3-5 ‘mm.,
at first acute, becoming blunt with recurved edges, white: ovary
often pinkish.

Distribution—India, Indo-China, Siam.

Type:—Wallich Cat. 6166.

This species is easily recognised from its tree-like
habit, its narrow, often blunt, leaves with rather few veins
and short petiole tas int; javanica), the white fragrant.
flowers, linear bracteoles and oblong green-veined sepals
exceeding the ovary in length. It is frequently cultivated
in India, Indo-China, Siam and Malaya, and its native
country appears to be Siam.

Vol. XI. (1941).

194

Specimens from the plants cultivated in the Singapore
Botanic Gardens were compared recently with the
Wallichian type of I. finlaysoniana by Mr. M. R. Henderson
who reported that they matched very well. The Malayan
plants also agree with Pitard’s description of the species
except that Pitard describes the leaves as larger, with more
pairs of lateral veins (10-12) and the corolla with short
tube (19-21 mm. long) and smaller petals (4-5 x 2-25
mm.) (FI. Gen. Indoch. 3, p. 312). But, as in other Ixoras,
[ find that the flowers vary considerably in size in different
collections, the larger-flowered forms being always the more
desirable for cultivation. Rea

The species has been received from India for cultivation
in Singapore as I. stricta var. alba which it cannot be
because of the long sepals and bracts and the fragrant
flowers. I think it has yet to be proved that there is really

a white-flowered variety of any red-flowered Ixora.

1. fulgens Roxb.

Hort. Beng. 1814, 10. (name and citation of Flamma

silvarum of Rumph. Herb. Amb. 4, p. 105, t. 46).
Typus:-—-lamma silvarum of Rumphius loc. cit.
Interpretations :—

Roxburgh, Fl. Ind. 1, 1820, p. 387:

Wallich, Cat. 6152.

Wight, te Pb. iat.) fob.

Miquel, Fl. Ind. Bat. 2, 1857, p. 264 (as Pavetta fulgens).-

Hooker fil., F.B.I. 3, 1880, p. 146.

Koorders and Valeton, Bijdr. Booms. Java 8, 1902, p. 158.

King and Gamble, Mat. Fl. Mal. Pen. 4, 1904, p. 79.

Merrill, Interpr. Rumph. Herb. Amb. 1917, p. 487: Enum.

Phil, “PES, p. 546.

Pitard, Fl. Gen. Indoch. 3, 1924, p. 325.

Craib, Fl. Siam. Enum. I, 1934, p. 157.

Bremekemp, Journ. Bot. 75, 1937, p. 111: Bull. Jard. Bot.

Burt.’ ser.+3, 14, 1987, paskiss
From Roxburgh’s description in the Flora Indica it is

clear the binomial J. fulgens was made for a plant cultivated
in the Royal Botanic Gardens at Calcutta, the plant having
been introduced from the Moluccas. There is no specimen
of Roxburgh’s plant but a drawing of his which is that
reproduced by Wight. There is also a collection of Wallich’s
6152 which Wallich, who was Roxburgh’s successor,
identified with J. fulgens, and which was evidently seen and
passed as such by Wight. Therefore, though Roxburgh’s
plant may have disappeared, it should be possible to deter-
mine what Roxburgh intended for I. fulgens from his
drawing and from Wallich’s specimens. Unfortunately,
according to the Rules of Botanical Nomenclature, neither
of these clues indicate what, taxonomically, must be intended
by I. fulgens. The first publication of this binomial is in

Gardens Bulletin, S.S.-

195

Roxburgh’s Hortus Bengalensis. This is accepted as a valid
publication and, because a species can be validated merely ~
by citation of a plate, Roxburgh’s binomial J. fulgens which
is without description is validated by reference to Rumphius’
plate. Hence, taxonomically, the identity of J. fulgens hangs
upon the identification of Rumphius’ Flamma silvarum,
about which no more is known than what Rumphius has
written. Nevertheless, several interpretations have been
made which I will now elaborate.

Roxburgh’s own interpretation by his description and
drawing, reproduced by Wight, is the first. It has been
denied recently by Merrill and Bremekamp, after a century’s
acceptance, though I cannot discover that it has ever been
amplified or critically examined in the light of modern
systematics. Wallich’s interpretation, by his collection
6152, has never been published, there being no description
of his specimen: I imagine that his interpretation must
coincide with Roxburgh’s. Hooker’s interpretation com-
bines that of Roxburgh’s and Wallich’s (by quotation), with
I. Lobbu by citation and by specimens (according to King
and Gamble) : it seems that the Ixoras which botanists now
call I. Lobbiui, Hooker called I. fulgens. Koorders and
Valeton follow Hooker and make Pavetta salicifolia Bl. into
a variety of J. fulgens in their sense (=I. Lobbi). King
and Gamble deny Hooker’s interpretation and, distinguish-
ing I. Lobbit from I. fulgens, refer to I. fulgens a narrow-
leafed Ixora which appears to be my I. javanica retinervia
(at least, according to the collections Wray 3024 and
Scortechini, cited by King and Gamble, and according to
their description). Merrill (1918) arrived at no decision
concerning the identity of J. fulgens, though he named
as such the Amboina collection, Robinson 169: there is no
description of the collection so that Merrill’s interpretation
is hid, like Wallich’s, until the specimens be examined
and described. Later, 1923, Merrill identified, without
advancing reason, J. congesta Roxb. with I. fulgens Roxb.,
thus refuting Roxburgh’s own interpretation: (the acute
petals shown in Rumphius’ plate ill suit the blunt ones of
I. congesta described by Roxburgh). Pitard’s interpreta-
tion is a copy of that of King and Gamble (‘teste the
misquotation of Wight’s plate). Craib also follows King
and Gamble and identifies a Siamese collection with I.
fulgens. Most recently Bremekamp has given the problem
a wholly new turn by remarking that “Roxburgh described
this species [J. fulgens] from a plant cultivated in the
Botanic Gardens, Calcutta. It was supposed to have been
introduced from the Moluccas, but this was obviously a
mistake, probably due to a superficial resemblance to the
plant described and figured by Rumphius under the name

Vol. XI. (1941).

196

Flamma silvarum”. Further, Bremekamp adds for the
distribution of J. fulgens “Known with certainty from
Tenasserim only”; and he seems to doubt Craib’s identifica-
tion of the Siamese plant. As regards Bremekamp’s first
point, if it is true, then he has discovered some information
hitherto unknown to botanists, even to Roxburgh himself,
which will give a different value to Roxburgh’s interpreta-
tion, yet one cannot except such a bare contradiction without
proof. But from Bremekamp’s second point, it is clear that
he has misunderstood the taxonomic intricacy of the identity
of I. fulgens which can be determined, if at all, only from
the Amboina-flora. Moreover, like that of Merrill and
Wallich, his interpretation of J. fulgens must remain a
mystery to botanists who cannot examine the specimens
cited, because there is no description. (Such interpretations
seem to me valueless and of the same category as nomina
nuda: the essence of systematic botany being to classify
plants with sufficient accuracy that other botanists may
recognise them from the published descriptions, hence
nomina nuda are illegal, and so, too, should be interpreta-
tiones nudi: the mere citation of specimens or expression
of opinion should be as _ insufficient to validate the
interpretation).

It must be remarked, now, that from King and Gamble
onwards there has passed unnoticed a discrepancy between
the description of I. fulgens given by King and Gamble and
the descriptions of Roxburgh and Rumphius. King and
Gamble give the leaves of I. fulgens as 314-514 & 114-114”:
Rumphius gives 8-12 2-314”, for Flamma silvarum: and
Roxburgh gives 6-8 < 1-3”. I think that it is extremely
improbable that J. fulgens sensu King and Gamble, with
such small narrow leaves, should be identical with Flamma
silvarum of Rumphius. Hence, it is evident that the inter-
pretation of J. fulgens by King and Gamble and also those
by Pitard, Craib and Bremekamp, who cite King and Gamble
are doubtful. Bremekamp’s interpretation is rendered yet
more uncertain because he has not followed up the clue to
Roxburgh’s interpretation which the Wallichian specimens
must afford.

I. fulgens Roxb. has therefore been interpreted as J.
Lobbu, I. congesta and I. javanica, the three common species
of Western Malaysia, from which I conclude that it is still
an unknown species awaiting identification by a resident
of Amboina. However, I may give my own interpretation
which is that my I. javanica form 7 is the Rumphian plant,
for the following reasons. |

There is no record how this form 7 which is the
Common Red Ixora of Malayan gardens, was introduced to

Gardens Bulletin, SS.

197

Malaya but it has evidently been in cultivation for very
many years. It agrees so well with Roxburgh’s description .
and drawing as well as with Rumphius’ that I can find no
discrepancy: it also has large leaves, 4-11 * 134-434” and,
as Roxburgh remarks, it blossoms most part of the year.
Wallich’s collection 6152a came from Penang and/or
Singapore; it agrees presumably with Roxburgh’s drawing;
it may even have been our Common Red Ixora because there
is no other common big-leafed Ixora with red, pointed
petals in Penang or Singapore. Now, if Roxburgh’s plant
of J. fulgens came from the Moluccas, is it not likely that
this same species was brought by the ear ly travellers: to
other countries and to what place more likely than Penang,
especially on its way to India? Rumphius called his other
red Ixora, Flamma silvarum peregrina, for the very reason
that it had been widely dispersed and cultivated by travellers
so early as the seventeenth century. If this be so, then
we Shall have the explanation of the origin of our common
garden plant and it will be necessary to reduce /. javanica
to I. fulgens.

Some Collections determined as I. fulgens

Merrill 1362 (Palawan) is J. Lobbit var. stenophylla

_ form 10.

Moulton 78 (Borneo) is J. Lobbu form 8.

Ridley 8970 (Sumatra) is J. Lobbw form 5.

Wray 3024 (Perak) is J. javanica var. retinervia.

I. grandifolia Zoll. et Mor. Pink River Ixora
Syst. Verz. 1846, p. 65 (= Pavetta macrophylla Bl.).
Synonyms :—

Pavetta odorata Bl., Bijdr. 1825, p. 952 (haud JI. odorata
Spreng., Syst. Veg. I. 1825, p. 407).

Pavetta macrophylla Bl., Bijdr., 1825, p. 953 (haud
I. macrophylla DC. Prodr. 4, 1830, p. 487).

Ixora Blumet Zoll. et. Mor., Syst. Verz. 1846, p. 65.
C= /F..odoratg Bi.) .

2? Pavetta Teysmanniana Miqu., Fl. Ind. Bat. 2, 1857,
p. 267 (fide Hooker fil, F.B.I. p. 143).

? Pavetta macrocoma Miqu., loc. cit. p. 274 (fide Hook.
fib loc.: cit.).

? Ixora arborescens Hassk., Retzia I, 1858 p. (? 22) (fide
eer Aa loc. cit., et Koorders et Valeton, Bijdr. Booms. 8,
p. ; .

Ixora macrosiphon Kurz., Journ. Bot. I, 1875, p. 327 (fide
Hook. fil., loc. cit.).

Ixora rosella Kurz., For. Fl. Burma II, 187, p. 28 (fide
Hook: fil, loc.-cit.).

Ixora odorata (Bl.) K. et V., Bijdr. Booms. Java, No. 8,
1902;:p. 165.

Ixora coriacea Ridley, Journ. St. Br. R. As. Soc. 79, 1918,
p. 83: Fl. Mal. Pen. II, 1923, p. 98 (= I. elliptica Ridley).

Ixora crassifolia Ridley, loc.-cit..p, $3: Fl: Mal. Pen: IT,
p. 98. (haud I. crassifolia Merrill, 1910).

Ixora fluminalis Ridley, loc. cit, , 0427 91,- Mal. Pen. I,

Vol.. XI. (1941).

198

p: 97.

Ixora gigantea Ridley, loc. cit p. 84.

Ixora elliptica Ridley, Fl. Mal. Pen. II, 1923, p. 98.
Ixora Ridleyi Merrill, Gard. Bull. 8, 1935, p. 1382.

Ixora lancifolia Ridley, Journ. Bot. 72, 1934, p. 256.
Ixora Valetonii Hochr., Candollea 5, 1934, p. 261.

Varieties

var. coriacea Hook. fil., F.B.I., III, 1880, p. 143.

var. arborescens Hook. fil., loc. cit.

var. Kurzeana Hook. fil., loc. cit.

var. rosella Hook. fil., loc. cit.

var. gigantea King and Gamble, Mat. Fl. Mal. Pen. 4,
1904, p. 156.

var. acutifolia Hochr., Candollea 5, 1934, p. 257.

var. lancifolia var. nov.

A shrub or tree up to 20 m. high, often flowering as a
spindly bush 14%2-5 m. high: twigs pale, greyish white or silvery
buff: glabrous except the puberulous panicle and often puberu-
lous style.

Leaf-blade 10-32 « 4-18 cm., rather large to very large,
elliptic, often inclined to ovate, the apex blunt or acute, the
base cuneate, rounded or even cordate, more or less coriaceous;
primary lateral veins 6-16 pairs, directed obliquely forward,
inarching near the edge, with coarse faint reticulations, the
venation generally visible on both sides of the dried leaf:
petiole -6—-3-5 cm. long, stout: stipules 5-12 mm. long, broadly
triangular with a minute point, generally more or less
sheathing.

Corymb 12-23 cm. wide, erect, sessile, with 3 main, long-
stalked branches from the base, the peduncles of the branches
generally 5-9 cm. long, the branchlets not articulate, green
flushed reddish, the: corymb falling entire: flowers crowded,
some sessile, others shortly stalked, fragrant or not: sepals
shorter than the ovary, blunt to subacute: corolla-tube -6—3 cm.
long, pink or reddish: petals 4-5-5 mm. long, 1-5-2-5 mm. wide,
oblong, blunt: or subacute, reflexed along the tube, not spread-
ing, white, often tipped pink: anthers yellow: style projecting
3-5 mm. glabrous or the part in the corolla-tube more or less
hairy, white.

Berries 10-12 mm. wide, subglobose, distinctly longitudi-
nally suleate when dried, green turning red then black.

Distribution :—Burma, Indo-China, Andaman and Nicobar
Islands, Malaya, Sumatra, Java, Borneo, Philippines, ? Celebes.

Vvar. lancifolia var. nov. Marsh Ixora

~

Frutex vel arbor minor, vix ramosus, usque 5 m. altus.

Lamina 20-36 X 3-5-12 em., lanceolata, linearis (30
3-5 cm.) vel lanceolato-obovata, obtusa vel subacuta, wtrinque
gradatim attenuata praecipue basim versus, carnoso-subcoria-
cea, subtus sicco minute papilloso—rugulosa, nervis reticulatis
haud distinctis, costis lateralibus utrinsecus 10-14.

Corymbus 8-16 cm. latus, semper minor, ramis 3 breviori-
bus pedunculo 1-5-5 cm. praeditis: corollae tuba 12-16 mm. vel
17-20 mm. longa.

Distribution :—Malaya, Sumatra.

Typus:—Ridley 2213 (Tahan River, Pahang).

Gardens Bulletin, SS.

£99

That I. grandifolia is a variable species can be judged
from the list of synonyms, as well as from the number of
forms which I have distinguished. As it grows in Malaya, ”
the species could be split into six or seven according to the
inadequacy of the herbarium material, for one sheet typical
of each form might be considered a species. However,
from the numerous collections in the Singapore Herbarium
and from the numerous plants which I have studied in the
forest, I have discovered that these forms are by no means
clearly defined and that with our present knowledge the
subdivision of J. grandiflora is impossible. Indeed, in the
wide sense, it is easily and conveniently recognised as a
common species of swampy forest in Western Malaysia.
In this sense, the species was understood by Hooker, King
and Gamble; and so too by Koorders and Valeton, except for
their separation of J. odorata (Bl.). Recently, for reasons
which appear to me wholly inadequate as well as conflicting,
Ridley has divided the species into six. As in the red-
flowered Ixoras, the variability of J. grandifolia is not yet
appreciated.

For the legitimate name of the species in this wide
sense I have adopted I. grandifolia. The alternative,
simultaneous name J. Blumei I have made the synonym
because it is not well-known while the first is. Both these
names of Zollinger and Moritz were proposed for Blume’s
two species without any reason being given, though evidently
because Blume’s specific names would make later homonyms
._ in the genus Ixora—a point overlooked by Koorders and
Valeton in their J. odorata (Bl.). For the interpretation of
the species of Blume, Miquel, Hasskarl and Kurz, I have
followed Hooker, Koorders, Valeton, King and Gamble,
having no access to the types. Ridley’s species I have been
able to examine personally in Singapore.

As regards I. grandifolia, in the wide sense, the
variability of the leaves in shape and venation has not been
properly explained and it seems that the variability in the
length of the corolla-tube, in the fragrance of the flowers
and in the hairiness of the style is not realised at all: there
is confusion also about the colour of the flower. That some
trees have pointed and others blunt leaves I am certain
from the numerous trees of my form 3 which I have studied
in the East of Johore, but some collections are intermediate.
The number of veins is also a fairly constant distinction but
_also imperfect through the occurrence of specimens with

9-12 pairs of lateral veins: unfortunately, because of
the large size of the inflorescence, herbarium-specimens
generally have only one or two pairs of leaves on the twig
below the panicle and these uppermost leaves of the twigs

Vol. MI. 6iG9 41).

200

are commonly reduced. As for the leaf-base it varies
greatly: even on the same plant, I have seen leaves with
gradually tapered, cuneate, rounded and cordate bases, so
that one unacquainted with the living trees might well make
four species from separate leaves of the same plant.
Concerning the texture of the leaf, I think there is great
variability which cannot be used satisfactorily in defining
varieties because of the numerous intergrades, with the one
exception of var. lancifolia in which the texture is accom-
panied by other peculiarities. That the length of the
corolla-tube is no sign of specificity is obvious from the fact
that there are specimens differing only in this one character.
And, likewise, regarding the hairiness of the style which
I find to be as often rather sparsely puberulous as glabrous:
(this point cannot be determined without pulling the style
from the flower because it is the part of the style in the
corolla-tube which is hairy). The colour of the flower may
be described generally as white or pink, according as the
corolla is open or not. When the flowers of a corymb are
not yet open they look pink because the reddish pink corolla-
tubes are conspicuous: but when the flowers open, the white
petals are reflexed, concealing the corolla-tubes almost
completely. Thus, I believe, have arisen the ‘conflicting’
statements of the flowers as white or pink, just as with
I. pendula. The flowers are generally very fragrant but
I have seen several trees in Johore (my form 3, as 32277)
which had scentless flowers. I think this as important a
varietal character as any other in the species, though it is —
one that will not commend itself so long as we are limited
to dried specimens.

As regards I. odorata (Bl.) K. et V., the authors of the
combination say that from I. grandifolia it is “toto coelo
diversa (N.B. fructus valde diversi)”. Nevertheless I am
unable to follow them. The fruit of J. grandifolia they
describe as 6-12 mm. wide and distinctly divided in 2 lobes,
or didymous: that of J. odorata they describe as only 6-7
mm. wide and long, globose and never didymous, although in
their plate (Baumart. v. Java, fig. 554) a slightly didymous
fruit is shown. I find that the normal 2-seeded ripe fruits
of J. grandifolia are 10-12 mm. wide and, when dried, are
distinctly didymous as is general in Ixora, but that
immature fruits 4-8 mm. wide, even when dried, do not
show a longitudinal sulcation between the two seeds. _Hence
I conclude that Koorders and Valeton drew their opinion
from immature fruits. But these authors also distinguish
I. odorata by its short corolla-tube and hairy style. But as
I have said, these characters vary much in different
collections of J. grandifolia which. in other respects are

Gardens Bulletin, S.S.

201

identical. It seems a little fastidious, moreover, to distin-
guish two species of tree solely on the hairiness of the.
filiform style. J. odorata (Bl.) K. et V. is my I. grandifolia
form 1: and I. grandifolia sensu K. et V. consists of my
form 5 and my form 2 (as shown by the figure, Baumart.
v. Java. fig. 545, in which the leaves are drawn with few
lateral veins).

Hooker’s varieties, as amplified by King for the
Malayan flora, I interpret in this way. The var. gigantea
is based on the size, shape and texture of the leaf and so
it covers my form 2 (Ridley 4120), form 3 (Curtis 3384),
form 4 (Wray 2973, 3971) and form 5 (King’s coll. 5609,
10294, Wray 3678). The var. coriacea, based on the more
or less lanceolate, very coriaceous leaf, covers my form 5
(Curtis 2486, Wray 2140) and, perhaps, var. lancifolia
(Ridley 2213, which King doubtfully referred to var.
coriacea). The var. arborescens covers my form 1 (Derry
310) and form 2 (Derry 571, King’s Coll. 4686, Ridley
4165). The var. rosella may be my form 3, on account of
its pointed leaves, but Kurz describes the corolla-tube “about
an inch long’, though Hooker gives 4-3”. It seems to
me that the varieties gigantea, coriacea and arborescens are
unsatisfactory because from their definitions so many
variable features have been omitted though it may be
possible to define accurately var. arborescens and var.
coriacea from the single, and therefore the type, specimens
cited for each by Hooker.

As regards I. macrosiphon Kurz (= I. grandifoiia var.
Kurzeana Hook. fil), it differs from I. grandifolia only in
the longer corolla-tube (2-5-3 cm. long) and my form 6
connects the two, having the tube 1-7-2 cm. long. The three
collections of I. macrosiphon in the Singapore Herbarium
leave no doubt that it is conspecific. These collections,
which are cited by King and Gamble as var. Kuryzeuna
are:—King’s Collector 277, s.n. 3.3.94, Prain’s Collector
sn. 1899 (all from the Andamans). The variety has not,
been found in Malaya, though it probably occurs.

Ridley’s species are hard to understand through the
_ different treatment which he has given them in his Flora
and his previous paper. Thus J. coriacea of his paper
(1918), based on Wallich 6151 which is the type of J.
grandifolia var. coriacea Hook., is assigned to both /.
crassifolia and I. elliptica in his Flora, while I. gigantea of
his paper (starting in the middle of line 8 from the top
of p. 84) which is there said not to be true J. grandifolia of
Zollinger, is omitted from the Flora, though under J.
grandifolia of the Flora it is remarked that var. gigantea
King and Gamble “‘is a typical form” of the species.

Vol. XI. (1941).

202

I. coriacea Ridley, based on Wallich 6151, is evidently
my form 5, according to Ridley’s description, and this agrees
with King’s interpretation of var. coriacea Hook. (also based
on Wallich 6151). In his paper, Ridley doubtfully refers
Wallich 6153 to J. coriacea, but in his Flora he refers both
Wallichian collections to I. elliptica. As the specific names
“coriacea” and “elliptica’” are manuscript names of R.
Brown taken up by Ridley, it follows that J. elliptica ex
Ridley is an exact synonym of I. coriacea ex Ridley. I may
add that both Hooker and King regarded Wallich 6153 as
typical I. grandifolia. |

I. crassifolia is based firstly, in the paper (1918), on a
specimen of Griffith’s from Ayer Panas in Malacca and to
it, in the Flora, Ridley assigns many other specimens as
well as var. coriacea Hook. It is possible that this collection
of Griffith’s should be taken as the basis of var. arborescens
Hook., but Griffith’s specimens are not available in Malaya.
In any case, Griffith’s specimen is the type of I. crassifolia
and, from Ridley’s original description, I find it hard to
distinguish from I. coriacea: it is said to have a blade
8-9 X 8-814” with a blunt base. J. crassifolia of Ridley’s
Flora is said to have a blade 13 X 3”, lanceolate and
narrowed to both ends, as my var. lancifolia, and the cited
collections, which J can check for certain in Singapore,
namely G. Janeng (Kelsall), Tahan River and Pekan, are
my var. lancifolia to which Griffith’s specimen from Ayer
Panas cannot belong if Ridley has correctly described it.
Hence, I have no hesitation in making the new variety
lancifolia, for it is not covered by the original definition
of any of Ridley’s species or of Hooker’s and King’s
varieties.

I. Ridleyi Merrill was proposed for IJ. crassifolia Ridley,
which is a later homonym of I. crassifolia Merrill. Merrill
merely cited Ridley’s references so that he threw but
another cloud over the problem.

I. fluminalis, as originally described, covers my form
1 (Derry, s. Rumbai), form 2 (Ridley 4165; Kunstler,
Goping) and form 5 (Haniff 4049). It represents the
common forms of J. grandifolia with leathery blunt leaves.
A distinction of seeming importance between J. fluminalis
and I. grandifolia which must occur to the reader studying
Ridley’s description is that J. grandifolia is said to have
white flowers and I. fluminalis yellow. The corollas of both,
however, have the white limb and pink tube typical of J.
grandifolia. The reason for ascribing yellow flowers to /.
fluminalis seems to have been the field-note of Kunstler’s
collection 4686, the “yellow flowers” of which refer, how-
ever, to the tiny calyces from which the corollas have

Gardens Bulletin, SS.

203

dropped. The distinction in flower-colour, of white or pink,
given by Ridley for his species depends on whether the
flowers of the specimen were open in which case they would
have looked white, the reflexed limb concealing the corolla-
tube, or not open, the pink tube then being more striking.
It seems, in any case, that J. fluminalis must be discarded.
Ridley based his species on “J. grandifolia var. arborescens
King not J. arborescens Hassk.” By this, Ridley meant var.
arborescens Hook. fil in the definition of which Hasskarl’s
I. arborescens was directly cited. In other words, taxo-
nomically, J. fluminalis Ridley is JI. grandifolia var.
arborescens Hook. fil. which is J. arborescens Hassk. with
Hasskarl’s plant as the type of each.

A specimen of Haviland’s collection 2157, which is the
type of I. lancifolia Ridley, in the Singapore Herbarium
shows this species to be identical with the small-flowered,
few-veined J. Blumei which is my I. grandifolia form 1.

I. timoriensis is very close to I. grandifolia but it differs
in the laxer inflorescence with fewer flowers, all of which
are distinctly pedicellate, and in the hairy corolla-mouth.
I should be inclined to consider it as a variety, in which
case it would be the legitimate name for the species.

KEY TO THE FORMS AND VARIETY OF I. GRANDIFOLIA

Leaves lanceolate, very gradually tapered to the base,
the reticulations invisible in the dried leaf, and
the undersurface finely wrinkled a

. Leaves more or less elliptic: reticulations visible:

undersurface even
Leaves with 6-10 pairs of side-veins: blade up to

var. lancifolia

9 cm. wide
Corolla-tube -6—-9 cm. long as died Form 1
Corolla-tube 1-1-2 cm. long
Leaf-blade blunt .. ne ir Form 2
Leaf-blade acute .. ” Me Form 3
Leaves with 10-16 pairs of side-veins: blade
7-18 cm. wide
Corolla-tube -7-1-0 cm. long cx he Form 4
Corolla-tube 1-1-2 cm. long Mf: a Form 5
Corolla-tube 1-7-2-0 cm. long .. fe Form 6
Corolla-tube 2-5-3 em. long de .. (var. Kurzeana)
Form 1

Leaves with 6-10 pairs of lateral veins, acute, subacute or
blunt, 3-5-9 em wide.
Corolla-tube short 6-9 mm. long.

Malaya
Pahang; Cons. For. 3603.
Negri Sembilan; Alvins s.n. (S. Ujong).
Malacca; Alvins 2254, 2332, Derry 310 (det. I. grandifolia
var. arborescens by King). ©

Bornean Collections
Carr 26520, For. Dept. B.N.B. 4721, Wood 2033.
Haviland 1673 (det. J. fluminalis), Omar 395 (det. I. fuminalis).
Haviland 2157 (det. I. lancifolia).

Mot XT. (1947).

204

Javan Collections
v. Steenis 2415 (det. J. odorata and I. Blume?).

Two forms might be made of this one, as in Forms 2
and 3, according as the leaves are blunt or acute. But my
object is merely to explain the variety.

Form 2
Exactly as Form 32 but the leaves blunt.

Malayan Collections
Perak; King’s Collector 4686 (det. J/. grandifolia var.
arborescens by King).
Pahang; For. Dept. 15659, Burn Murdoch 200 (no flowers: det.
I. concinna).
Malacca; Derry 571.
Johore; Corner 25892, 28628, s.n. 3.4.34 (S. Kambau pr. S.
Sedili), Ridley 4165 (det. J. grandifolia var. arborescens
by King), 13490, 15387.
Singapore; Ridley 4120 (det. J. grandifolia var. gigantea by
King).
Sumatra Collections
Boden-Kloss 14567 (Siberut: det. J. coriacea).

Javan Collections
Koorders 343868, Teysmann 18651 (det. I. macrophylloides and
I. fluminalis).
(Koorders et Valeton, Baumart. v. Java fig. 545).

Form 3
Leaves 3-5-9 cm. wide, the apex acute and generally
accuminate with a tip 6-12 mm, long: lateral veins 6-10 pairs,
the reticulations often very faint.
ms Corolla-tube 1-1-2 cm. long, the petals 3-45 « 1-5-2 mm.
unt.
} Malayan Collections
Kedah; For. Dept. 17890.
Penang; Curtis 3384 (det. as J. grandifolia var. gigantea by
King).
Malacca; Alvins 797.
Johore; Corner 32052, 32277, 32498, s.n. 11.4.36 (S. Kayu, pr.

S. Sedili).
Sumatran Collections
Boden-Kloss 14669 (Sipora Il.), Krukoff 4088, 4207, 4328.

Bornean Collections
J. & M. S. Clemens 26255, 31294.

Form 4
As Form 5 but the corolla-tube 7-10 mm. long.

Malayan Collections
Perak; Wray 2978, 3971 (both det. I. grandifolia var. gigantea
by King).
Sumatran Collections
Bartlett 2887.
Javan Collections
sn. cult. Hort. Bog. XIII (det. J. odorata).

Form 5
Leaf-blade 7-18 cm. wide, blunt, with 10-16 pairs of side-
veins: petiole 1-5-3-5 cm. long.
Corolla-tube 1-1-2 cm. long.

Gardens Bulletin, S.S.

Form

gap

205

Siamese Collections
Haniff and Nur 4049 (det. J. fluminalis Ridley by Craib).

Malayan Collections

Kedah; Haniff 15480 (det. I. fuminalis Ridley by Craib), 21020.

Penang; Curtis 2486 (det. J. grandifolia var. coriacea by King).

Perak; King’s Collector 5609, 10294 (both det. J. grandifolia
var. gigantea by King), Wray 2140 (det. I. grandifolia
var. coriacea), 3678 (det. I. grandifolia var. gigantea by
King).

Pahang; For. Dept. 17133.

Negri Sembilan; Alvins 1177.

Johore; Corner 28578, s.n. 11.10.36 (S. Kayu, pr. S. Sedili),
Henderson 28419, Kiah s.n. 6.10.36 (S. Kayu, pr. S. Sedili),
Lake and Kelsall s.n. 22.10.92 (G. Janing).

Sumatran Collections
Krukoff 4397.

Bornean Collections
For. Dept. B.N.B. 4302 (Brunei), Wood 2121.

G
As form 5 but the corymb smaller and denser, 10-13 cm.
wide, with the peduncles of the main branches 2-5-4 cm. long:
corolla tube longer, 1-7-2 cm., the petals -5 cm. long, blunt to
subacute.
: Malayan Collections
Johore; Kiah s.n. 6.10.86 (S. Kayu, pr. S. Sedili).

The length of the corolla-tube in this form bridges the
between J. grandifolia and var. Kurzeana (=

macrosiphon Kurz) in which it is 2-5-3 em. long.

var.

Vol.

lancifolia Corner
Malayan Collections

Kelantan; Nur and Foxworthy 11953.

Trengganu; (Kemaman), Corner 30719, s.n. 13.11.35 (2 sheets).

Pahang; Burkill sn. June 1915 (Pulau Tioman), Ridley 2213,
(det. J. grandifolia var. coriacea by King), s.n. 1891
(Pengkalan Kasai), s.n. 1893 (Taka Tahan).

Negri Sembilan; Burkill 4478.

Malacca; Alvins s.n. (5 sheets), 2081, 2153, 2224, Derry 67.

Johore; Corner §.n. 9.3.36, s.n. 10.4.36 (Ulu Segun, Gunong
Panti), Holttum 9282, Kiah s.n. 12.10.36, s.n. 6.3.37, Lake
and Kelsall s.n. 21.10.92 (G. Janing), Nur 19988, Ridley
sn. Jan. 1910 (Panti River, Kota Tinggi).

Sumatran Collections
Bartlett 7479 (pr. Loemban River, Asahan).
Forbes 1986 (det. J. grandifolia var. coriacea by King).

COLLECTIONS INTERMEDIATE BETWEEN J. GRANDIFOLIA

ND VAR. LANCIFOLIA

Kelantan; Haniff and Nur 10090.

Perak; Derry s.n. 1899 (Maxwell Hill, det. Tarenna grandi-
folia), King’s Collector 5935 (det. J. gvrandifolia var.
coriacea by King).

earn oe 22416 (det. J. crassifolia), For. Dept.

Selangor; For. Dept. 16904 (det. I. crassifolia).

XT. (1941).

206

I. javanica (Bl.) DC Javanese Ixora
emend. Koorders et Valeton, Bijdr. Booms. Java 8, 1902,
p. 160.
Basinym:—Pavetta javanica Bl., Bijdr. 1826 p. 949.
Synonyms :— al

Ixora stricta auctt. non. Roxb.

I. fulgens auctt. non. Roxb.

I. amoena Don, Gen. Syst. 3, 1834, p. 571.

I. stricta var. amoena Ridley, Fl. Mal. Pen. 2, 1923, p. 94.

I. aurorea Ridley, Journ. Bot. 72, 1934, p. 251.

Icones :—

Bot. Mag. t. 4586.

Baumart. Java 4, t. 553.

An evergreen shrub or small tree to 8 m. high, the trunk
1-8 em. thick, with smooth brown bark: glabrous.

Leaf-blade 5-25-5 & -8-11-5 em., elliptic, varying narrowly
elliptic to lanceolate in riverside forms, simply acute or at
most subacuminate, the base varying from gradually tapered
and narrowly cuneate to rather suddenly rounded, widest about
the middle, the proximal half often wider than the distal,
membranous to subcoriaceous or distinctly coriaceous in some
cultivated forms, rather dull green, often light or yellowish
green: lateral veins 6-14 pairs, generally 6-10 or 9-13, inclined
obliquely forward and without distinct marginal loops or in
var. retinervia distinctly inarching: petiole -25-1-5 cm. long:
stipular point -1—75 em. long, longest in the narrow-leafed
riverside forms.

Corymbs 3-5-20 em. wide, with peduncles 0-9 cm. long,
the branches articulate and coral-red: sepals blunt to subacute,
variable in the same flower: corolla-tube 2—5 em. long, the
limb 1-3-8 em. wide, the petals varying blunt to acute.

Berries 1-2-2 cm. wide, depressed globose, ripening red
then black, 1-2 seeded.

Distribution:—India, China, Siam, Malay Archipelago.

v . .
var. multinervia var. nov.
costis lateralibus utrinsecus 9-14, marginem versus non
arcuato-connexis: lamina usque 25-5 & 11-5 em., non vel vix

acuminata.
Typus:—Corner 33438 (Kota Bahru, Kelantan, 22.4.37).

Y var. paucinervia var. nov.
costis lateralibus utrinsecus 6-9 (-10), marginem versus
non arcuato-connexis: lamina. usque 155 « 7-8 em., non

acuminata.
Typus:—Burkill and Haniff 17342 (Telok Sisek, Pahang,
| 4.12.24).
Vv var. retinervia var. nov. Ixora Lotong

costis lateralibus utrinsecus 7-12, marginem versus con-
spicue arcuato-connexis, costa media angulo lato excurrentibus:
lamina 6-15-55 < 1-7-5-8 em., elliptica vel oblongo-elliptica, -
plus minus accuminata.

Typus:—Corner 29020 (E. Johore, 9.2.35).

KEY TO THE VARIETIES

Lateral veins of leaf making a wide angle with
the midrib, conspicuously inarching with wide
loops near the margin: leaf-blade generally more
or less acuminate, up to 15 & 5-5 em. Eye var. retinervia

Gardens Bulletin, S.S.

“a

/

207

Lateral veins of leaf inclined obliquely forward,
not distinctly inarching: leaf-blade not acumi-

nate
eae veins 6-9 (-10) pairs; blade up to :
5 xX 5-5 em. : var. paucinervia
falters veins 9-14 pairs: blade up to 25 x
i &- Se “ var. multinervia

There is much sala oek over the nomenclature of the
Malayan plants which I refer to I. javanica. They have
been called J. stricta Roxb. until recently when Ridley has
made them a new species, 7. aurorea, and Burkill has
referred them to J. chinensis Lam. (Dict. Econ. Prod. Mal.
Pen. 1935). But, as I have explained under J. chinensis,
I. stricta Roxb. is to be interpreted as a synonym of 7.
chinensis, which differs from the Malay specimens in its
very short petiole and very blunt petals so that, provided
this distinction is held, the Malayan specimens are not /.
chinensis. I find that they agree exactly in all essentials
with I. javanica as interpreted by Koorders and Valeton
for the Javanese flora. Craib identified them with I.
amoena but, as I have shown under that species, I can find
no difference between it and J. javanica. Ridley’s latest
interpretation I consider no solution because it deals with
very few specimens and, while resting on such extremely
variable features as the size of leaf, stipule, corymb and
corolla and the shape of sepals and size of corolla-tube, it
ignores differences in the venation, the shape and size of
the petals, and the colour of the flowers, and omits the
common Red Ixora of Malayan gardens and villages.

I. javanica is the commonest and mgst variable I[xora in
Malaya. I have distinguished 3 varieties and 30 forms so
that the variation may be perceived. The one which I have
found to be most constant is var. retinervia, which has not
been recognised previously, yet there are intermediates
between them all. Until the species has been studied
genetically, I do not see how it can be dissected profitably
in the herbarium.

I. pseudojavanica Bremekamp, I[. ovalifolia Bremekamp
and J. lacuum Bremekamp (Journ. Bot. 75, 1937, pp.
170-171) are to me only forms of J. javanica. Such
“splitting’”’ is quite impracticable.

var. retinervia var. nov. Ixora Lotong

The Malayan specimens are remarkably constant in
having rather small, subacuminate glossy leaves, drying
dark-brown, fuscous or even black, inarching veins often
making almost a right angle with the midrib, small corymbs
4—9 cm. wide and small corollas (tube 3-4 em. long; limb
1-1-6 cm. wide; petals -25—3 cm. wide, oblong or subelliptic,
more or less acute, light orange-yellow turning scarlet or

Vol. XI. (1941).

208

deep red.). One might, therefore, regard the variety as
a species but there are several collections intermediate with
var. paucinervia and as these lack the full characteristics of
var. retinervia I have classed them with var. paucinervia.
The variety retinervia also resembles J. Lobbi var.
stenophylla in the Herbarium: the distinctions are given
under J. Lobbit. Compare also J. concinna.

It appears that there are forms of this variety in
Borneo which have not only larger flowers but leaves which
dry pale. The systematic position of these Bornean
specimens, which I have listed below, is problematical
because they resemble also J. Robinsonii. Many more
collections will be needed to establish their status from their
variability.

Malayan Collections
Penang; Curtis 1112 (det. I. stricta by King), s.n. B. Laksa-
mana March 1890, (det. J. stricta var. amoena by King):
For. Guard. '1112.
Perak; Wray 3024 (det. J. fulgens by King).
Pahang; Burkill 15704, 17229; Henderson 22009; Ridley 2226,
2227, 16243 (det. I. concinna); Strugnell 23453.
Trengganu; Moysey 336438.
Selangor; Burkill 6842.
Negri Sembilan; Holttum 9609; Kloss sn. B. Tangga 1910;
Nur 11669; Ridley sn. G. Angsi, Feb. 1900.

Malacca; Alvins 62, 567, s.n. 29.1.86; Derry 257; Goodenough
‘758; - tBAG.

Johore; Best 7717; Corner 29020, 29939; Goodenough s.n. 18938;
Ridley 10940 (? 16940), s.n. 1892, s.n. 1898, s.n. 1906
(? 1908).

Singapore; J.S.G. 3999, 5004; Ridley s.n. 1904.

Bornean Collections

? Winter 2154 (S.E. Borneo, 1908), det. J. amoena: (the
inarching of the veins not always distinct): (det. J. Winkleri
Brem. Bull. J. Bot. Buit., XIV, 1937, p. 266).

2? Kijoss 19136, For. Dept.“ B.N-B. 4256, 4276 (all det.
I. stenophylla) : corolla- tube 3-3-8 cm. long, limb 2-5-3 cm. wide,
petals -6—75 cm. wide.

? For. Dept. B.N.B. 4037: corolla-tube 3-7—-4:3 cm. long,
limb 2-5-3-3 cm. wide, petals -5—-6 em. wide.

KEY TO THE FORMS OF I. JAVANICA VAR. MULTINERVIA

Corolla-tube 2-3 cm. long
Limb 1-1-5 em. wide

Petals -5 em. wide, blunt, yellow then pinkish orange Form 1
Petals -3 cm. w ide, pointed, rich yellow then intense

brick-red d s * Form 2

Corolla-tube 3-3-8 cm. ‘long

Limb 2-5-3-8 em. wide oe My .. I. congesta

Form 10

Limb 2-3-2-5 em. wide a A ue Form 4

Limb 2-2-3 cm. wide: petals acute os re Form 9

Limb 1-5-2 em. wide
Petals blunt,
Petals -5—-6 cm. wide ie ee ch Form 5
Petals -8—-5 em. wide Mey: tat _ Form 6

Gardens Bulletin, S.S.

209

Petals acute, -5 cm. wide

Petals turning deep brick-red »% Form 7 -
Petals pale orange buff then pale red as Form 8
Limb 1-1-5 cm. wide: petals subacute -
Petals -3-—-4 cm. wide 7H a Form 10
Petals -2 cm. wide, narrowly oblong .. oe Form 11
Corolla-tube 3-8-4:3 cm. long
Limb 1-2-1-5 cm. wide
Petals -3 cm. wide, salmon pink a Re Form 12
Limb 1-7-2:3 cm. wide
Petals -5—-6 cm. wide, acute: (leaves narrow, 2-5-
6-5 em. wide) ~ a ap Form 138
Petals -8--4 cm. wide, blunt —
Petals yellow then brick-red Form 14
Petals pale reddish oe (leaves 2-5-3-8 cm.
wide) Form 15
Limb 2-3-2.5 em. wide
Petals -6—-7 cm. wide, acute, pale yellow then salmon
pink my Va Sx - Form 16

KEY TO THE FORMS OF I. JAVANICA VAR. PAUCINERVIA

Flowers yellow, not reddening: “Yellow Javanica”’
Corolla-limb 1-5-1-9 cm. wide .. ne a Form
Corolla-limb 2-2.5 cm. wide .. x =, Form
' Flowers yellow turning red or pink
Flower pale yellow turning ae clear pink: “Pink
Javanica” - Form
Flower turning orange red to brick- ‘red; “Red
Javanica”
Corolla-tube 2-2-8 cm. long
Limb 1-1-5 cm. wide

Petals blunt: leaves thin ne ie Form
Petals subacute: leaves coriaceous e. Form
Limb 1-8-2-3 cm. wide: petals blunt a Form

Corolla-tube 2-5-3-8 cm. long
Limb 1-1-3 cm. wide: petals -2 cm. wide,

subacute .. ak my} Form
Limb 1-3-1:8 cm. wide
Petals -3—-5 cm. wide, blunt or subacute .. Form

Limb 1-8-2-:2 em. wide
Petals pointed

Petals -3—-5 cm. wide, elliptic m: Form
Petals -2—-3 cm. wide, oblong ‘< Form
Petals blunt, -4—-6 cm. wide .. Form
Limb 2-2-5 cm. wide: petals -4--6 cm. wide,
acute Form
Corolla-tube 3-8-5 cm. long: petals 3—5 em. wide,
acute
Limb 1-7-2 cm. wide me 2 Form
Limb 2-2-5 em. wide 33 RY Form

Form 1

17
18

£9

29
30

Corolla-tube 2-3 cm. long, the limb 1-1-5 em. wide: petals

‘5 em. wide, blunt, yellow then pinkish orange.
Malayan Collections

Kelantan; Corner 33443.

Penang; Ridley s.n. 1900 (P. Tikus).

Trengganu; Rostado s.n. Feb. 1904.

Singapore; (cult. in Hort. Bot.) Corner 30795 (det. I. con-

gesta), Nur s.n. 14.6.29 (det. I. congesta).

Vol. XI. (1941).
Cc

Form

Form

Form

Form

that

210

Siam
Keith 129 (det. I. a@moena by Craib).
The two Singapore collections have coarsely veined leaves
like J. congesta but only 10-13 pairs of side-veins like J.
javanica. They might be classed with either species.

2

~

As Form 1 but the petals narrower, .35 cm. wide, pointed
and turning deep red.

Malayan Collections
Perlis; Henderson 22950.
Penang; Curtis s.n. (cult. Bot. Gdns.).
Kelantan; Corner 33434.
(? Selangor; For. Dept. 17493).

Siam

Kiah 24360 (det. J. amoena by Craib).
Java

Hallier 142.

A

Corolla-tube 3-3-8 cm. long, the limb 2-2-2-5 em. wide:
petals -6 cm. wide, acute or subacute.

Malayan Collections
Singapore; (cult. in Hort. Bot.) Corner.

5
Corolla-tube 3-3-8 cm. long, limb 1-5-2 cm. wide: petals
-5—-6 em. wide, turning reddish pink.

' Malayan Collections
Kelantan; Corner 33437.
6 .
As Form 5 but the petals narrower, -3—-5 cm. wide.
Malayan Collections

Kedah; Ridley 8354.
(? Singapore, Ridley s.n. 9.11.1892, Kranji).

7 Giant or Common Red Ixora
Corolla-tube 3-3-8 em. wide, limb 1-5-2 cm. wide: petals

-) em. wide, pointed, turning intensely brick-red.
Leaves rather large, the blade 8-28 x 5-12-5 cm.

Malayan Collections

Kedah; Ridley s.n. June 1898.

Kelantan; Corner 33436, s.n. 21.4.37 (Kota Bahru), s.n. 22.4.37
(Kota Bahru).

Perak; Ridley 9710, Seimund s.n. 6.12.25 (P. Lalang), s.n.
8.3.26 (P. Rumbia) s.n. 23.83.26 (P. Rumbia).

Pahang; Burkill s.n. June 1915 (P. Tioman).

Malacca; Seimund s.n. 8.4.21 (P. Jarak).

Singapore; (cult. in Hort. Bot.) Cantley s.n., Nur s.n. 15.8.18,
14.6.29. Corner.

Anamba Islands
Henderson 20225.

This is the commonest garden Ixora in Malaya, yet one
offers the greatest difficulty in classification. It has

generally larger leaves with more numerous veins than /.
javanica, besides pointed petals, and it flowers continuously
all the year, whereas the wild forms of J. javanica like those

Gardens Bulletin, SS.

211

of I. congesta and I. Lobbu are seasonal. In the Singapore
Botanic Gardens it has for many years been called J.
macrothyrsa but it does not answer in any detail with the
description of that species in Bot. Mag. t. 6853. It seems
that dried specimens have never come into the hands of
systematists working on the flora of Malaya or of neigh-
bouring countries, unless as Wallich 6152 as I have
suggested under J. fulgens. It corresponds so remarkably
well with Roxburgh’s interpretations of J. fulgens that it
seems to be the species which he intended though whether
it is the true I. fulgens, taxonomically, has still to be
discovered.

It is known, colloquially, as the Common Red Ixora or
the Giant Ixora. The only other garden-variety with which
it may be mistaken is J. congesta form 9, which differs in
its larger, redder flowers that develop seasonally, its much
less bushy habit and its more numerous side-veins in the
leaf.

Form 8
. As Form 7, but the petals pale orange-buff turning pale
red.
Malayan Collections
Kelantan; Corner 33435.

Trengganu; Holttum s.n. 10.5.25 (K. since aetna
Singapore; (cult. in Hort. Bot.).

Form 9
As Form 7 but the corolla-limb 2-2-3 em. wide.

. Stam
Haniff & Nur 3910 (det. J. amoena by Craib).

Malaya
Pahang; Corner 29910.
Form 10
As Form 7 but the corolla-limb 1-1-5 cm. wide.

Malayan Collections
Kedah; Curtis 2607, Ridley 8354: (both det. I. amoena by
Craib).
Penang; Ridley 7092.
Perak; Kunstler ‘ieee
Form 11- :
As Form 10, bie the petals only -2 em. wide, narrowly
oblong.
Malayan Collections
Province Wellesley; Ridley 10776.
Form 12

Corolla-tube 3-8—4-3 em. long, limb 1-2-1-5 em. wide; petals
-3 cm. wide, subacute, salmon-pink, never red.

Malayan Collections
Kelantan; Corner s.n. 22.4.37 (Kota Bahru).
Form wc,

Corolla-tube 3-8—4-3 cm. long, limb 1-7-2-3 em. wide; petals
-D—-6 em. wide, pointed.

Vol. Xf. (1941):

212

Malayan Collections
Trengganu; Moysey 33891.

Form 14
As Form 13 but the petals narrower, -3—4 cm. wide, blunt,
turning deep brick red.
Malayan Collections
Kelantan; Corner s.n. April 1937 (2 collections).
Form 15
As Form 14 but the flowers pale reddish pink.
Malayan Collections
Kelantan; Corner s.n. 22.4.37 (Kota Bahru).

Form 16
Corolla-tube 3-8—4-3 cm. long, limb 2-3-2-5 cm. wide: petals
-6—-7 cm. wide, pointed, pale yellow then salmon pink.
Malayan Collections
Kelantan; Corner 33438, Nur & Foxworthy 12255.

Form 17 Yellow Javanica

Corolla-tube 2-5-3 cm. long, limb 1-5-1-9 em. wide: petals

blunt, persistently light orange buff or apricot-buff never
reddening.

Malayan Collections
Singapore; (cultivated) Corner 34513, Hort. Bot. Sing. s.n.
10.8.21.

This and the following form are cultivated in Malayan
gardens as a “Yellow Ixora” but they may be more
accurately known as the “Yellow Javanica”, to distinguish
them from the “Yellow Coccinea”. The colour is not a
true yellow but a pale dull orange slightly pink.

Form 18 Yellow Javanica
As Form 17 but the corolla-limb wider, 2-2-5 cm. wide.
Malayan Collections
Singapore; (cult.).
Form 19 Pink Javanica
Corolla-tube 3-2-3-8 cm. long, limb 2-2-6 em. wide: petals
-4—-6 cm. wide, pointed, elliptic, pale yellow buff eo wholly
pale rose pink, never red.

Malayan Collections
Negri Sembilan; Corner 31502.
This collection is from a rather narrow-leafed shrub,
10 ft. high, growing in thickets by a sluggish, forest stream
beside the main road from Tampin to Gemas. Plants with
rather broader leaves are also cultivated in Malayan gardens
as a “Pink Ixora’”: they may be known more accurately as
the “Pink Javanica’’.
Form 20 !
Corolla-tube 2-2-8 cm. long, limb 1-1-5 cm. wide: petals
-3—4 cm. wide, blunt. Leaves thin.
Malayan Collections
Trengganu; Rostado s.n. Feb. 1904 (K. Kemaman).

Form 21
As Form 20 but with subacute petals and very coriaceous
eaves,

Gardens Bulletin, S.S.

Form

Form

213

Malayan Collections
Singapore; (cult. in Hort. Bot.) Nur s.n. 22.6.29.

22
As Form 20 but the corolla-limb wider, 1-8—2-3 cm

Sumatra
Ridley 8979 (Penasa) (det. J. Lobbiit ex. enum. Herb. Kew).

23
Corolla-tube 2-7-3-3 cm. long, limb 1-1-3 cm. wide: petals

-2 em. wide, subacute.

_ Malayan Collections
Perlis; Ridley 14997.
Penang; (cult. in Hort. Bot.) Curtis s.n. 29.11.02.

Siam

- Setul; Ridley 14998 (det. J. amoena by Craib).

Form

24
Corolla-tube 2-8-3-5 cm. long, limb 1-3-1-8 em. wide: petals
-3—-5 cm. wide, blunt to subacute.

Malayan Collections

Kelantan; Henderson 19572, 22595 (both det. J. stenophylla).

Trengganu; Corner 25924 (det. I. stenophylia).

Penang; (cult. in Hort. Bot.) Curtis s.n. Aug. 1900.

Perak; Corner s.n. 29.7.36 (Pangkor), Ridley 7187, 10276.

Pahang; Burkill and Haniff 17342, 17602, Burn Murdoch 158,
319, Cons. For. 4767, Evans 13231 (det. I. concinna ex
enum. Herb. Kew), Henderson s.n. 19.8.29 (Kuantan),
Ridley s.n. 19.8.89 (Pekan), s.n. 1891, s.n. 25.8.91 (Pasir
Mandi, det. J. concinna), 1085 (det. J. concinna), Seimund
962, Walker 23326.

Selangor; Hume 7107, Ridley s.n. 1.1.21 (Klang Gates), 13408.

Malacca; Burkill 1327, Derry 84, Seimund 1139.

Johore; Corner 25960, 25961, Henderson 18218.

Singapore; (cult. in Hort. Bot.) s.n. 31.10.21, Nur s.n. 12.6.24,
s.n. 9.9.24.

Anamba and Natuna Islands
Henderson 20450, v. Steenis 966.
Javea
v. Steenis 5416.
Sumatra
v. Steenis 3847.

This is one of the common wild forms of 7. javanica in

Malaya. It is abundant by the sea-shore and, as a narrow-
leafed bush of the flood-zone, by riversides.

Form 25

Corolla-tube 2-8-3-8 cm. long, limb 1-8-2 cm. wide: petals
-3—-5 cm. wide, elliptic, pointed.

Malayan Collections é‘
icici Rostado 11971, s.n. 1904 (Bundi).
Pahang; Ridley 1075.
Singapore; (cult. in Hort. Bot.) Burkill s.n. 23.10.17, Corner
s.n. 26.10.37.

Form 26

Vol.

As Form 25 but the petals narrower and oblong, 2-3 cm.

' wide: corolla-tube 2-5-3 cm. long.

AT. (1941).

214

Malayan Collections
Singapore; (cult.) Corner s.n. 26.10.37.

Form 27
Corolla-tube 3-2-3-8 cm. long, limb 1-7-2 cm. wide: petals
4—6§ em. wide, blunt.

Malayan Collections
Kedah; Ridley 15890 (Pulau Rawei).
Johore; Holttum 24958 (Pulau Pelandok).

Form 28
Corolla-tube 3-3-8 cm. long, limb 2-2-5 em. wide: petals
-4—5 cm. wide, acute.

Malayan Collections
Pahang; Ridley 2114 (det. I. concinna).
Johore; Ridley 15403 (rather narrow-leafed, riverside shrub).

Form 29
Corolla-tube 3-8-5 cm. long, limb 1-7-2 em. wide: petals
-3—-D cm. wide, acute.

Malayan Collections
Kelantan; Nur and Foxworthy 12087.
Pahang; Burkill and Haniff 16107, Cons. For. 5176.

Java
Koorders 39079.

Sumatra
Robinson and Kloss s.n. 19.3.14 (Korinchi Exped.: det.
I. amoena).
All these collections are from riverside shrubs with small,
narrow leaves, up to 12 * 2-5 em.
Form 30 /
As Form 29 but the corolla-limb wider, 2-2-5 cm.
Malayan Collections
Perak; Ridley 14312.
Pahang; Kiah and Strugnell 23981 (det. I. stenophylla).

Both collections are from narrow-leafed, riverside shrubs
(blade 6-20 x 1-5-3 em.).

Without open Flowers
Kedah; Dolman 21705.
Perak; Burkill 12435, Wray 3448.
Pahang; Burkill and Haniff 16936 (riverside bush with
lanceolate ieaves: det. J. stenophylla).

COLLECTIONS INTERMEDIATE BETWEEN I. JAVANICA
AND I. LOBBII

Borneo
Elmer 20678, 21049, Wood 2262.

These three collections have been distributed as /.
Lobbiu but their leaves, which are narrowly oblong-elliptic
(10-25 3-5-8-5 cm.), are simply pointed or subacute (not
acuminate) and have only 10-16 pairs of side-veins. There-
fore they are not J. Lobbu. I think they are extreme forms.
of J. javanica but the Ixoras of Borneo are too little known
to be rightly understood.

Gardens Bulletin, S.S.

215,

COLLECTIONS INTERMEDIATE BETWEEN I. CONGESTA
AND I. JAVANICA ;

Perak; Murton 43 (det. I. congesta: but more like I. javanica ©
form 7 with leathery leaves).

Selangor; Hume 8433 (det. I. congesta: but more like I. java-
nica form 7 with leathery leaves and remarkably strong
veins; “flowers pink” on the field note does not apply to
I. congesta).

Malacca; Alvins 781 (det. J. congesta: but with 11-17 pairs
of side-veins; exactly between IJ. javanica multinervia and
I, congesta).

Singapore; Baker s.n. 17.7.17 (Gardens Jungle), Goodenough
5.2.89 (Selitar), Ridley 5067, 5674, sn. 1892 (Chan Chu
Kang), s.n. 1892 (Changi): (all these collections, having
10-16 pairs of side-veins, seem exactly between I. congesta
and I. javanica).

I. Kingstonii Hook. fil.
| Fl. Br. Ind. III, 1880, p. 140.

A shrub or small tree up to 9 m. high, glabrous except
for the puberulous branches of the inflorescence, or even wholly
glabrous.

Leaf-blade 15-30 * 5-9 em., rather narrowly oblong elliptic
or narrowly obovate, more or less acuminate, the base cuneate,
thinly coriaceous; side-veins 13-24 pairs, not distinctly
inarching: petiole -6-1-5 cm. long.

Corymb 5-12 cm. wide, dense, sessile or with a peduncle
up to 3 ecm. long, the branches articulate, 2-3 mm. wide, rather
short, finely and often sparsely puberulous, or even glabrous:
bracts, at the base of the main branches, 6-12 * 4-7 mm.,
broadly lanceolate, subacute, rather leafy: bracteoles 3-6
2-5-3-5 mm., oblong, acute or subacute like the sepals: flowers
yellow-red, ? inodorous: sepals 2-5-5 & 1-5-2-5 mm., longer
than the ovary, glabrous, oblong, blunt or subacute, unequal in
length in the same flower: corolla-tube 3-3-5 em. long, glabrous,
the petals 6-7 x 3-5-5 mm. elliptic, blunt or subacute.

Distribution :—Malaya.

Type:—Griffith 2996 (Malacca).

Collections:—Perak; King’s Collector 5972, Burn Murdoch
202, Wray 726, s.n. 1889 (Changkat Serdang).

Selangor; Curtis 2346, Ridley 5674, 7300, 8536, sn. Aug. 1904
(Sempang Mines): For. Dept. 122, 4941, 8274, 24258.
Malacca; Goodenough 1271, Hervey s.n. 20.9.90, Ridley s.n.

1890 (S. Baru).

Johore; Curtis s.n. April 1901 (Muar).

Very little is known of this species which has seldom
been collected within the last thirty years. I interpret the
species from the specimens identified by King in the
Singapore Herbarium. According to King, and to the field-
note on Burn Murdoch’s collection, the flowers are yellowish
red, perhaps yellow turning red and this character
distinguishes it radically from the similar species I.
multibracteata, I. merguensis etc. with which it may easily
be mistaken in the herbarium. It is most like J. wmbeilata
but differs in the narrower, more coriaceous leaves, the
broader bracts, stouter branches of the corymb as well as
in the colour of the flower.

Mol. AT. .(19%1).

216

I. Lobbii King et Gamble | Glossy Ixora
Mat. Fl. Mal. Pen. 4, 1904, p. 152. '
Synonyms :—

I. Lobbu var. angustifolia K. & G., loc. cit. p. 153 pro parte.
I. fulgens Roxb. sensu Hook. fil., F.B.I. 3, p. 146 pro parte
(fide King and Gamble).

I. aurorea var. major Ridley, Journ. Bot. 72, 1924, p. 252.

An evergreen shrub or treelet, 1-10 m. high like J. javanica.

Leaf-blade 10-28 X 48-5 cm., lanceolate-elliptic to
lanceolate-obovate, more or less suddenly acuminate, widest at
or above the middle, the base gradually tapered, dark glossy
green, with 12-26 pairs of lateral veins inclined forward and
distinctly inarching near the margin.

Sepals subacute to acute.

Petals yellow or pinkish yellow turning red, generally
acute.

Distribution:—Lower Siam, Malaya, Borneo, Sumatra,
Anamba and Natuna Islands.

’ var. stenophylia var. nov. Red River Ixora

Synonyms:— |

I. stenophylla (Korth.) Kuntze sensu Ridley, Trans. Linn.

Soc...d, 2894; p. 311,- FL Mal. Pen: 2.1922, 9. 82

I. salicina Ridley, Journ. Bot. 72, 1934, p. 253.

Lamina lanceolata, -7—-2-3 cm. lata, utrinque attenuata,
apice longe acuminata, costis lateralibus utrinsecus 8-15; frutex
riparius.

A riverside shrub -5-3 m. high.

Leaf-blade 3-5-15 -7-2:3 em., lanceolate, gradually
tapered to the long acuminate apex and to the base, broadest
at the middle, with 8-15 pairs of lateral veins distinctly
inarching at the margin and making a wide angle with the
midrib: petiole -2-1-3 cm. long: stipular point up to -9 cm. long.

Distribution:—Malaya, Borneo, Palawan (?).

Type:—Ridley 2215 (Tahan River, Pahang).

This species has hitherto been attributed to Loudon for
authorship. On turning to Loudon’s reference, (Encycl.
Suppl.II, p. 1548) it will be seen, however, that there is no
description. Loudon based his name on Pavetta Lobbii
Teysm. et Binn., which is also a nomen nudum as well as
a nomen invalidum from its manner of publication. Hence
the authorship must be attributed to King and Gamble who
nrst described it. These authors give a reference to
“Pavetta Lobbii Teysm. et Binn. in Mig. Ann. Mus. Lugd.
Bat. IV, 194’, but I cannot find the name on this page, nor
in the whole volume, nor in any other work of Miquel’s.
Misgivings arise, therefore, whether the name J. Lobdbii
has not been antedated because the species is one of the
commonest in western Malaysia. This is, unfortunately,
impossible to decide from the earlier descriptions of red
Jxoras and so one must wait until herbaria have been
ransacked for the types of the old species. Surely it did
not escape Miquel.

Gardens Bulletin, S.S.

217

- The narrowly obovate, acuminate, many-veined, dark
glossy green leaves and the acute petals will generally .
distinguish J. Lobbw. But there are narrow-leafed collec-
tions of J. congesta which seem to pass into J. Lobbii and
likewise many-veined collections of J. javanica, particularly
of J. javanica retinervia which has the acuminate blade,
pointed petals, and inarching veins of J. Lobbit.

My basis for the identification of the Malayan material
with I. Lobbi is the specimens so named by King in the
Singapore Herbarium. King doubtless saw the specimen
of Teysmann and Binnendijk in the Kew Herbarium. It
seems that the species is not recognised outside Malaya,
though Bremekamp reports it from the neighbouring islands
of the Rhio Archipelago.

The variety angustifolia King and Gamble, as pointed
out by Ridley, is a mixture of a narrow, lanceolate-leafed
state of J. Lobbi and a small lanceolate-leafed riverside
shrub. Of the collections cited by King and Gamble there
belong to the first state Wray 519, Scortechini 1893 and
King’s collector 2718, and to the second only Ridley 2215.
Therefore Ridley referred the first state to var. angustifolia,
defined it anew and made Wray 519 as the lectotype: the
second state he referred to the Bornean Pavetta stenophylla
Korth. as I. stenophylla Ridley which name he changed
subsequently to I. salicinuw as he considered the Bornean
plant to be different. As regards var. angustifolia sensu
Ridley, I am unable to distinguish it because there is every
gradation to the typical, slightly broader-leafed state of /.
Lobbit, and the variety has no particular habitat or distribu-
tion, specimens from all parts of Malaya being referable
to it. Hence I merge this variety with J. Lobbii. On the
other hand, J. stenophylla sensu Ridley (= I. salicina) is a
well-marked riverside shrub with narrow, shorter leaves
and fewer veins than typical J. Lobbu and in Malaya, it
appears to be limited to the rocky streams that are the
tributaries of the rivers flowing into the China Sea: in
which relation, may be, it occurs also in Sarawak. and
probably, Palawan. Nevertheless, I am unable to follow
Ridley in maintaining his J. stenophylla (I. salicina) as a
species. The low shrubs by the banks of swift flowing
rivers that are prone to frequent and sudden flooding, like
the Tahan and its tributaries, have small narrow leaves
very different at first glance from those of J. Lobbii in the
quiet, deep shade of the forest, but the smaller and more
precipitous the river the less are its riparian growths
subject to flooding and the larger are the leaves and the
bushes so that it is not easy to decide whether they should
be classed as “‘stenophylla” or ‘“Lobbii’’, e.g. the Johore

Vot.. XP (1941).

218

collection of var. stenophylla (shrubs 2-4 m. high). I
prefer, therefore, to reduce I. stenophylla sensu Ridley as a
variety of J. Lobbii without any decision about its identity
with Pavetta stenophylla Korth. The narrow-leafed,
riparian form of Melastoma malabathricum affords a
comparable case.

The variety stenophylla has been confused in Malaya
with the narrow leafed riparian states of I. javanica but
these differ in having fewer lateral veins, which do not
inarch distinctly but are directed obliquely forward, and,
generally, in having less acuminate leaves and smaller
flowers with blunt or subacute petals. The type of I.
aurorea var. major Ridley, namely Nur 20007, is a riverside-
form intermediate between J. Lobbii and var. stenophylla,
perhaps a riverside-form of my form 7.

I. salicifolia (Bl.) (= I. fulgens Roxb. var. salicifolia
K. & V.) differs from I. Lobbii var. stenophylla in having
much longer, linear-oblong leaf-blades (12-38 X 12-25
cm.) with many more pairs of lateral veins (15-30 prs.).
Nothing approaching I. salicifolia has been found in
Malaya: and J. Lobbii seems not to occur in Java.

I. javanica var. retinervia may well be mistaken for
I. Lobbii var. stenophylla in the herbarium. The var.
retinervia is a‘small tree, or shrub, of lowland forest, not
a riverside bush: its leaves are broader than those of var.
stenophylla, have on the whole fewer lateral veins and are
oblong elliptic to elliptic. It seems, moreover, that the
Malayan specimens of var. retinervia are always distin-
guishable by their smaller flowers. Nevertheless, I feel
there is some close relation between the two.

KEY TO THE FORMS OF I. LOBBII

Corolla-tube 2-2-7 cm. long

Limb 1-1-3 cm. wide: petals -25 cm. wide, acute ve Form 1
Corolla-tube 2-5-3-8 cm. long

Limb 1-2-1-8 cm. wide: petals -2—--4 cm. wide

Petals acute 2 ay a: Sid Form 2
Petals blunt ; bi: Form 3
Limb 2-2-5 em. wide: petals pay cm. wide, acute... Form 4
Limb 2-7—3-3 em. wide: petals -5—8 cm. wide, acute .. Form 5

Corolla-tube 3-8—5 cm. long
Petals -4—-5 em. wide, acute

Limb 1-2-1-7 cm. wide: tube 3-7-5 em. long if Form 9

Limb 1-8-2-3 em. wide: tube 3-7-4-3 em. long a2 Form 6
Petals -5—8 cm. wide: tube 3-7-5 em. long

Limb 1-8-2-5 em. wide: petals -5-—6 cm. wide “*s Form 7

Limb 2-5-3-3 em. wide: petals -7—--8 cm. wide ai Form 8

KEY TO THE FORMS OF I. LOBBII VAR.. STENOPHYLLA

Corolla-tube 3-3-7 cm. long: rine -A—-6 cm. wide
Limb 1-7-2 cm. wide : : .. Form 10
Limb 2-2-3 cm. wide s ae .. Form 11

Gardens Bulletin, S.S.

219

Corolla-tube 3-7-5 cm. long

Limb 2-2-8 em. wide: petals -4--6 cm. wide ef Porm 12,
Limb 2-8-3-5 cm. wide: petals -5—-7 cm. wide sean TAOETN ches
Form 1

Corolla-tube 2-2-7 cm. long, limb 1-1-3 cm. wide; petals
-25 cm. wide, acute.

Form 2
Corolla-tube 3-2—3-8 cm. long, limb 1-25-1-8 cm. wide: petals
-25—-4 em. wide, oblong, acute.

. Malayan Collections

Kedah; Henderson 21264.

Penang; Burkill 813, 2599, Curtis s.n. March 1892 (Govt. Hill),
s.n. April 1884, 479.

Province Wellesley; Ridley 6991.

Perak; Cantley s.n., Curtis 3338, Fox 158, For. Dept. 11567,
Ridley s.n. 1892 (K. Kangsar), s.n. 1892 (Maxwell’s Hill),
Scortechini s.n.

Pahang; Henderson 10695, Kiah 23901.

Selangor; Burkill 7580, Cons. For. 2278, Hume 9254, Pestana
28430.

Malacca; Alwins s.n., Alwins 60, Burkill 462, 1321, Corner s.n.
23.2.36, Derry 10, Hullett s.n. April 1888 (Mt. Ophir).
Singapore; Cantley s.n., Corner s.n. 1.4.37, Nur s.n. 23.2.29,

Ridley 5673.

Henderson 20124.
This is the commonest form of J. Lobbii, but whether it is
taxonomically the typical state I do not know.

Form 3

Anamba Islands

as Form 2 but the petals blunt to subacute.

Siam
Curtis 3258 (Kasoom).
Borneo
Moulton 6703/15 (Upper Baram, Sarawak: det. I. cf. amoena).
Form 4
Corolla-tube 3-3-8 cm. long, the limb 2-2-5 cm. wide: petals
-4—.6 em. wide, lanceolate, acute.

Malayan Collections
Pahang; Henderson 21655 (Pulau Tioman).

Anamba Islands
Henderson 20185 (det. I. javanica).

Form 5
Corolla-tube 3-3-8 cm. long, the limb 2-7-3-3 em. wide:
petals -5-—-8 cm. wide, acute.

. Sumatra
Ridley 8970 (Batang Mandan, det. J. fulgens).
Except for its short corolla-tube, this form agrees with
Form 8. Ridley’s collection has, however, much smaller leaves
than is general in the collections of Form 8.

Form 6
Corolla-tube 3-8-4-3 cm. long, the limb 1-8-2-3 em. wide;
petals -4—-5 em. wide, elliptic, acute.

Malayan Collections
a ptaes pao and Holttum 8607, Henderson 10584, Nur

Vol. AT... (2947).

220

Sumatra
Posthumus 724 (Djambi; det. J. javanica).

Form 7 Ixora Sedili
Corolla-tube 3-8-5 cm. long, limb 1-8—-2-5 cm. wide; petals
-5—6 cm. wide, elliptic, acute: (corymb branches pale green in

32293).
Malayan Collections

Johore; Corner 28166, 30787, 30962, 32293, 32495, Fielding s.n.
Oct. 1892 (Pulau Tinggi).

This fine Ixora is known only from the watershed of
the Sedili River in the East of Johore and from the island,
P. Tinggi, off the East Coast of Johore. One may expect it
therefore in Borneo, where its nearest ally, Form 8, occurs.
It commonly develops into a spindly tree, 15-25 ft. high,
with chocolate brown, cracked bark; the leaves have rather
large, thinly leathery, dark glossy green blades which in
shape and venation are inclined to resemble those of /.
congesta, though they are always acuminate and generally
_more or less obovate and narrower. The more numerous
veins and pointed petals distinguish it from J. congesta
Form 10: it seems also to have a different distribution but
our knowledge of the flora on the East side of Malaya is
yet very meagre.

Form 8 Giant Glossy Ixora

As Form 7 but the corolla-limb 2-5—3-3 em. wide, the petals
-65—-8 ecm. wide: (corymb branches pale green in 32292).
Malaya
Johore; Corner 32292.
Borneo
Allen 629 (Payao R., B.N.B., det. J. congesta).
Moulton 78 (Baram R., Sarawak, det. J. fulgens).

This is the finest red-flowered Ixora that I have seen.
The big shining flowers contrast splendidly with the dark
glossy leaves.

Form 9
Corolla-tube 3-5-5 em. long, limb 1-2-1-7 cm. wide, petals
-4—-5 cm. wide.
Malayan Collections
Perak; Anderson 152, Burkill and Haniff 12713, Haniff 21043,
_ Ridley 152, Wray s.n. 1889 (Taiping).

This variety from the Perak Hills has lanceolate leaves
12-24 & 2-5 cm. It much resembles var. stenophylla but
the longer leaf-blade and a habit that is not riparian
distinguish it.

STERILE, FRUITING OR NOT FLOWERING COLLECTIONS

Malaya
Penang; Curtis s.n.
Province Wellesley; Burkill 6538, Ridley s.n. Dec. 1895.
Perak; Haniff 10555, Ridley s.n. March 1896 (Gunong Tung-
gul), Wray 519, 952, 2891, 3449.
Pahang; Burkill & Haniff 17003, For. Dept. 28273, Machado
11569.

Gardens Bulletin, S.S.

221

Selangor; Ridley s.n. June 1889 (K. Lumpur).

Negri Sembilan; Nur 797.

Malacca; Alwins s.n., Burkill 6419, Derry 966.

Johore; Corner s.n. 11.3.37, Ridley 11324, sm. March 1908
(S. Tebrau), s.n. Aug. 1908 (Sedenah).

Singapore; Cantley 83, Hullett s.n. 1893 (Changi Road).

Sumatra
Bartlett and La Rue 29 (det. I. salicifolia).
Bunnemeyer 3204 (det. J. salicifolia).

Anamba Islands
v. Steenis 897 (det. I. javanica).

_ Borneo
For. Dept. B.N.B. 4705.

FORMS OF VAR. STENOPHYLLA

Form 10
Corolla-tube 3-3-7 cm. long, limb 1-7-2 cm. wide: petals
-4—-6 cm. wide.
Malayan Collections
Johore; Holttum 10870.

Form 11
As Form 9 but the corolla-tube 2-2-3 cm. wide.

Malayan Collections
Trengganu; Corner 25825, s.n. 24.6.832 (Kemaman).

Borneo
Ridley 12443 (Puak) (det. I. stenophylia).

Philippine Islands
Merrill 1362 (Palawan: det. I. fulgens).

Form 12
' As Form 11 but corolla-tube 2-2-8 cm. wide, petals -4—6
cm. wide.

Malayan Collections
Kelantan; Haniff & Nur 10280.
Pahang; Burkill and Haniff 16562, Corner s.n. Sept. 1937
(S. Tahan) (2 collections), Haniff and Nur 8089, Ridley
2215, Seimund 496, 585, Strugnell 22468.

Form 13
As Form 11, but corolla-limb 2-8-3-5 em. wide, petals
-‘D—-7 em. wide.
Malayan Collections
Johore; Corner 32490.
Pahang; Corner s.n. 7.9.37 (S. Tahan).

STERILE, FRUITING OR WITHOUT FLOWERS

Malayan Collections
Perak; Ridley s.n. Sept. 1890 (Bujong Malacca).
Pahang; Seimund 586, Nur 18924 (det. IJ. stricta).
Johore; Ridley s.n. 1904 (Castlewood). -

Borneo
Anderson 218 (Gunong Rumput), Ridley 117438 (Bidi).

Vol. XI. (1941).

222

COLLECTIONS BETWEEN I. LOBBII AND
VAR. STENOPHYLLA

Johore; Nur 20007 (type of J. aurorea var. major Ridley):
Corner 31434.
These collections may well be narrow leafed riverside
plants of my form 7. Their leaves are not lanceolate enough
for true var. stenophylla.

J. micrantha Ridley Mountain Ixora

Journ. St,. Br. KR.’ Ag. Soe. 61, 7212, 9. 2.
Synonym :—

I. patens Ridley Journ. St. Br. R. As. Soc. 79, 1918, p. 83.

A mountain shrub or small tree to 8 m. high: glabrous
except for the puberulous branches of the corymb.

Leaf-blade 9-18 * 3-8 em., oblong elliptic or slightly
obovate, more or less acuminate, the base cuneate or rather
abruptly rounded, scarcely coriaceous, with 9-16 pairs of
side-veins directed obliquely forward and inarching near the
edge: petiole -3-1-5 cm. long: stipular point 3-7 mm. long.

Corymb 6-20 cm. wide, 5-14 em. high, large, sessile or
very shortly stalked, the branches articulate and puberulous,
the flowers densely crowded and more or less sessile, very
fragrant: sepals as long as or shorter than the ovary, subacute:
corolla-tube 3-9 mm. long, pink or red, or yellow turning red;
petals 25-3 * 1-5 mm., oblong, blunt or subacute, reflexed
along the tube, yellow gradually turning red.

Berry 1 cm. wide, subglobose, scarcely sulcate.

Distribution:—Malaya; apparently only in the mountains,
1000-1700 m.
Typus:—Ridley 15830, Sempang Mines, Selangor, April

1911
KEY TO THE FORMS
Corolla-tube 3-5 mm. long ie me Form 1
Corolla-tube 7-9 mm. long eo Se Form 2
Collections ©

Form 1:—Haniff 2455 (Perak, G. Hijau), For. Dept. 27552

(Pahang), Anderson 31 pro parte (Perak, G. Hijau).
Form 2:—Anderson 31 pro parte, Burkill and Haniff 12629,

Henderson 11819 (all three from G. Hijau, Perak), Wray

3934 (Perak, G. Bubu), Hose 78 (Selangor, Fraser’s Hill),

Ridley 15880 (Selangor), Robinson s.n. 19.1.93 (Selangor,

G. Menkuang: type of I. patens).

Without full-grown flowers:—Ridley 9708 (Perak, G.
Keledang), 138640 (Pahang, Telom Valley) For. Dept. 28117
(Pahang, S. Lemai), 23394 (Selangor, Fraser’s Hill).

This species is closely allied with J. concinna, having
the same small fragrant flower with yellow corolla turning
red, and similar leaves, but it is easily distinguished by its
much larger inflorescences with stouter, puberulous
branches. Most of the Malayan specimens have been
identified with J. concinna. Ridley remarked that the
Species was related to I. grandifolia but the bracteate,
articulate ramuli of the corymb forbid such an alliance.

Gardens Bulletin, S.S.

223

I can find no specific difference whatever between the
types of J. micrantha and I. patens. Ridley’s statement that -
I. micrantha has only 9 pairs of side-veins to the leaf is
wrong: the type-specimen has as many as 13. The
differences in size of leaf and corymb, given by Ridley, are
trivial and such as occur in any plant.

The collection Wray 3934 is the basis of the record of
I. undulata in Malaya, being determined as such by King
and Ridley. Wray’s fieldnote is ‘flower yellow, tube tinged
pink, as the fiower gets older it turns to red”, but I. undulata
has. white flowers, and this is to me conclusive that Wray’s
plant is not J. undulata: indeed, Wray’s plant is identical
with the type of J. micrantha.

I. nigricans W. et A. White Needles

Prodr. FI. Ind. -I,. 1834, p. 428.
Wight le. Pl, t. 318.
Synonyms :—

I. affinis Don, Gen. Syst. 3, 1824, p. 571.

I. erubescens Don, loc. cit.

I. arguta King, Mat. Fl. Mal. Pen. 4, 1904, p. 148.

I. plumea Ridley, Journ. Str. Br. R. As. Soc. 59, 1911,
ob, 227.

I. nigricans var. erubescens Kurz, For. Fl. Burma 2,
1877, p. 24.

I. nigricans var. arguta Hook. f., F.B.I. 3, 1880, p. 149.

I. affinis var. arguta Craib, Fl. Siam.-En. 2, 1934, p. 147.

I. affinis var. plumea Craib, loc. cit. p. 148.

Pavetta acutiflora Korth. Ned. Kruidk. 2. 1851, p. 262
(fide auctt.).

Pavetta subulata Teysm. and Binn. Ned. Kruidk. Arch. 3,
1855, p. 408 (fide auctt.).

A bush or treelet to 4 m. high, flowering at 1 m., glabrous
except for the puberulous branches of the corymb.

Leaf-blade 8-20 *« 2-5-8 cm., narrowly elliptic to elliptic
obovate, acuminate, gradually tapered to the base, the upper
leaves on flowering branches often shorter, broader, and ovate
or even cordate, membranous, blackening in drying: side-veins
7-12 pairs, not or vaguely inarching: petiole -5-1-5 em. long:
stipule-point -3—-8 cm. long.

Corymb 5-15 cm. wide, rather lax, erect, sessile or with a
peduncle up to 4 cm. long, the branches slender, articulate,
more or less puberulous, often reddish: bracts 4-8 « -1 mm.
linear-lanceolate: bracteoles like the sepals but slightly broader:
flowers white, fragrant often only faintly: sepals -7-1-5 mm.
long, linear-lanceolate, glabrous, as long as, longer than or
shorter than the ovary (ca. 1 mm. long), green: corolla-tube
-8-1-3 cm. long, glabrous, the mouth also glabrous, white: petals
4-65 < 1-1-5 mm. linear-oblong, acute or becoming blunt on
expansion, recurved against the tube; style projecting 3-6 mm.:
ovary green or purplish.

Distribution :—India, Indo-China, Siam, Malaya, Sumatra,
Anamba and Natuna Islands, Java, Bali.

Type:—Wight 1335.

Collections :—
Siam; Haniff and Nur 3590, 3851, Keith 277, Ridley 15098.
Perlis; Ridley 14995, 15097, Henderson 23007.

Vol. XI. (1941).

224

Kedah; Best 21213, Curtis 3206, Holttum 15083, Ridley
8299.

Kelantan; Corner s.n. 1.5.87 (Kota Bahru).

Trengganu; Corner s.n. 1.5.37 (S. Besut).

Penang; Curtis 986, For. Guard..s.n. Ayer Itam, s.n.
Balek Pulau.

Province Wellesley; Ridley 7013.

Perak; Curtis 1303, Haniff 14184, Henderson 10074, 10295,
King’s Collector 870, 10901, Ridley 7188, 11905, s.n.
1891, s.n. 1898, Wray 2135.

Pahang; Burkill s.n. June 1915 (P. Tioman), Evans 13229,
Henderson 10665, 21678, Nur 18552.

Selangor; Ridley 8240, s.n. 1897.

Negri. Sembilan; Burkill 4962.

Malacca; Alvins 1084.

Johore; Fielding s.n. 1892 (P. Dayong), Henderson 18212.

var. ovalis Pitard
Fl]. Gen. Indoch. 8, 1924, p. 322.
Synonyms :—
I. lucida Hook. f., F.B.I. 3, 1880, p. 148.
I.. lucida var. densipila Craib, Fl. Siam Enum. 2, 1934,

AGL

(earblnde 5-14 x 2-6 cm., broadly elliptic to ovate-
elliptic, generally broadly rounded at the base, the apex dis-
tinctly acuminate, drying shiny, the side-veins arising at a
wider angle from the midrib and distinctly inarching near
the margin: petiole rather short, 2-7 mm. long.

Corolla-tube 2-3 cm. long, puberulous, the mouth glabrous
or more or less bearded: petals 5-7 & 2 mm.: ovary and
sepals puberulous.

Distribution :—Indo-China, Siam, Malaya (north).

: Coin pean: specimen from Tan-Huyen (prov. Bien-
oa).
Collections :—

Siam; Kerr 7108.

Penang; Curtis 2451, Nur 1478.

Kelantan: Henderson s.n. 22.10.28 (Gua Ninek), s.n. 15.

7.35 (Gua Teja).
Perak; Henderson 23856, Ridley sn. Aug. 1909 (K.
Kenering).

King’s reason for rejecting the name J. nigricans for

the Malayan specimens in favour of IJ. arguta rested on a
misunderstanding. King regarded the Wallichian collection
6154 as the type of I. nigricans and, having discovered that
the number included a variety of plants, he rejected I.
nigricans as a confused species. But the publication of J.
nigricans dates from Wight.and Arnott’s Prodromus where,
in the paragraph following the description, it was clearly
stated that the description was based on the collection 1335
of Wight’s with the remark that this collection was not
included under the Wallichian aggregate 6154. Wight’s
illustration is taken from his own collection and it
undoubtedly depicts the Malayan~« species. Wight and
Arnott described the corolla-tube as three-quarters of an
inch long but the illustration shows the corolla-tube 10-12

Gardens Bulletin, S.S.

225

mm. long as in most Malayan specimens and as described
by Koorders and Valeton, King and Pitard. Hence I have
no doubt of the identify of our Malayan plants.

Don’s species, /. affinis and I. erubescens, have generally
been regarded as identical with J. nigricans. They were
published in the same year without any indication of
priority over J. nigricans and, as this name is the better
known, Don’s have been reduced to synonyms (Bremekamp,
Journ. Bot. 75, 1935 p. 296). Craib considered that Don’s
species were different from IJ. nigricans but gave no reason.

I. plumea Ridley is identical with I. nigricans. Craib
endeavoured to distinguish it as a variety of I. affinis
on the character of the sepals being longer than the ovary.
But I find that there is every gradation to states with the
sepals as long as the ovary and those with the sepals shorter:
and the differences are trivial, concerning a length of
d—75 mm. Craib also distinguished J. arguta King as a
variety of I. affinis without saying how it differed. The
distinction between J. affinis and I. erubescens lies in the
more oblong leaf and shorter petiole of J. eruwbescens accord-
ing to Craib but he does not give the length of the petiole
in either or say from what part of the plant the leaf or the
petiole is to be taken for comparison. I consider that none
of these distinctions has even varietal value especially when
the general variableness of the plants in the length of the
corolla-tube and petals, the size of the leaf and corymb,
and the colour of the ovary has been disregarded.

On the other hand, var. ovalis Pitard does seem worth
varietal rank through the longer corolla-tube, the different
venation and the rounded leaf-base. Its leaves, in fact,
throughout the plant are like those just below the corymbs
of typical I. nigricans. Nevertheless there are some
collections intermediate between the two, and the variety
has all the important specific characters of J. nigricans
though Bremekamp excludes it. Craib’s J. lucida var.
densipila seems to be only an extreme form of var. ovalis
with densely puberulous corolla-tube and heavily bearded
corolla-mouth. The collection, Curtis 2451, has a rather
thinly bearded corolla-mouth whereas all the other Malayan
collections have a glabrous mouth.

Of the typical J. nigricans the flowering specimens in
the Singapore Herbarium can be separated into two forms,
those with the corolla-tube 8-9 mm. long and those with the
tube 10-13 mm. long. However, the length of the petals
does not always vary concurrently there being short-tubed
collection with long petals (55-65 mm.) and long-tubed
collections with short petals (4-5 mm. long) as well as
vice versa. |

Vol. XI. (1941).
i |

226

I. pendula Jack Pink Needles

Mal. Misc. 1, 1820, p. 11.

Synonyms :—
I. opaca Don, Gen. Syst. 3, 1824, p. 573.
I. montana Ridley, Journ. Str. Br. Roy. As. Soc. 79,

1918,.p. BZ

I. candida Ridley, Journ. F.M.S. Mus. 10, 1920, p. 141.
I. pendula var. opaca Ridley, Fl. Mal. Pen. II, 1922, p. 96.
I. Parkinsoniana Craib, Kew Bull. 1932, p. 428.

A shrub or small tree up to 8 m. high, glabrous except
for the puberulous cyme-branches and corolla-tube, occasionally
wholly glabrous or with the corolla-tube glabrous.

Leaf-blade 10-35 3-12-5 em., elliptic, oblong-elliptic,
lanceolate elliptic or more or less obovate, the apex subacute,
acute or distinctly acuminate, the base gradually tapered or
rather suddenly rounded, thinly leathery or distinctly so, dull
dark green: side-veins 12-24 pairs, directed forward and
distinctly inarching near the edge: petiole -6-1-8 cm. long,
rather stout: stipular point -2-1 cm. long.

Corymb 3-23 cm. wide, deflexed perpendicularly on a long
peduncle, 1-28 cm. long, the peduncle and branches generally
puberulous and more or less suffused purple, articulate, the
corymb terminal or from the twigs behind the leaves and
generally with a pair of leaf-like, often sessile and cordate,
bracts at the base; the flowers not scented: sepals varying
lanceolate, acute and as long as the ovary (1 mm. long) to
bluntly triangular and shorter than the ovary, the ovary and
sepals glabrous or more or less puberulous: corolla-tube
1-8-3-5 cm. long, deep rose-red, generally puberulous outside,
but often glabrous or nearly so, the limb -9-1-6 cm. wide:
petals 1-5-2-5 mm. wide, oblong, subacute, more or less reflexed
along the tube, with reflexed edges, white, or tipped pink, or
partly suffused pink.

Distribution:—Siam, Malaya, Sumatra.

Type:—Non extant, the type-locality being Penang.

Form 1
(= I. diversifolia Kurz sensu King and Ridley).

Corolla-tube 1-1-2 cm. long, the limb -8—--9 cm. wide.
Collections :—

Lower Siam; Haniff and Nur 2718, Kiah 24363.
Malaya; King’s Collector 5737, Ridley s.n. July 1909 (Ulu
Temango), Wray 3491.

I can distinguish only one species in Malaya which has
deflexed, scentless inflorescences, deep pink corolla-tubes and
white petals, and such must therefore be [. pendula Jack.
It is nevertheless exceedingly variable in detail, especially
in the leaf. The variable features are the leaf-size, shape,
venation and texture, the size of the corymb and the length
of the peduncle, the length of the corolla-tube and the
hairiness of the cyme-branches and corolla-tube. If one
attempted to distinguish varieties or forms reckoning with
all, these characters, the number would be impractical.

Some of the Penang specimens in the Singapore
Herbarium have large, rather blunt, not acuminate, leaves
with rather few veins. (blade 18-35 xX 6-12.5 cm., veins

Gardens Bulletin, S.S.

227

12-16 pairs): these are Burkill 6898, Curtis 2240, Curtis
s.n. July 1887, April 1890, February 1900. On the sheet.
of Curtis s.n. July 1887 King has written “exactly the same
as the type Wallich Cat. 6127”: this collection is not the
type but may be regarded as the neotopotype. Other
collections have linear leaves 10-20 « 2-4 cm. There is,
however, such diversity of the leaves in detail, with every
intermediate condition, that I do not see how varieties can
be maintained on such characters: I have noticed, too, that
the leaves on the short flowering twigs are often smaller,
more pointed and with rounded bases than those of the
stout vegetative shoots.

The size of the inflorescence and the length of the
peduncle vary extraordinarily. King distinguished JI.
opaca from I. pendula by its short peduncle, 1-5 cm. long,
and its erect corymb. The first point indicates merely the
lowest part of the range in size of the inflorescence and is
connected by all intermediates with the largest. The second
point is certainly an error by no means easy to detect from
herbarium specimens: the collections Ridley 3071, King’s
Collector 4061 and 4118, identified by King as J. opaca have
the typical deflexed inflorescences. Ridley’s definition of
I. pendula var. opaca is not applicable, and there is no
evidence that it is a mountain form more than a lowland.
Hence I reduce J. opaca to J. pendula, as being founded on
an inadequate range of material. The type of J. opaca is
Wallich 6141 from Penang.

As regards the size of the corolla some collections have
the tube 2:2 cm. long with the limb 1:2 cm. wide, others
have the tube 3-5 cm. long and the limb 1-6 cm. wide, but
most collections have some intermediate size. A few
collections have the tube only 1-1-2 cm. long and the limb
8—9 mm. wide and they have been identified with J. diversi-
foka Kurz by King and Ridley. Some of these have
unusually large leaves (up to 35 X 18 cm.) but I cannot
see any specific difference between them and J. pendula,
for they have the same characteristic inflorescence and
flowers and they display the same diversity of leaf.
Whether they are true J. diversifolia which comes from
Burma and has ‘white’ flowers I do not know. Hence I am
content to regard them as a form only, like the forms of
other Ixoras variable in size of corolla.

Of J. montana Ridley, there is a part of the fragmentary
type-collection (Haniff 16310) at Singapore, consisting of
‘two twigs, an inflorescence and two leaf-bases. It is
- identical with J. pendula except for the rather shorter
corolla-tube (1-8-2 cm. long) and thus it connects with my
form 1. The inflorescence is on a short peduncle like that
of J. opaca but it is a wide inflorescence typical of J. pendula.

Ol. XT. 6 P9AT).

228

Whether the inflorescence was decurved, or the twig bearing
the inflorescence was decurved, is now impossible to
discover but, as the specimen agrees in all other points with
other mountain specimens of J. pendula, I have not the
slightest hesitation in reducing it.

I. candida Ridley is merely a narrow-leafed form of
I. pendula var. opaca; that is, of the state of I. pendula
with a small, shortly pedunculate inflorescence. Specimens
of both type-collections are at Singapore: (it seems that the
number 6219, given by Ridley, for Robinson’s collection at
Dayang Bunting should be 6189 as numbered in the
Singapore Herbarium). There are numerous intermediates
connecting these lanceolate-leafed specimens (with blades
only 6-12 x 1-5-3 cm.) to the typical J. pendula. Both
type-collections show the deflexed inflorescence and both
have a puberulous ovary as in Craib’s I. Parkinsoniana.
The reason for considering the flowers as white is the
field-note “flowers white” to Robinson’s unnumbered syntype
from Telok Wau, Terutan: it is likely that the pink-corolla-
tube was. concealed by the expanded white petals and so
escaped the notice of the collector.

I. Parkinsoniana, according to the collections named
by Craib in the Singapore Herbarium, is merely a form of
f. pendula with puberulous ovary and sepals, short-corolla-
tube (1-5 em. long) and rather small, though long-pedun-
culate, corymb. These collections show ‘that the inflorescence
was deflexed as in I. pendula. Craib does not explain how I.
Parkinsoniana is to be distinguished from I. pendula. If
!. Parkinsoniana can be split off on the evidence of a
puberulous calyx and corolla-tube, what should one do with
the other forms of I. pendula with puberulous corymb and
corolla-tube, puberulous corymb and glabrous corolla-tube,
glabrous corymb and puberulous corolla-tube, and those
glabrous in all parts? Then add to these variations, all
the varieties of leaf-size, texture and venation, of width
of corymb.and of length of peduncle. To separate such as
“species” seems to me not only impracticable in the extreme
but also inability to distinguish “the wood for the trees’,
losing thereby the major aspects of natural features. The
specimens named J. Parkinsoniana by Craib in the
Singapore Herbarium are:—Holttum 15126, Haniff and
Nur 3557, Ridley s.n. Jan. 1897 (Langkawi), Haniff 15531
(all from the island of Lankawi).

I. pumila Ridley

F.M.S. Mus. Journ. 10, 1920, p. 142.

A dwarf shrub, 30 cm. high, with finely villous stem,
petioles and branches of the inflorescence, and thinly villous
undersides to the leaves.

Gardens Bulletin, S.S.

a -—_—

re

229

Leaf-blade 12-18 * 5-712 cm., obovate elliptic, simply
pointed, gradually attenuate to the base, thin: side-veins 10-14.
pairs directed obliquely forward, not inarching, the tertiary
veins rather close and transverse: petiole 3-8 mm. long: stipule
lanceolate with a long point, 4-6 mm. long.

Corymb 5 cm. wide, on a peduncle 4 cm. long, the branches
articulate, hairy, erect, the corolla white: bracts lanceolate, the
lowest pair 10 « 15 mm., the bracteoles like the sepals: sepals
-2—3 em. long, lanceolate or linear-lanceolate, acute, hairy,
longer than the ovary: corolla-tube 3 cm. long, glabrous: the
petals 5 K 2 mm., oblong, subacute: corolla-mouth glabrous.

Distribution :—Malaya.

IE ee tga 13226 (Gunong Senyum, Pahang, June

This species, known only from the type-collection,
should be easily recognised from the hairy stems and
undersides of the leaves, the hairy lanceolate sepals and
ae ha white corolla as well as from its apparently dwarf

abit.

Compare /. clerodendron with broad red sepals and
I. Brunonis with differently shaped leaves, hairy corolla-
tube, and longer sepals.

J. Robinsonii Ridley Tahan Ixora

Journ. F.M.S. Mus. VI, 1915, p. 154.

A shrub up to 2 m. tall: glabrous.

Leaf-blade 7-5-20-5 « 2-7-6 cm., rather narrowly elliptic
to broadly lanceolate or lanceolate-obovate, the apex tapered,
simply acute or acuminate, the base very gradually tapered
into the petiole, leathery, the edge incurved: lateral veins 9-18
pairs, passing obliquely out at a wide angle from the midrib,
inarching near the edge, the reticulations coarse: petiole -5-2
em. long, rather flattened and woody, passing so gradually
into the blade as to have its upper limit indiscernible: stipular
point 1-5 mm. long.

Corymb 7-5-15 em. wide, the peduncle 0-2-5 cm. long,
the branches red: sepals acute or subacute: corolla-tube 3-8-5
em. long, the limb 2-3—3-8 cm. wide, the 4 petals -5—-9 cm. wide,
elliptic, acute, deep salmon pink to scarlet.

Distribution:—Malaya (Gunong Tahan, G. Padang and ?
neighbouring mountains, 600-1800 m.).

Type:—Robinson 5304 (Gunong Tahan, 1300 m. alt.).

Key to the Forms
Corolla-tube 3-8-4:5 em. long: limb 2-3-2-8 cm. wide: petals
‘5 em. wide ne a2 see Form 1
Corolla-tube 4-3-5 cm. long: limb 3-3-8 cm. wide: petals -6—9

cm. wide Form 2
Collections
Form 1
Pahang; Robinson & Wray 5304 pro parte (G. Tahan).
Form 2

Kelantan; Henderson 19736 (ca. 650 m. Gua Ninek).

Pahang; Robinson & Wray 5304 pro parte (G. Tahan),
Haniff & Nur 8115 (G. Tahan, ca. 1800 m.), Ridley
16244 (G. Tahan 1300 m.).

Trengganu; Moysey 33360 (G. Padang, ca. 1300 m.).

Vol. XI. (1941).

230

The leathery leaves with extremely attenuate base,
stout petiole and coarse reticulation and the mountain habit
distinguish this species, the range of which is curiously
limited. I believe that one may say that it occurs only in the
mid-mountain forest on the quartzite ridges of the Tahan-
group of mountains from G. Padang on the Trengganu
boundary westward to the Telom River. Nevertheless, .
there are some Bornean collections from sea-level—one,
indeed, from the sea shore—which bear a_ strong
resemblance to this species though I prefer to classify them
with I. javanica, viz. Kloss 19186, For. Dept. B.N.B. 4037,
4256, 4276 (vide I. javanica var. retinervia).

On the type sheet of J. Robinsonii in the Singapore
Herbarium there are two specimens, one each of the two
form of flower-size. As Ridley described the corolla tube
as 2” long, one should regard my ‘Form 2’ as the type.
Ridley’s statement that the corolla has 5 lobes is erroneous.

Our collection are yet too few to say if the species is
as variable in size of flower as the other red I[xoras.

I. Scortechinii King et Gamble Dwarf Red Ixora

Mat. Fl. Mal. Pen. IV, 1904, p. 154.
Synonyms :—
I. humilis K. & G., loc. cit. non J. hwmilis O. Kze (1891).
I. Kingiana Craib, Fl. Siam. Enum. II, 1934, p. 159.
I. humilis var. Scortechinii Ridley, Fl. Mal. Pen. II, 1923,
p. 95.

A dwarf, evergreen, little-branched or unbranched shrub
up to % m. high, the stem sometimes decumbent with a few
upright branches: leaves tufted at the top of the stem with
hee internodes more or less concealed by the overlapping leaf-

ases.

Leaf-blade 13-36 5-18 ecm., elliptic to obovate, simply
pointed or subacute, generally rather blunt, narrowed to the
widely or narrowly cordate, base, generally subsessile, broadest
at or above the middle, dull green: lateral veins 11-21 pairs,
passing obliquely forward and distinctly inarching near the
margin: petiole 0-5 mm. long, generally concealed by the
cordate leaf-base: stipular point 1-5 mm. long.

Corymb 4-9 cm. wide, terminal, subsessile, very dense,
with red branches: sepals subacute: corolla-tube 2-4-5 cm. long:
corolla-limb 1-2:75 cm. wide, the petals -5—-85 cm. wide, broad,
rounded, blunt, pinkish yellow or orange pink generally
deepening to brick-red.

Distribution:—Malaya (Pahang, Perak, Treneeeny, Kelan-
tan, Penang, Province Wellesley).

Syntypes:—As given by King and Gamble.

var. caulescens var. nov.

A typo differt caule evoluto, usque 214 m. alto, internodis
distinctis, corymbis terminalibus vel lateralibus longe pedun-
culatis (pedunculis 4-10 cm. longis) haud congestis.

Distribution:—Malaya, Lower Siam, (Selangor to Patani).

Typus: Hume 7940 (Selangor, Semyih, 13.7.21).

Gardens Bulletin, S.S.

251

King and Gamble distinguished two species on the
basis of the number of primary lateral veins to the leaf,
namely I. humilis (15-18 pairs) and I. Scortechinu (8-12
pairs). I find that some collections have 11-16 prs. of
veins, and others have 14-21. I have also seen plants with
11-14 prs. of veins on small leaves and 16-21 pairs on
large leaves. But in no case have I seen as few as 8. King
and Gamble referred, moreover, Curtis 400 and Wray 4157
to both species. The distinction seems to me inconsistent
and therefore I have reduced both species, following Ridley,
but have been obliged to re-install the binomial I.
Scortechinii as the earliest legitimate name. Accordingly,
Craib’s later binomial J. Kingiana becomes a synonym.

Craib has given Penang as the type-locality of I. humilis
K. & G., but these authors indicated no type, having cited
six collections which must be taken as syntypes. Likewise
there is no holotype of I. Scortechinii but 5 syntypes. Of
the syntypes of J. humilis K. & G. I remove one (Ridley
7422a) to my variety caulescens.

This variety caulescens is abundantly distinct both in
the herbarium and the field. It has the habit of a slender
plant of J. congesta but the cordate, subsessile leaves and
broad blunt petals of J. Scortechinui. There can be no doubt
that I. Scortechinii is closely related to I. congesta, perhaps
even derived directly from it, and therefore var. caulescens
appears to retain its ancestral habit while typical J.
Scortechiniui has declined into a dwarf. The distribution
of I. Scortechinii is curiously limited to the northern half
of Malaya and that of var. caulescens scarcely exceeds it.

I. Scortechinii inhabits the dense, lowland forest. It
is in cultivation at Singapore, where it makes a good front
to a shady path or rockery, but it will not tolerate the open.
In the size of the flower, perhaps also in the colouring, it
is aS variable as the other red-flowered species J. congesta,
I. javanica and I. Lobbit.

KEY TO THE VARIETIES AND FORMS OF I. SCORTECHINII
Stem up to 4% m. high, leaves crowded at the nee

corymbs sessile : I. Scortechinii
Corolla-tube 2-2-5 cm. long
Limb 1-1-3 cm. wide: petals -5 cm. wide .. Form 1

Corolla-tube 2-3-8 em. long
Tube 2-3 cm. long: limb 1-3-1-6 cm. wide:

petals -5--65 em. wide , Form 2
Tube 3-3-8 cm. long: limb 1-8-2 cm. wide:
petals -6—-75 cm. wide i Ae Form 3
Corolla-tube 3-8-4-5 cm. long
Limb 1-8-2 cm. wide: petals -5—-65 em. wide Form 4
Stem 1-2% m. high, with distinct internodes:
corymbs with peduncles 4-10 cm. long .. var. caulescens

TO. AT (LOL ).

232

Corolla-tube 3-2-3-8 cm. long

Limb 1-5-1-8 cm. wide: petals -5—-6 cm. wide Form 5
Limb 2-2-3 em. wide: petals -6—9 cm. wide Form 6
Limb 2-3-2-8 cm. wide: petals -6—9 cm. wide Form 7
Corolla-tube 3-8—4-5 cm. wide
Limb 2-3-2-8 em. wide: petals -7-1 cm. wide Form 8
Limb 1-8-2 cm. wide: petals -5—-65 cm. wide Form 9
Collections
| I. Scortechintui (typical)

Form 1

Penang; Curtis s.n.: Perak, Wray 4157 (syntype).
Form 2

Penang; Curtis 400 (syntype): Province Wellesley, Burkill
6596.
Form 3

Perak; Ridley 7186 (syntype): Pahang; Burn Murdoch
215 (21.6.13): Kemaman; Corner s.n. (cult. Hort. Bot. Sing:
flowers becoming pinkish orange, never red).

Form 4
' Trengganu; Corner 25836 (the flowers of this form appear
to be persistently orange pink).

Specimens without flowers:—Province Wellesley; W. Fox’s
Collector s.n., Ridley 6992 (syntype): Pahang; Henderson
25081.

var. caulescens
Form 5
, Pahang; For. Dept. Mal. Pen. 20364: Selangor, Ridley
8234.
Form 6
Selangor; Ridley 7422a (syntype of J. humilis K. & G.).
Form 7
Selangor; Ridley 7422, Hume 7770A, 7940 (type).
Form 8 ;
Patani; Machado s.n. Sept. 1898.
Form 9
Kelantan; Henderson 19518.
Specimens without flowers:—Kelantan; Haniff and Nur
10110, Henderson 29660: Perak; Corner s.n. 22.8.37.

I. umbellata Koorders et Valeton Malayan White Ixora

Bijdr. Booms. Java 8, 1902, p. 162.

Baumart. Java, IV, t. 550.

le. Bog. II, t. 183.

Bot. Mag. t. 8577.

Synonyms :—
? Pavetta Korthalsiana Miq., Ann. Mus. Lugd. Bat. 4,
1868-9, p. 197. :
I. Korthalsiana (Miq.) Boerl., Hand. Fl. Ned. Ind. 2,
1891,).p. 130.
non I. Korthalsiana Kurz, Journ. As. Soc. Bengal 46 (ii),
Tee) 149. (= Pavetta calycina Korth. non I. calycina
w.).

A shrub or treelet up to 6 m. high, glabrous except for
the puberulous branches of the inflorescence, the hairy corolla-
mouth and the puberulous corolla-tube: twigs. pale.

Leaf-blade 13-33 « 5-13 cm., elliptic, acuminate or simply
pointed, occasionally subacute, narrowed gradually to the base,
drying light greenish: side-veins 12-20 pairs, not inarching
or only vaguely with reticulations, the secondary veins coarsely

Gardens Bulletin, S.S.

233

and conspicuously reticulate in the dried leaves: petiole
6-20 mm. long, rather stout: stipular point 1-4 mm. long.

Corymb 8-17 cm. wide, almost sessile or with a peduncle
up to 4-5 em. long, the branches articulate, puberulous and
slender (1-2 mm. wide): bracts 3-7 * 1-5-2-5 mm., linear
lanceolate, green, narrowly leafy with minutely ciliate edge:
bracteoles and sepals 3-6 1-5-3 mm. long, oblong, obtuse or
subacute, minutely ciliolate on the edge, green, like tiny leaves,
the sepals unequal: flowers white, fragrant: corolla-tube
2-5-4 em. long, white, puberulous externally, the mouth sparsely
to thickly bearded with white hairs: petals 6-10 «K 3-5 mm.,
white, oblong-elliptic, obtuse: anthers yellow: style green:
bracts, bracteoles and, generally, the sepals caducous in the
infructescence.

Distribution :—Malaya, Borneo, Java, ?Sumatra.

Type:—Hallier 719 (Java)-

Collections :—

Perlis; Henderson 23027 (det. I. crassifolia), Ridley 14999a
(det. I. multibracteata).

Kelantan; Henderson 19615 (det. I* Kingstonii).

Perak; Wray 3953 (det. J. multibracteata).

Pahang; Burkill 996 (det. J. multibracteata), Corner s.n.
11.6.32 (Pulau Tioman), Henderson 21677 (det. J. multt-
bracteata), Nur 18754 (det. J. multibracteata).

Selangor; Kloss s.n. Feb. 1912 (det. J. Kingstonti) Ridley
8570, sn. July 1897 (Gua Batu).

Johore; Corner 28489 (det. J. Kingstonii), 30685, Hender-
son 18219 (det. J. multibracteata), Ridley 12199 (det. J.
Kingstoni).

Singapore; Ridley 15463.

Borneo; Mondi 219 (Karimata Archip.) (det. J. Korthal-
siana Val.), Hallier 381 (det. J. Korthalsiana Val. and I.
umbellata).

In the Eastern half of Malaya, this species is rather
common by streams in the lowland forest. I have found
it abundant in the sandy coastal woods in the East of Johore.
The Malayan specimens have been identified mostly with
I. multibracteata which I consider a distinct variety with
larger bracts and sepals. Though I have not seen the type
of J. umbellata, the original description and the three
illustrations leave no doubt of the identity of the Malayan
specimens. The white fragrant flowers, the large green
leafy sepals and bracts, and the many-veined rather large
leaves together characterise the species. It has been
generally supposed related to I. con§esta but the resemblance
is superficial and suggests how little the genus can be
interpreted from the herbarium. :

Concerning P. Korthalsiana I am uncertain because
there is no figure or recent description of it, though both
Bornean specimens which I have cited were identified with
it by Valeton.

The corolla tube and petals are very variable in length
in different collections, so that it is impossible to define
the species narrowly on flower-size.

Vol. XI. (1941).

234

‘var. multibracteata var. nov. Lesser Malayan White Ixora
Synonym :—
| I. multibracteata King and Gamble, Mat. Fl. Mal. Pen. 4,
1904, p. 148.

Leaf-blade rather smaller and narrower (? always) 8-30
<< 3-10 cm., narrowed gradually to each end, the apex not
acuminate, often subcoriaceous, dark glossy green with wavy
edge: side-veins 10-15 pairs, generally fewer, the reticulations
often faint: petiole up to 1-5 em. long.

Peduncle of corymb up to 2-5 em. long: bracts, bracteoles
and sepals larger: bracts 8-12 * 5-10 mm., ovate, shortly
acuminate or acute: bracteoles and sepals 6-12 &K 4-7 mm.:
corolla-tube 2-5-3-5 em. long, glabrous or sparsely puberulous
outside, the mouth glabrous or hairy.

Distribution:—L. Siam, North Malaya.

Syntypes:—As enumerated by King and Gamble.
Collections :—

Siam; Curtis 2954 (syntype).

gare tS Curtis 3408 (syntype), Ridley 5540 (syntype),

Perlis; Henderson 23049, Ridley 14999.

Perak; Corner 31640, Wray 3317 (syntype).

The larger bracts and sepals distinguish this variety
from typical I. umbellata. The difference is at first so
striking that no doubt it led King and Gamble to overlook
the resemblance of their species to Valeton’s. Nevertheless
the difference is barely double the linear dimensions of the
parts, and little more than one finds in the variation of the
size of the corolla, or of the sepals, in other species such as
I. concinna and I. Lobbii, though in these, the sepals being
much smaller, the variation is almost microscopic. More-
over fruiting specimens, from which the bracts and sepals
have been shed, can be distinguished from those of typical
I. umbellata only by a close comparison of the venation
which reveals slight differences between the two which
cannot be rendered in words: and, even so, I doubt if anyone
could infallibly distinguish such.

Whether this variety is wide-spread in Malaya has still
to be discovered. I have seen it only in villages where it
is rather frequent from Ipoh northwards. Mr. J. N. Milsum
informs me, however, phat it occurs wild in the forest on
the road to Grik.

I. undulata Roxb.

Fl. Ind. 1, 2820, p. -395.

This species is recorded as occuring in Malaya both
by King and Ridley. The record is based on the collection,
Wray 3934, which is J. micrantha, as I have shown under
that species. There is no evidence that J. undulata is wild
in Malaya. .

Gardens Bulletin, S.S.

255

Ixora Sunkist nom. hort. nov.

A dwarf bush up to 1 m. high, glabrous, flowering —
throughout the year.

Leaf-blade 1-7-8 -7-2-5 em., small, subsessile, lanceolate
elliptic to lanceolate obovate, acute or subacute, the base
gradually or rather suddenly narrowed, subcoriaceous, rather
dull and yellowish green; side-veins 6-13 pairs, directed
obliquely forward and distinctly inarching near the edge:
petiole -1—3 cm. long: stipule-point -1—3 cm. long.

Corymb 5-10 cm. wide, as in J. chinensis, the branches red:
sepals as in I. chinensis, blunt to subacute: corolla-tube 2-2-3-3
em. long, the limb 1-3-2-5 em. wide, the petals -2—3 cm. wide,
lanceolate acute or subacute, pinkish apricot-yellow turning
brick red.

This description is based on specimens cultivated in
Malaya, probably introduced from southern China, but their
origin is unrecorded. They do not match any description
I have seen. Pending a decision on their botanical status,
I give to them the fancy garden name Sunkist. They grow
best in the full sun and the round inflorescences are not
unlike oranges in shape and colour. I find that there are
two forms :—

1. leaves with 6-9 pairs of side-veins: corolla-tube
2:2-2:5 cm. long, the limb 1:3-1-6 cm. wide.

2. leaves with 8-13 pairs of side-veins: corolla-tube
3—3°3 cm. long, the limb 2-2-5 cm. wide.

As I have remarked under I. chinensis, this Ixora is
what one imagines a hybrid of J. chinensis and I. javanica
or J. Lobbi would be like. It has the foliage of I. chinensis
and pointed petals of J. coccinea. :

SUM MARY

The following new varieties of Ixora are described :—
I. coccinea var. decolorans, var. auwreo-rosea, and var. rosea.
I. grandifolia var. lancifolia.
{. javanica var. multinervia, var. paucinervia, and var.
retinervia.
I. Lobbii var. stenophylla.
I. Scortechinii var. caulescens.
I. wmbellata var. multibracteatn.
“Ixora Sunkist” (a horticultural name).

Vol. XI. (1941).

236

PALMAE MALESICAE

IX.—TWO NEW CORYPHACEOUS GENERA
IN MALAYA

By C. X. FURTADO,
Botanic Gardens, Singapore

1. General Remarks

While arranging, in the Singapore Herbarium, the
specimens of the Malayan palms having fan-shaped leaves,
I detected many anomalous characters in the specimens
referred to Livistona rupicola Ridl. The petiole is not
prolonged as a thick rachis into the Jamina but ends in
numerous, almost equal, primary ribs or costae of the
lamina. Further the spadices or flowers are usually
unisexual, the segments of the perianth are imbricate, and
the three carpels of the ovary are free almost to the base
and are provided with free styles. .

These characters are sufficient to remove the specimens
from the section to which the genera Licuala and Livistona
belong and place them in a different section together with
Trachycarpus and Chamaerops. The genus Trachycarpus
is not recorded from any tropical regions, it being confined
to sub-temperate regions from N. W. Himalaya to Tibet,
Assam, Burma, Indo-China, China and Japan. It is closely
allied to Chamaerops, which is confined entirely to the
Mediterranean countries as far as Persia, and with which
Trachycarpus is sometimes united. However BECCARI (Ann.
Roy. Bot. Gard. Calc., XIII, 1933, pp. 7, 253 & 272) has
shown that Chamaerops has ruminate seeds, and Trachy-
carpus non-ruminate ones grooved on the raphal side with
a more or less deep intrusion of the integument. The
Singapore specimens referred to L. rupicola have no mature
fruits, but the seeds of the very young fruits present seem
to agree with the description of the fruits of Trachycarpus.
However the flowers exhibit details on which the specimens
can be arranged into two definite groups, neither of which
belong to Trachycarpus or Chamaerops.’

The specimens in the first group are all from the
limestone hills in the Langkawi Islands and have their
stamens united into a conspicuous epigynous cup as in
Licuala, whereas in Chamaerops the stamens are united into
a small hypogynous cup and in Trachycarpus they are
entirely free. These Langkawi specimens are made here
the basis of a new genus Liberbaileya named in honour of
the American palm specialist Prof. L. H. BAILEY. Its type
species has been named here as L. lankawiensis Furtado
to show the exact area where it is native. The imbricate
perianth and the staminal cup would make the genus a close
ally of the American palm genus Acanthorhiza.

Gardens Bulletin, S.S.

237

The specimens of the second group are from the
limestone hills in Selangor and are referred here to a new
genus Maxburretia, named in honour of the European palm
specialist Prof. M. BURRET. These specimens are generically
distinguished from Liberbaileya by their stamens being
epipetalous; and were it not for their free styles and
imbricate perianth, Maxburretia would have been a close
ally of Livistona and Rhapis. The type species of this new
genus is M. rupicola (Ridl.) Furtado (Livistona rupicola
Ridl. emend. Furtado). The generic description is derived
not from the lectotype of M. rupicola, but from its apotype,
Nur 34370. ae

liberbaileya lankawiensis is a caespitose palm pro-
ducing, under f2vourable'conditions, a stem 10 ft. high. Its
leaf-blade is small, about 1-114 ft. in diameter and bears, as
in some varieties of Chamaerops humilis, small punctiform
brown marks all over in the lower surface. The fiowers are
arranged in groups of 3—5 on a simple spadix which bears
two or three spathes in the peduncular parts and many,
scattered, simple or divided flower-bearing branchlets at
the apex. On the other hand, Mazxburretia rupicola is
apparently a stemless or scarcely caulescent, solitary palm;
it produces larger leaves with no brown marks in their
lower surface, compound spadices with 4-6 partial inflores-
cences, and one separate spathe at each internode, and
usually solitary flowers (rarely in groups).

Both these species have unarmed petioles, thus differing

from both Chamaerops and Trachycarpus and resembling
the American genera Trithrinax and Acanthorhiza. The
spadices in both these new genera appear to be dioecious,
though traces of polygamous spadices have also been
observed.
: Since both these genera are confined only to the
limestone hills in the Peninsula and. have no allies in its
immediate neighbourhood, it is reasonabte to suppose that
they are vestiges of an ancient flora of the Peninsula; and
since the nearest allies are in America, in the sub-temperate
regions of Asia and in the Mediterranean countries, it seems
reasonable also to suppose these two Malayan genera are
immediate descendents of a stock that existed somewhere
at the Cretaceous period and that gave rise also to
Chamaerops, Trachycarpus, Acanthorhiza, etc.

The fact that in some cases open limestone vegetation
produces conditions favourable for the growth of plants of
colder or seasonal climates or of higher altitudes (cf.
KRASAN. in Engl., Bot. Jahrb. II, 1882, pp. 240 & 249-250;
VAN STEENIS in Bull. Jard. Bot. Buitenz. XIII, 1935, p. 296)
might account for the persistence of these two genera in
the Peninsula only on the limestone hills. This hypothesis
would also explain why HENDERSON (Journ. Roy. As. Soc.
Malayan Br. XVII, 1939, pp. 13-87) found that almost all

Vol. XI. (1941).

238

the genera occurring in the peninsular limestones, if not
common in the Peninsula, represented the southernmost
extension of the range of the genera found in the north of
the Peninsula as in Thailand, Burma and Indo-China. On
the other hand, the absence of these two palm genera or
their very close allies in the countries lying immediately
to the north of the Peninsula might be more apparent than
real, for botanically these regions are very imperfectly
known, especially as regards the occurrence and the
distribution of palms; further explorations may reveal in
“ie area the existence of these two new genera or their close
allies.

However, there are cases, though very rare, where
genera or species occur in the Peninsula only on the lime-
stone hills and in distant temperate and sub-temperate
climates but not in other parts of the Peninsula or in
countries lying in the immediate neighbourhood of the
Peninsula. Thus the genus Pistacia (ANACARDIACEAE)
which has a wide distribution from the mountains of Luzon
and Formosa to China, westwards to Yunnan, Upper Burma,
Himalayas, Afghanistan, the Mediterranean region, and the
Canaries in the Old World and in Texas and Mexico in the
New World, is represented by Pistacia malayana on the
limestone hills of Selangor and Upper Perak in the
Peninsula, though the genus has so far not been recorded
either from Lower Burma, Thailand and Indo-China, or
from the northern parts of the Peninsula.

2. Systematic Notes

A. Liberbaileya Furtado gen. nov. (CORYPHOIDEAE)

Palmae nanae, dioceae vel polygamae, caespitosae vel
solitariae. Petioli inermes, apice in costas aequicrassas vel
fere in lamina producti. Laminae frondium flabelliformes.
Spadices interfoliares, simplices (semper?); flores uni-
sexuales vel hermaphroditi. YTepala imbricata; calyx
brevior quam corolla. Stamina 6, in annulum conspicuum
unita ut in LICUALA. Carpella 3, libera. Fructus erectus,
semine aequabile(?). ;

A ‘Labetoaea lankawiensis Furtado sp. nov.

Caulis humilis, ad 3 m altus, caespitosus vel non, inferne
annulatus superne infra coronam foliarum vaginibus fibrosis
caulem amplectantibus, nigris, tectus. Petiolus Mmermis,
metralis, semiteres, utrinsecus angulatus, apice ut rachis
haud productus, sed in uno latere in ligulam parvam et in
altero latere in costas aequicrassas vel fere terminatus.
Lamina frondis 25—40 cm. longa, 34—orbicularis basi cordata,
subtus glaucescens fusco punctulata, junior //, -orbicularis
basi cuneata; segmentis 25-30, unicostulatis, apice cuneatis-
simis, acutis, pungentibus vel breviter bifidis, basi ad 5-12
cm, connatis. Spadix gracilis, unisexualis vel polygamus,

Gardens Bulletin, S.S.

A re i
Vet th a WN)
rhs wy) PEIAS
i

i a
. Liberbaileya lankawiensis (Holotypus: HENDERSON 29134).

a. Folium cum parte petioli superiore. b. Spadix
integer. c. Fructus juvenilis. d. Alabastrum. e. Alabas-
trum verticaliter discissum. f. Carpella. g. Alabastrum,
tepala resecta ut aestivatio staminum apparyeat.

240

simplex, circa 80 cm. longus, inferne ad 50 cm. pedicelli-
formis, cum spathis 2-3 bicarinatis, tubulosis, gramineis,
plus minusve striatis apice oblique truncatis, acutis vel
bifidis involutus, et interdum infra partem ramosam 1-2
bracteolos liguliformes vel lineares ferens; superne in ramos
remotos simplices vel ramulosos primo dense pubescentes
tarde puberulos vel glabros, plerumque in axilla bracteoli
orientes terminatus. Flores feminei, vel interdum hermaph-
roditi, oblongi, 2 mm. longi, 1:2 mm. crassi, in glomerulos
distantes spiraliter dispositos 3-5 aggregati; flores masculi
ignoti. Calyx circa 0-8 mm. longi, glaber, rugosus sepalis
late ovato-reniformibus, imbricatis. Corolla calyce duplo
vel triplo longior, glabra, striata, petalis oblongis, imbricatis.
Stamina vel staminodia 6, in annulum membranaceum
conspicue unita. Carpella 3, ventricosa, dense pubescentia,
libera, in stigma glabrum abrupte terminata. Fructus
juvenilis tantum visus, erectus, vestigio stylare apicale vel
paulo excentrico; semen aequabile(?).

MALAY PENINSULA: Langkawi, Pulau Dayang Bunting
in collina calearea (Henderson, 29134, Holotypus; Curtis,
ace vern. nom. Serdang Batu) ; loc. incert. (Fox in Dec.
1 A

Hermaphrodite flowers which are also seen in this
species resemble the female flowers. The fruits are too
young for an investigation into the nature of the albumen,
the position of the embryo, etc.; in the shape and arrange-
ment of the perianth and of the young developing ovaries
this species appears to be like Maxburretia rupicola.
| CURTIS mentions that the stems reach a height of 6-10
ft., and HENDERSON notes that they attain about 10 ft.
amongst dry rocks, but that the plants are dwarfed when
growing in cracks of the rocks near the sea. That the palm
is caespitose has not been recorded by any of the collectors;
but I have noticed this character in a plant growing in the
Botanic Gardens, Singapore.

B. Maxburretia Furtado gen. nov. (CORYPHOIDEAE)

Palmae dioeciae, nanae vel acaulescentes. Petiols
inermes, apice in costas aequicrassas in lamina producti.
Laminae frondis flabelliformes. Spadix interfoliaris, uni-
sexualis, in inflorescentias partiales divisus; flores solitarii
vel glomerati. Tepala imbricata; calyx brevior quam
corolla. Stamina 6, epipetala. Carpella3,libera. Fructus
erectus; semine erecto; albumine aequabile; intrusione
integumentali brevi, ad latus raphale sito.

Specimen typicum NwuR 34370.

1. Maxburretia rupicola (Ridl.) Furtado comb. nov.

‘Livistona rupicola in Journ. Roy. Asiat. Soc. Straits
Settl. XLI (1903) 41 et Mat. Fl. Malayan Pen., Monocot.,
II (1907) 16: partim; Becc. in Webbia V (1921) 16 et 20

Gardens Bulletin, | S.S.

Se

370).
b. Pars spadicis

Maxburretia rupicola (Apotypus: Nur 34
a. Folium cum parte petioli superiore.
c. Fructus juvenilis. d. Fructus verticaliter dis-
cissus ut albumen integumentumque appareant. e. Flos
f. Flos verticaliter discissus.

feminei.
femineus post anthesin.
g. Alabastrum. h. Petala cum staminibus.

‘ol. XT. (1941).

242

(§ LIVISTONELLA) ; Ridl., Fl. Malay Pen. V (1925) 23 p.p.;
Bece. in Cale. Annals XIII (1933) 105 t. 53 (§ LiIVvIs-
TONELLA) ; Henderson in Journ. Roy. Asiat. Soc. Malayan
Br. XVII (1939) 85 p.p.. (ex altera parte = Liberbaileya
lankawtensis) .

Planta nana, vel acaulescens, solitaria (semper?),
dioecia. Petiolus metralis, semiteres, inermis, postice
convexo planus, untrinsecus angulatus, apice in costas fere
aequicrassas in lamina percurrens, basi vagina fibrosa,
brunneo-nigrescente, liguliforme, 10-20 cm. longa, caulem
amplectante, in senectute in fibras persistentes soluta.
Lamina frondis 50-60 cm. longa, flabelliformis, 4/-orbi-
cularis, subtus albescens, basi cuneata, apice in lacinias
acutas 12-18 cm. longas, aut 30-35 unicostulatas 2 cm. latas,
apice bifidas, aut pauciores, latiores, 2—4 costulatas, apice tot
fidas quot costulas. Spadix unisexualis, 45-60 cm. longus,
masculus femineo longior, in 4—6 inforescentias partiales,
pubescentes, masculas femineis ramosiores, divisus. Spathae
brunneae, glabrae, valde striatae, acutae, vel acuminatae,
interdum bifidae. Flores solitarii, rarissime 2-3 glomerati,
spiraliter dispositi; masculi circa 25 mm. longi, lineares
vel obovato-oblongi, 0-8-1 mm. crassi; feminei ovato-oblongi
paulo minores masculis. Calyx circa 0.8 mm. longus; sepala
imbricata, ovata, acuta, rugosa. Corolla calyce triplo
longior ; petala striata, imbricata. Stamina 6, petalis paulo
breviora, epipetala, filamentis subulatis, antheris cordatis.
Staminodia petalis valde minora, 6. Carpella 3, valde
ventricosa, fusiformia, dense sericea, 1-3 fertilia, singula
in stylum conicum liberum glabrum subito contracta;
stigmate punctiforme. Carpellodia minuta, glabra. Fructus
juvenilis tantum visus, glaber, erectus, obovato-oblongus ;
semine erecto; albumine aequabile; intrusione integumentali
breve; embryone non viso; vestigio stylare apicale vel paulo
excentrico.

MALAY PENINSULA: Selangor (Gua Batu, Ridley 8285,
Lectotypus); Batu Takun prope Kanching (Nur, 34570,
Apotypus).

_ RIDLEY notes that this produces a stem about 3 ft. high
and 6 inches through; but NuR informs me that he found
it always stemless and solitary. The situation of NUR’S
collection was steep, exposed limestone rocks where the
vegetation usually makes a poor growth.

The basinym, L. rupicola, was based on two specimens,
one collected at Gua Batu by RIDLEY and the other in the
Langkawi Islands by CuRTIS. As the two specimens
represent two distinct species, and as the Gua Batu one
appears to have been the principal basis of the description
given by RIDLEY, I have typified the species on the last
named specimen, making the other specimen the paratype
of Liberbaileya lankawiensis.

Gardens Bulletin, S.S.

243

The leaf in RIDLEY 8285 contains 2-4 costulate
segments; and since this was unusual in the genus Livistona ~
except in juvenile stages, BECCARI employed the character
to create the section LIVISTONELLA under Livistona. But
the many stages represented in NUR’S collection lead me to
conclude that the degree of leaf division in RIDLEY 8285
represents that of young plants.

There are no flowers in the lectotype. preserved in
Singapore, a reason why BECCARI failed to discover the
erroneous generic identity made of the specimen by RIDLEY.
However NUR 34370 has supplied the important diagnostic
characters lacking in RIDLEY 8285, for which reason NUR’S
specimen has been called here the Apotype of the species
(for the definitions of the nomenclature of types, see
FURTADO in Gard. Bull. Straits Settl. 1X, 1937, pp. 285-309).

Vol. XT. (1941).

244

ALOCASIA MACRORRHIZA AND ITS VARIETIES
By C. X. FURTADO,

Botanic Gardens, Singapore

1. Introduction : ap 244
2. The Authorship of the Name ag 244
3. The Typification of the Species os 245
4. Splitting the Species i 247
5. Confusion with Alocasia odor a 248
6. Confusion with Colocasia and Xanthosoma
Species Z, i aye 250
7. ‘Varieties . oF Zon
8. Varieties Deen vel Excludendae Tes Zoe
9. Literature consulted Pas Ni 256

1. Introduction

The binomial Alocasia macrorrhiza has been used in
more than one sense. Some of the plants which have passed
under this name are economically very important and found
in cultivation or as escapes in the tropics of both
hemispheres; others are of little economic value and not
known in cultivation or as escapes except in their native
country and perhaps in some botanic gardens. Further,
many experiments and analyses have been made in order
to test their value as food to men and animals, and the
results have been sometimes conflicting. The existence of
many varieties in some of the species called by this name
is also a factor that has caused further ambiguity. The
present paper is an attempt towards the clarification of this
ambiguity.

It is my pleasant duty to record here my thanks to
Dr. R. C. BAKHUIZEN VAN DEN BRINK, Agricultural
Department, Buitenzorg, Dr. G. J. A. TERRA, Horticultural
Division, Batavia, Dr. C. G. G. J. VAN STEENIS, Botanic
Gardens and Herbarium, Buitenzorg, Dr. K. BIswAs, Royal
Botanic Gardens, Calcutta, and Mr. E: FE. ALLEN,
Agricultural Department, Teluk Anson, Perak, for procur-
ing for me planting material of the species and varieties
connected with the nomenclatural problems of Alocasia
macrorrhiza and its varieties; without this material it would
have been impossible for me to clear many. difficulties
discussed in this paper.

2. The Authorship of the Name

The genus Alocasia was published first by NECKER,
Elem. Bot. III (1791 p. 289), a work proposed for rejection
by some botanists (see FURTADO, 1937, p. 252 footnote).

Gardens Bulletin, SS.

245

But whether by rejecting this we should also reject A.
macrorrhiza (L) Sweet and A. cucullata (Lour.) Sweet [ .
have not been able to investigate; both these names have
been recorded with their respective basinyms in STEUDEL’S
Nomenclator Botanicus ed. 2, vol. I (1840), though the
authorities so far consulted by me attribute these combina-
tions to SCHOTT (1854). No doubt in Index Kewensis these
two combinations have been recorded as if they were
published by SCHOTT et ENDLICHER in their Meletemata I
(1832 p. 18); but this is an error, for there Alocasia was
accepted only as a section of Colocasia. Since NECKER’S
work is not yet rejected, I propose in this treatment to
follow STEUDEL and cite the name as Alocasia macrorrhiza
(L) Sweet.

3. The Typification of the Species

Alocasia macrorrhiza (LL) Sweet is an isonym derived
from Arum macrorrhizon, a binomial coined by LINNAEUS
in Species Plantarum ed. 1 (1753, p. 965) for a species
from Ceylon. The specific prologue! given by LINNAEUS
runs as follows :—

“Arum acaule, foliis peltatis cordatis repandis:
basi bipartitis: Fl. Zeyl. 327. Roy. lugdb. T[76].

Arum maximum macrorrhizon zeylanicum
Herm. par. TS 1.13.- Raz. Suppl. 574.

Habitat in Zeylona, perennis.”’

I have not been able to consult the last reference given
by LINNAEUS. In the second reference the number ‘7”’
appears to be a misprint of “6”, but in neither of these
references (6 or 7) are the leaves described as peltate.
In the third reference (Herm. Parad. Bat. p. 73, t. 73),
the species is described and depicted with sagittate leaves.
In Flora Zeylanica (the first reference given by LINNAEUS),
reference is made to ROYEN’S (the second reference) and
HERMANN’S (the third reference) species, but the plant is
described as peltate-leaved.

I have not been able to consult two more books cited
under the species in Flora Zeylanica, and so I do not know
_ whether these references refer to any peltate-leaved species.
However, there is evidence to show that the Linnean phrase
“foliis peltatis, basi bipartitis’ refers to sagittate leaves
in the modern sense. Thus the sixth Arum in Hort. Cliff.
p. 485, which is said in Flora Zeylanica to have the same
appearance as Arum macrorrhizon, is described as having
“foliis peltatis, ovatis, basi bipartitis’, a -description
subsequently expanded to mean “modo latus foliis a petiolo

1. A PROLOGUE is the printed matter (description, etc.) accom-
panying the original publication of a name or epithet.—A. J. WILMOTT
in Journ. Bot [London] LXXVITI (1939) 206. :

Vol. XI. (1941).

246

ad basin totum bipartitum” (RICHTER, 1840). LINNAEUS
also described his Arum peregrinum in Hort. Cliff., p. 435,
No. 7, as having leaves “folia peltata, cordata usque ad
petiolum”’; Arum peregrinum L. is an accepted synonym
of Alocasia macrorrhiza.

The other alternative is to assume that LINNAEUS
confused a species of Colocasia with Alocasia macrorrhiza,
overlooking the warning given by HERMANN (1698, p. 74) in
the following observation under his Arum maximum
macrorrhizon zeylanicum, t. 73:

“Idem Arum videtur quod COMELINUS in Catal. Horti. Med. Amstel.
nominat Arum maximum Ceylanicum radice crassa longa rotunda
geniculata Colocasiae folio. Welila Hort. Malab. part. 13 nondum
editae. Ast appellatio isthaec non arridet, Nostri enim folia nihil cum
Colocasiis habent commune, sed forma ad Arum accedunt, pedicu-

lumque cui incumbunt non in medio sed extremo foliorum insertum
habent.”

Welila or Weli-Ila in RHEEDE’S Hort. Malab. XI, t. 22, is a
special variety of Colocasia esculenta which produces long,
thick, edible subterranean suckers, and it is possible for
persons not acquainted with the plants to apply the
description of the runners of this variety of Colocasia to the
thick, edible, sometimes inclined, rhizomes of Arum
macrorrhizon (quoad HERMANN’S plate).

But there is not the slightest evidence to support this
view. On the contrary there are abundant reasons to justify
the typification of Arum macrorrhizon on HERMANN’S
plate 73: (1) Arum macrorrhizon has always been inter-
preted as a_ species of Alocasia and not of Colocasia (both
genera in the modern sense); (2) “foliis peltatis, basi
bipartitis” in Linnean sense probably means sagittate leaves ~
with a somewhat acute sinus at base: (3) Flora Zeylanica,
which supplied the definitive phrase cited under Arum
macrorrhizon L., was based mainly on HERMANN’S
herbarium (TRIMEN, 1887; BOULGER, 1900; ARDAGH, 1931;
OOSTROOM, 1937); (4) HERMANN’S plate and description
indicating an Alocasia with sagittate leaves was cited both
in Flora Zeylanica and again independently in the prologue
of Arum macrorrhizon; (5) the specific epithet macrorrhizon
adapted by LINNAEUS was derived from HERMANN’S specific
phrase-name quoted in the prologue of A. macrorrhizon;
(6) there are no specimens of the species either in the
Linnean herbarium in London (JACKSON, 1912), or in
HERMANN’S collection in Leiden (OOSTROOM, 1937), and the
drawing which exists in HERMANN’S herbarium in the
British Museum, London, represents the same species as
HERMANN’S plate; (7) the species depicted by HERMANN
is a common species in Ceylon, specially in the lowlands
where HERMANN made the bulk of his Ceylon collection

Gardens Bulletin, S.S.

247

(TRIMEN, 1887; BOULGER, 1900) ; (8) no species of Alocasia
having large peltate leaves has been discovered in Ceylon °
(PETCH, 1919; ALSTON, 1931).

4. Splitting the Species

The particulars given by HERMANN (1698) concerning
the vernacular name and economic uses coupled with the
description and the plate led some early botanists to
interpret Arum macrorrhizon L. in the sense indicated by
the present typification. Thus G. FORSTER (1786 & 1786
bis) identified as Arum macrorrhizon, the sagittate-leaved
Alocasia cultivated in Ceylon, New Zealand and Polynesia
for its edible stems and depicted by RUMPHIUS as Arum
mdicum sativum, Herb. Amb. V (p. 306, t. 106) (FORSTER
gave an erroneous description of the flowers). ROBERT
BROWN (1810, p. 192) also cited Arum macrorrhizon L.,
A. maximum macrorrhizon zeylanicum Herm. Parad. 73, t.
73, and A. indicum sativum Rumph. Amb. V (p. 308, t. 106)
under his Calladium ? macrorrhizon, observing at the same
time that both HERMANN’S plate and RUMPHIUS’S plate
(without the spadix) were good representations of the plant
in question.

LOUREIRO (1790), who was working in Cochin-China
with a small library at his disposal, was probably not able
to consult HERMANN’S Paradisus, or any other books which
identified A. macrorrhizon with the large sagittate-leaved
Alocasia furnishing edible stems. The Linnean statement
that A. macrorrhizon was stemless had raised difficulties
even to LINNAEUS, who suggested the possibility of Arwm
mdicum sativum Rumph. being a different species, Arum
arborescens (cf. STICKMANN, 1754). LOUREIRO was also
misled by the phrase ‘“‘foliis peltatis’”’ used by LINNAEUS in
the prologue of Arum macrorrhizon. Consequently
LOUREIRO identified with the last a stemless species, having
peltate leaves and [practically?] no appendix to the spadix
(probably Colocasia gigantea), and named as Arum indicum
the sagittate-leaved species depicted in the Rumphian plate
106 and cultivated also in Cochin-China for its edible stems.

The name Arum indicum of LOUREIRO was adopted
by ROXBURGH (1832) for an edible Alocasia cultivated in
Bengal. ROXBURGH found some difficulties in identifying
A. macrorrhizon owing perhaps to the statement in the
prologue that the species was stemless; he was thus inclined
at first to identify A. macrorrhizon with A. montanum Roxb.
(see ROXBURGH’S remarks under the latter op. cit., p. 497).

Since ROXBURGH had described Arum indicum in detail,
giving its vernacular names and economic uses, and since
LOXBURGH’S drawing of the plant was later published by

Vol. XI. (1941).

248

WIGHT (Ic. Pl. Ind. III, t. 794), the name Arum indicum
and its isonym Alocasia indica became widely current in
India: for the plant cultivated in Bengal.

Since there were some discrepancies in the descriptions
of Arum indicum Lour., A. indicum Lour. sensu Roxb. and
A. macrorrhizon L., SCHOTT (1854) was inclined to consider
these three descriptions as representing three distinct
species. This view of SCHOTT was contested by KOCH
(1854) who maintained that all these three represented
but one single species. But SCHOTT’S opinion, after his
defence in pointing out the discrepancies (1855) and the
revision of all the Aroids then known (1860), was followed
by many botanists, notably ENGLER (1879) and later by
ENGLER & KRAUSE (1920).

MIQUEL (1855) seemed to have agreed with KOCH in -
reducing Arum indicum Lour. sensu. Roxb. to Alocasia
macrorrhiza, but for some unaccountable reasons he also
retained Alocasia indica with the synonym Arum indicum
ou or Roxb. as the correct name for Caladium giganteum

PETCH (1919) pointed out that in Ceylon there is only
one large-leaved species of Alocasia and that this should be
called A. macrorrhiza, that specimens of this species from
Ceylon were very rare in Europe, the only one he found
being THUNBERG’S, and that the plants which pass in Ceylon
as A. indica are cultivated species of Xanthosoma. ALSTON
(1931) further pointed out that the only large-leaved
Alocasia that occurs in Ceylon has sagittate leaves, that on
geographical considerations the name A. macrorrhiza should
be applied to this species, and that the real A. indica is
probably a variety of A. macrorrhiza. ALSTON did not
consider the problem either of the interpretation or of the
typification of A. macrorrhiza on the original description
of the species; nor did he consider whether Arum odorum
Roxb. or Caladium odorum Roxb. ex Lindl. was a distinct
Species or not.

Perhaps influenced by the statements of ALSTON,
MERRILL (1935), who had previously (1922) maintained
Alocasia indica and A. macrorrhiza as specifically distinct,
the former not occurring in the Philippines, considered A.
macrorrhiza, A. indica and Arum indicum Lour., and Arum
indicum Rumph. as representing the same species. He did
not refer to the confusion of other species either with A.
indica or with A. macrorrhiza.

5. Confusion with Alocasia odora
There is an Aroid growing wild in Burma, Khasia,
China, etc. which has peltate leaves and which was described
as Arum odorum by ROXBURGH (1832) and figured by

Gardens Bulletin, SS.

249

WIGHT (le. Pl. Ind. III, t. 797), This species was named
by KocuH (1854) as Alocasia odora and by SCHOTT (1854)
as A. commutata. It does not occur in Ceylon, Malaya,
Polynesia, etc. either in cultivation or wild.

But ten years before the posthumous publication of the
binomial by ROXBURGH (1832), LINDLEY (Bot. Reg., 1822,
t. 641) published under Caladium odorum a Latin version
of ROXBURGH’S manuscript description of Arum odorum
and a plate which in general aspects of the plant appears to
correspond to A. indicum of ROXBURGH, though the inside of
the spathe is depicted green. ENGLER (1879) and ENGLER
& KRAUSE (1920) who keep Arum macrorrhizon L. and
A. odorum Roxb. as two distinct species with Alocasia
macrorrhiza and A. odora as the correct names, identified
Caladium odorum Lindl. in Bot. Reg., t. 641, as A.
macrorrhiza which normally produces a whitish or yellowish
spathe-limb.

Since the drawing of the entire plant in LINDLEY’S
plate was made from the living plant, whereas the details
of the spathe, spadix, flowers, etc. were drawn from the
specimens communicated to LINDLEY, it might seem
legitimate to suspect that the different colours in these detail
figures were given in order to make them conform not with
the colours of the material (which might have faded) but
with the colours indicated by ROXBURGH’S description
quoted. However, this explanation does not receive any
support from the plate published of apparently the same
species, by HOOKER in Bot. Mag., (1842), t. 3935, under
the name of Colocasia odorata; in this plate the colour of the
spathe agrees with that given by LINDLEY in Bot. Reg., t.
641. However, if the determination by ENGLER (1879) is
_ correct, and ENGLER had examined practically all the

available Aroid material in the great botanical institutions
in England, it is possible that the peculiar colour variation
is a product of the stove environment in which the plants
are grown in Europe. The two plates are so poor that it
is difficult to decide whether the identification made by
ENGLER (1879) is right or not. There is, however, ample
evidence to show that the name A. odorum and its isonyms
were used from LINDLEY (1822) onwards in more than one
sense, perhaps for any large-leaved Alocasia grown in hot-
houses of Europe (cf. the synonymy given by ENGLER
(1879) and ENGLER and KRAUSE (1920) under Alocasia
indica, A. macrorrhiza and A. odora).

This indifferent use of the name Arum ‘odorwm and of
its isonyms, and the misleading use of the word “peltate”
by LINNAEUS in describing his Arum macrorrhizon must
have led THWAITES (1864), TRIMEN (1885 & 1898),
HOOKER (1893), RIDLEY (1907 & 1925), KooORDERS (1911),

Vox. (194 ).

250

HAINES (1924) and FISCHER (1931) either to describe
Alocasia macrorrhiza as having peltate leaves or at least to
reduce Arum odorum or Alocasia odora as its synonym, or
to do both, and then maintain A. indica as a distinct species.
The synonymy given by ENGLER (1879) and ENGLER &
KRAUSE (1920) in order to show the different uses of Arum
odorum or its isonyms was probably not understood because
these authors failed to note that, though Caladium odorum
Lindl., Bot. Reg. (1822), t. 641, quoad descriptionem was
Arum odorum Roxb., C. odorum Lindl., Bot. Reg. (1822),
t. 641, quoad tabulam was, in their opinion, a different plant
identical with A. macrorrhiza.

HAINES (1924), who noted the differences given by
ENGLER & KRAUSE (1920) between A. macrorrhiza and A.
odora and who had opportunities to study A. odora, could
not understand the distinction given by ENGLER & KRAUSE,
chiefly because he failed to recognize that A. macrorrhiza
sensu lato included the plants he described as A. indica.
Hence HAINE’S remarks that, if one were to follow ENGLER
& KRAUSE (1920), A. macrorrhiza should be used for Ceylon ~
plants and A. odora for the Indian ones. ALSTON (1931),
who pointed out certain discrepancies in the colour of the ~
spathes of Bot. Reg. t. 641, and in those of the Ceylon
plant, noted that the Ceylon plants referred to A.
macrorrhiza have sagittate leaves and on geographical
grounds appeared to be correctly named.

6. Confusion with Colocasia & Xanthosoma Species

In some parts of India, where Alocasia odora and A.
macrorrhiza occurred, it was not difficult to use A.
macrorrhiza for A. odora, and A. indica for A. macrorrhiza.
But in.Ceylon (PETCH, 1919; ALSTON, 1931) and in Malaya,
where no Alocasia with a large rhizome and large peltate
leaves occurs, the use of the two binomials, A. macrorrhiza
and A. indica, would have been extremely difficult, had it
not been for the fact that cultivated species of Xanthosoma
were, owing perhaps to their edible tubers and large leaves,
mistaken for Asiatic plants. Thus THWAITES (1864),
TRIMEN (1885, 1898), WILLIS (1911) & PARSONS (1926)
recorded A. indica for Ceylon as if it were an exotic plant
existing only in cultivation, and made no mention of any
Xanthosoma species which have been long in cultivation
there. According to PETCH (1919) the vernacular names
cited under A, indica by TRIMEN and others also suggest
the cultivated Xanthosoma species of Ceylon. BARRETT
(1910) also mentions his receiving Xanthosoma material
from Singapore, Buitenzorg, eic., .as Alocasia indica,
Alocasia violacea, A. macrorrhiza and A. javanica, or as

Gardens Bulletin, S.S.

251

Colocasia monorrhiza scriptum and Colocasia antiquorum
niger. And I have received Xanthosoma material from
Calcutta as Alocasia violacea (a binomial usually reduced
as a variety of A. indica).

Many factors must have contributed to allow this
masquerade of Xanthosoma spp. as A. indica to pass
undetected for so long. The basal sinus of the leaves of
the edible Xanthosoma spp. cultivated in the East is wide,
and since that of A. macrorrhiza is very narrow and was
described in most books as peltate, it was easy to take the
Xanthosoma species with a wide leaf sinus as fitting with
the description given under A. indica. Besides A. indica
is described as a large-leaved plant cultivated for its edible
tubers or rhizomes, a description also applicable to the
species of Xanthosoma. Further, owing to the large size
of the leaves, herbarium specimens of these plants are
usually very poor and are rarely made. Moreover the
cultivated Xanthosoma species flower but rarely and without
flowers it is not easy for systematists to identify the plants,
the general tendency in such cases being to identify the
plants on their vernacular names, economic uses, the
country of origin, etc. (a procedure eminently suited for
perpetuating the confusion if already made). In species
of both these genera there are many forms, some with
violet petioles. In fact BARRETT (1910) experienced such
great difficulty in distinguishing, on sterile material, the
species of Xanthosoma from those of Alocasia that he
evolved a key based entirely on the openness of the leaf
sinus, a key that, while enabling him to identify correctly
the best kinds of Xanthosoma in cultivation, misled him to
classify under Alocasia some inferior kinds of Xanthosoma.

Apparently the confusion existing in several botanical
works coupled with the absence of a satisfactory key to
enable one to distinguish generically between these plants on
vegetative characters led OCHSE (1931) to record under
Colocasia esculenta some information received by him
concerning the uses and the vernacular names of the
cultivated varieties of Alocasia, Colocasia and Xanthosoma,
and also to suppress altogether any discussion on Alocasia,
though A. macrorrhiza plays a very important role in the
diet of the peoples of the Netherlands Indies, much more .
important than many other plants discussed in the book,
e.g. Schismatoglottis. calyptrata. MACMILLAN’ (1935)
considered all edible species of Alocasia and Xanthosoma as
if they were generically identical and depicted a species of
Xanthosoma with ‘“‘Alocasia (Xanthosoma) indica var.” for
its legend, though he was personally inclined to regard all
these edible species of Xanthosoma and Alocasia as
varieties of Colocasia esculenta.

Vol. XI. (1941).

252

Such a confusion, very common in books written before
the beginning of the twentieth century, has now decreased
considerably, apparently due mainly to the writings of
BARRETT (1910) and of ENGLER and KRAUSE (1920), who
worked on many specimens derived from cultivated material
in Asia and America and who quoted in some cases
vernacular names.

7. Varieties

Some of the varieties here described are so distinct that
it may be better to separate them as species or sub- |
species; however I have retained the varieties, firstly to
show their affinities, secondly to warn systematists and
agriculturists against the confusion so often made over
these names, and thirdly to conform with current procedure.
In horticulture all these varieties usually pass as species.

The varieties described below may be divided into two
large groups: (1) one having the spathe very much longer
than the enclosed spadices: vars. typica and nigra; and (2)
the other having the spathes nearly as long as, or slightly
longer than, the spadices: vars. variegata, marmorata (?),
and rubra. There is a tendency to restrict the binomial A.
indica to the second group because it includes Arum indicum
Lour. sensu Roxb. (1832), but Arum indicum Lour. itself
appears to be the typical Alocasia macrorrhiza. If the
second group were separated specifically from A. macrorr-
hiza, then the status of Arum cordifolium Bory (1804),
Arum punctatum Desf. (1829), ete. will have to be
investigated as they seem to have a better claim than
Alocasia indica, if this is typified on ROXBURGH’S description.

1. Alocasia macrerrhiza (L.) Sweet in Steudel (1840)56;
Schott (1854)409 and (1860)146; Engler (1879)
502; Engler et Krause (1920)84: cf. also Petch
(1919), Alston (1931), Merrill (1935) and Furtado —
(1940). |
A. indica (Lour.) Schott (1854) 410: cf. also Koch
(1854), Schott (1860), Engler (1879), Hooker f.
(1893), Engler et Krause (1920) and Ridley (1925).
Arum indicum Lour. (1790)537 and Roxb. (1832) 498.
Arum macrorrhizon L. (1753) 965.
Colocasia indica (Lour.) Kunth, Enumer. Ili (1841)
539; Hassk. in Hoev. et De Vriese, Tijdsch IX (1842)
160 p.p.; Engler (1879) 494 p.p.; Engler et Krause
(1920)69 pro parte typica (ex altera parte=
Colocasia gigantea).
The binomial Alocasia macrorrhiza must be typified on
Arum maximum macrorrhizon zeylanicum Herm., Parad.

Gardens Bulletin, S.S.

5 2 . - -
| ii
Ps ey
.

253

Bat. (1698) 73 t. 73 which represents the large green form
common in Malaysia and Ceylon. The type was an edible
race cultivated in Ceylon.

A faulty terminology employed by LINNAEUS in
describing this species misled some botanists (e.g. HOOKER,
1893) to use the binomial to denote a mountainous or semi-
tropical species from India whose correct name should be
Alocasia odora (Roxb.) Koch. The latter has peltate leaves,
whereas the leaves of the former are sagittate. In India
the real A. macrorrhiza goes under the name of A. indica
(Lour.) Schott, though there are evidences that some species
of Xanthosoma are also included under that name.

In Ceylon and Malaya and other tropical countries
where A. odora does not occur even as an escape and where
A. macrorrhiza appears to be native, the name A. macrorr-
jiuza is usually correctly employed, but almost all its
published descriptions and the synonymies reveal the
influence of HOOKER f. who was one of the first to use A.
macrorrhiza as the correct name for A. odora (cf. also
THWAITES, 1864, and HooKER f. 1893); but the binomial
A. indica has been generally used to name the edible species
of Xanthosoma long introduced in these parts from
America. Some races of the type form of Alocasia
macrorrhiza furnish a very palatable rhizome, but others
are too acrid to be fit for human consumption. J have not
been able to detect any external differences in these races.

2. ‘Alocasia macrorrhiza var. variegata (Schott) Furtado
comb. nov.

“A. indica var. variegata (Koch et Bouché) Schott
(1860)145; Engler (1879)502; Engler et Krause
(1920) 88.

A. macrorrhiza var. brisbanensis Bailey, Queens]. Flora

V (1902) 1697. Syn. nov.

A. variegata Koch et Bouché in Index Sem. Hort.
Berol,, 1851 app. p. 5.

I have not been able a consult the original description
of this variety, and so I have followed the interpretation
given to it by ENGLER (1879) and ENGLER and KRAUSE
(1920) who had access to KoOCH’S original description as
well as to the living plants, or their progeny, and the type
herbarium material studied by KOCH.

It is a smaller form than the type and includes the
races described extensively by ROXBURGH (1832) under
Arum indicum Lour. It is characterised by the presence
of transverse brownish or purplish, irregular lines or marks
on the petiole. J have seen only two races of this variety;
in the one the markings are very conspicuous, and in the
other they are faint. The former race appears to be the

WOMANS (19LFY.

254

typical form of this variety, while the latter is cultivated
in Java and Bengal for its edible rhizomes, and also for
its marble-sized tubers produced at the end of long root-like
structures. ALSTON (1931) records the presence of this
variety in Ceylon. Apparently both races furnish edible
tubers.

ae VAlocasia macrorrhiza var. marmorata Furtado var. nov.

Planta minor quam forma typica. Petiolus viridis
maculis albescentibus, ivregularibus, paucis varie gatus.
Foliorum lamina maculis similibus, majoribus conspicuis
praedita, interdum per unum latus totum albescens.

This inedible variety is often cultivated in gardens as
Alocasia variegata or Alocasia indica var. variegata, a name
which, according to the descriptions given by ENGLER
(1879) and by ENGLER and KRAUSE (1920), should be
applied to the variety having transversely clouded petioles.
The petioles in this variety have a few large, creamy spots
and the leaf-lamina bears large similar markings; some-
times the entire half is cream.

The type specimen described here was taken from
plants grown in the Singapore Botanic Gardens from tubers
received from Calcutta Botanic Gardens.

4, Alocasia macrorrhiza var. nigra Furtado var. nov.

A forma typica, cut in aspectu et dimensionibus similis,
haec varietas differt petiolis et nervibus primariis im
superficie laminae inferiore nigro-purpurascentibus.

This form is large-leaved like the type form but has
purple or violet colour in petioles and in the primary nerves
in the lower surface of the leaves. In some measure it is
easy to confuse this variety with A. macrorrhiza var. rubra,
from which it is easily distinguished by the larger leaves,
by the absence of purplish colour in the lower surface of the
leaves (excepting nerves), by the upper leaf-surface being
dark green (not dark olivaceous green), by the spathes being
much longer than the spadices. The young leaves are
usually green whereas in the var. rubra they are practically
purplish in both surfaces.

The plant is quite common in Malaya and often passes
as A. indica var. metallica and A. indica var. violacea. The
stems attain a height of about 6 feet or more, but they are
at edible even after a prolonged boiling.

Alocasia macrorrhiza var. rubra (Hassk.) Furtado comb.

nov.

A. indica var. metallica Schott (1860)145 (the
trinomial was not published by Schott, though the
variety was indicated); Engler (1879)502; Engler
et Krause (1920)88.

~ Gardens Bulletin, S.S.

255

Yq. indica var. rubra (Hassk.) Engler in Engler et
Krause (1920) 88.

Colocasia indica var. rubra Hassk., Pl. Jav. Rar.
(1848) 145.

C. odorata var. rubra Hassk., Cat. Bogor. II (1844) 55.

This very ornamental variety has much smaller leaves
and spathes than A. macrorrhiza var. nigra with which it is
often confused. The spathe in var. rubra is slightly longer
than the spadix, and the purple colour is present in the
lower surface of the leaves. Juvenile leaves are nearly
purple in both surfaces, but the upper surface becomes
dark olivaceous green having a very pleasant polish.

The plant is used medicinally by Javanese who also
grow the plant around their houses as a protection against
thieves and burglers; for the Javanese believe that this
plant has the property of causing a cough in anyone who
approaches the plant at night, and coughing would warn the
inmates of the houses of the presence of. any unwanted
guests in the neighbourhood.

8. Varietates Dubiae vel Excludendae

1. Alocasia indica var. diversifolia Engl. in Engl. et
Krause (1920) and its synonym A. indica var. heterophylla
Engl. (1879) are synonyms of A. Portei (Shott) Engl. cf.
Merrill (1922).

2. Alocasia indica var. violacea Engl. (Engler et
Krause, 1920).

This was based on a specimen cultivated in the Royal
Botanic Gardens, Calcutta. Tubers received from these
Gardens under the name of A. violacea proved to be of a
form of Xanthosoma violaceum. ENGLER based this variety
apparently on sterile material and the description, which is
applicable also to X. violaceum, is insufficient to determine
its exact identity.

3. Alocasia talihan Elmer (1938).

The type of this species was cited by MERRILL (1922)
under A. macrorrhiza. ELMER, finding it to be quite distinct
from this species, described it as new, stating at the same
time that it might not be an Alocasia at all. Since in the
description of the spadix ELMER did not note a sterile
appendix which he described in a true Alocasia species, A.
talihan is, I suspect, a species of Xanthosoma. ELMER
recorded that the species was cultivated for its edible tubers.

4. Colocasia indica var. atroviridis Hassk. (1844).

HASSKARL did not give any description of this variety
but referred it to ‘Ari indici secunda species Rumph. amb.
V. p. 308”. The Rumphian description is too meagre for
its certain identification; but the words “caules subtus fusci

Vot.. XP, €1941 ).

256

sunt, at talibus lineis seu venis distincti, et maculati’”’ suggest
the form described here as A. macrorrhiza var. variegata.

5. Colocasia indica var. pallida Hassk. (1844).

HASSKARL referred this variety to “Ari indici tertia
species Rumph. amb. V p. 308’. It is difficult to identify
this variety from the Rumphian description, though “caules
albicantes, seu pallide virentes” suggest Colocasia gigantea.

6. Colocasia odorata var. viridis Hassk. (1844).

This was a varietal name proposed for the type form of
what HASSKARL called C. odorata; but it is difficult to find
how he really interpreted this species. C. odorata was
formerly used indiscriminately to name any large species
of Alocasia cultivated in the European hot-houses and the
binomial was supposed to be an orthographic variant of C.
odora Lindl. (1822).

9. Literature Consulted

ALsTon, A. G. H.—FIl. Ceyl. VI Suppl. (1931) 296-297.

ARDAGH, J.—Paul Hermann’s Ceylon Herbarium & Icones—
Journ. Bot. LXIX (1931) 137-138.

BARRETT, O. W.—Promising Root Crops for the South—U. S.
Bureau Pl. Ind. Bull. 164 (1910).

Bouucer, G. S.—History of Ceylon Botany—Trimen, FI. Ceyl.
V App. IV (1900) 367.

5. Brown, R.—FIl. Nov. Holl. (1810) 192.

6. Exumer, A. D. E.—Leafl. Philipp. Bot. X (1938) 3614.

7. ENGLER, A.—-DC., Mon. Phan. II (1879) 490-510.

8.

9

oe eee

a

ENGLER, A. et K. KrRAuse—Engl., Pflanzenreich IV, 23E.,
Heft 71 (1920) 1-132.
. FIscHer, C. E. C.—Fl]. Pres. Madras IX (1931) 1581-1582.
10. Forster, G.—Fl. Insul. Austral. Prodr. (1786) 62.
11. Forster, G.—De Plant. Escul. (1786) 58-59.
12. FurtTapo, C. X.—The Malayan Keladis and other Edible
Aroids—M. A. H. A. Mag. X (1940) 11-17.
13. Furtapo, C. X.—A Commentary on the Laws of Botanical
Nomenclature—Gard. Bull. Straits Settl IX (1937)
223-284,
14. JACKSON, B. D.—Index to the Linnean Herbarium (1912).
15. HASSKARL, J. C.—Cat. Hort. Bogor. II (1844) 56-57.
-16. Hatnes, H. H.—Bot. Bihar and Orissa V (1924) 868-871.
17. HERMANN, P.—Parad. Bot. (1698) 73-75, t. 73.
18. Hooxker, J. D.—FIl. Brit. Ind. VI (1893) 523-528.
19. Hooker, W. J.—Colocasia odorata—Bot. Mag. (1842) t. 3935.
20. Kocu, C.—Append. Gen. et Spec. quae Hort. Bot. Berol. 1894
—(cited by Schott and others).
21. Koorvers, S. H.—Exkursionfl. Jav. I (1911) 261-262.
22. LINDLEY, J.—Caladium odorum—Bot. Reg. (1822), t. 641.
23: LINNAEUS, C.—F Il. Zeyl. (1747) 155.
24. LINNAEUS, C.—Hort. Cliff. (17387) 4385 (cited by Ridhber}.
25. LINNAEUS, C.-_-Spec. Plant ed. 1 (1753) 965.
26. Loureiro, J.—Flor. Cochinch. (1790) 535-536.
a. MACMILLAN, H. F.—Trop. Plant. and Gard. ed. 4 (1935)
288-289, |
28. MERRILL, E. D.—Enum. Philipp. Pl. I (1922) 185.
29. Merritt, E. D.—A Commentary on Lovreiro’s Flora
Cochinchinensis—Trans. Amer. Philos. Soc. XXIV, Part
II (1985) 97-98.

Gardens Bulletin, S.S.

257

30. MiqueL, F. A. G.—FI. Ind. Bat. III (1856) 205.

31. Necker, N. J. de—Elem. Bot. III (1791) 289.

32. OCHSE, x J.—Veget. Dutch East Ind. (1931) 51-58.

3. OosTsTROOM, S. J. vAN—Harmann’s Collection of Gaylon
plants in the Rijksherbarium—Blumea, Suppl. I (1937)
193-209.

34. Parsons, T. H.—AlIph. List Pl. Roy. Bot. Gard. Perad. Ceylon
(1926) 9.

35. PETCH, T.—Alocasia indica Schott—Ann. Roy. Bot. Gard.
Perad. VII (1919) 53-55.

36. RHEEDE TOT DRAAKSTEEN, H.—Hort. Ind. Malab. XI (1692)
aa 4. 22

37. RICHTER, H. E.—Cod. Bot. Linn. (1840).

38. RimpLey, H. N.—Mal. Fl. Malayan Pen., Monocot III (1907)
18.

39. RipLey, H. N.—FI. Malay Pen. V (1925) 99.

40. RoxBurRGH, W.—FI. Indica III (1832) 498-500.

41. ROYEN, A. VAN—FI. Leyd. (1740) 7.

: 42. RUMPHIUS, G. E.—Herb. Amb. V (1750) 308-310.

43. ScuHoTt, H. G.—Oester. Bot. Wachenbl. IV (1854) 409—

(original not seen).
44. ScuottT, H. G—Araceen Betreffendes II (1855) 1-22.
: 45. Scuott, H. G.—Prodr. Syst. Aroid (1860) 140-157. .
: 46. Scuott, H. G. & S. Endlicher—Meletem. I (1832) 18.—

(original not seen).

47. STEUDEL, E. T.—Nomencl. Bot. ed. 2 (1840) 56.

48. STICKMANN, O.—Herb. Amb. (1754) 22.

49. TuHwaliTes, G. H. K.—Enum. Pl. Zeyl. (1864) 336—(nomen-
clature chiefly by J. D. Hooker).

50. TRIMEN, H.—Syst. Cat. Flower. Pl. Ceyl. (1885) 98.

51. TRIMEN, H.—Hermann’s Ceylon Herbarium and Linnaeus’s
Flora Zeylanica—Linn. Soc. Journ. Bot. Lond. XXIV
(1887) 129-155.

52. TRIMEN, H.—FIl. Ceyl. 1V (1898) 359-361.

55. WIGHT, R.—Ic. Pl. Ind. III (1844) tt. 794 & 797.

54. WILLIs, J. C.—Cat. Flow. Pl. Ceyl. (1911) 98 & 149.

Vol. XI. (1941).

258

THE TYPIFICATION OF CALOPHYLLUM CALABA L.
By C. X. FURTADO,
Botanic Gardens, Singapore

In the prologue of Calophyllum Calaba L., Spec. Pl. ed.
1 (1753) 514 LINNAEUS gave a number of references which
show that the species included at least three elements or
Syntypes: (1) a Ceylon element, (2) a Malabar element and
(3) an American element. The American element was
described and figured by PLUMIER under the Caribean name
Calaba (PLUM., Gen., 39 t. 18), and references to PLUMIER’S
plant are found in pre-1753 works of LINNAEUS, namely in
Fl. Zeyl. (1747) 90, no. 202, and in ROYEN, FI. Leyd. Prodr.
(1740) 476, both cited in the prologue of C. Calaba L. In
addition a reference to PLUMIER’S Calaba was made by
LINNAEUS in his Gen. Pl. ed v (1754) 229 and ed. vi (1764)
266, works having an important bearing on the interpreta-
tion of the genera published in LINN., Sp. Pl. ed. i (1753)
and ed. 11 (1763) (Art. 20). The particulars given in the
prologue concerning the habitat of the species includes also
the West Indies, for the habitat is stated to be “‘in Indiis’’,
a phrase frequently employed by LINNAEUS to indicate that
a plant occurs in the East as well as in the West Indies.

In view of this inclusion of the American element in the
prologue of C. Calaba and in view of the facts:—that
LINNAEUS adopted the American vernacular name as the
scientific epithet of the species; that the treatment given by
LINNAEUS in Gen. Pl. eds. v & vi included the American
element as a syntype of the genus Calophyllum; and that C.
Calaba was the only species of the genus as published in
1753 to include an American element, it is not surprising
that botanists in general should typify C. Calaba on the
American syntype and make this species as the lectotype of
the genus.

I have not been able to consult the treatment given to
the different syntypes of the species by WILLDENOW, who in
his Spec. Pl. (1800) 1160 suspected the Asiatic elements to
be different from the American one, and who is reported to
have later published C. apetalum Willd. (Mag: Ber. 1811,
p. 79) for an Asiatic element of C. Calaba. In the absence
of the original description of C. apetalum, it is impossible
to decide which of the Asiatic elements from C. Calaba was
included in C. apetalum; and the subsequent reports are
somewhat contradictory. However it is quite certain that
A. DE CANDOLLE (Prodr. I, 1824, pp. 562-563) reserved the
binomial C. Calaba exclusively for the Caribean element, and
published C. spurzum Chois. ex DC. to include C. apetalum
Willd. and at least the Malabar element of C. Calaba.

Gardens Bulletin, S.S.

259

Owing, however, to the contradictory typification of C.
apetalum and C. spurium, some having typified them on the ~
Malabar syntype and others on the Ceylon one, these two
binomials have been generally disregarded and later ones
are in use, namely, C. decipiens Wight (= C. Wightianum
Wall.) for the Malabar element, and C. Burmannii Wight
for the Ceylon element.

This exclusion of the Asiatic elements from C. Calaba
was so generally adopted that, at the Cambridge Botanical
Congress (1930), the British Botanists proposed that the
genus Calophyllum L. (1753) should be typified on C. Calaba
L. quoad the American element. According to a decision of
the 1935 Botanical Congress, Amsterdam, this proposal
should be followed unless there are grave reasons for
adopting another type; and so far no serious reasons have
been brought forward for disregarding the type proposed
by the British Botanists.

It is true that there have been a few opponents to this
general typification of C. Calaba. One of the first to select
a lectotype of C. Calaba from the Asiatic elements was
LAMARCK (Encycl. I, 1783, p. 553) who typified C. Calaba
on the Malabar element quoted as Tsjerou-ponna, RHEEDE,
Malab. 4, p. 81, t. 39, in the prologue of the species, inter-
preting at the same time C. inophyllum L. in a wide sense to
include not only the Ceylon and the American elements of
C. Calaba, but also a new one from Madagascar later named
as C. Tacamechaca Willd. (1811). Had it not been for the
fact that the genus Calophyllum has to be interpreted on
C. Calaba L. quoad the American element (included also
in LINN., Gen. Pl. eds. v & vi) on the recommendations by
the 1930-1935 Botanical Congresses, a very strong case could
have been made in favour of this Lamarckian typification of
the species; for RHEEDE’S plate of Tsjerou-ponna is quoted
directly in the prologue of C. Calaba as well as in two out of
the three additional references given in the prologue, while
there is an indirect reference to this plate also in the third
citation.

To my knowledge TRIMEN (Journ. Linn. Soc., Lond.,
XXIV, 1887, p. 143) was the first to assert that the Ceylon
element alone should be considered in typifying C. Calaba.
TRIMEN was then working on HERMANN’S herbarium in the
British Museum, London, and had typified, correctly as C.
Burmannii Wight, HERMANN’S specimen cited by LINNAEUS
under C. Calaba; but, apparently not realising that other
elements were also included in the prologue of C. Calaba,
TRIMEN remarked as follows :—

“The name C. Calaba has been generally abandoned for this

Eastern species to which it originally belongs, in consequence of
JACQUIN having figured in 1763 (Hist. Select. Stirp. Amer. t. 165)

Vol. XI. (1941).

260

as LINNAEUS’s species the Martinique plant, to which PLUMIER first
gave the generic name Calaba, taken from the Caribbee name.
LINNAEUS (Sp. Plant. ed. ii, p. 732) accepted JACQUIN’s determination,
and hence makes his own species to include both the E. and W.—
Indian plants. The name should not be maintained for either.”

This view of TRIMEN ignores the Malabar and the
American elements from the prologue of C. Calaba L. (1753)
and therefore cannot be accepted. Yet RENDLE and
FAWCETT (Fl. Jamaica V, 1926, p. 200) supported TRIMEN’S
views by creating the name C. Jacquini for the American
element of C. Calaba and by remarking as follows:—

“C. Calaba L. Sp. Pl. 514 (1753) is founded on a Ceylon plant
(Fl. Zeyl. no. 202) named later C. Burmannii Wight (IIL. i, 128,
1838) a different species from the West Indian plant described by
JACQUIN under the same name. In Sp. Pl. ed. ii, 732, LINNAEUS
added the reference to JACQUIN.”

The logical development to these erroneous assumptions
of TRIMEN and of RENDLE and FAWCETT concerning the
syntypes of C. Calaba L. (1753) culminated when ALSTON
in his Suppl. to TRIMEN, Fl. Ceyl. (1931, p. 22) adopted
C. Calaba L. as the correct name for the Ceylon plant with
C. Burmannii Wight as its synonym. In this treatment of
the Ceylon plant, ALSTON has been followed by VAN
OOTSTROOM (Blumea, Suppl. I, 1937, p. 196), and by
ABEYESANDERE and ROSAYRO (Descript. Check-List Ceyl. -
1959, p. 34), despite the fact that this treatment, based as
it is on erroneous assumptions, renders it impossible to
typify Calophyllum L. (1753-1754) on the American
element as recommended by the 1930 and the 1935 Botanical
Congresses.

In view of the foregoing considerations I submit that C.
Calaba L. should be accepted as the correct name for the
American element named as C. Jacquinii by RENDLE and
FAWCETT, and its use for any of the Asiatic elements
should be rejected as illegal. The nomenclatural treatment
proposed here is also in accordance with the procedure
which, in my opinion, should be followed in order to secure
stabilisation in the nomenclature of plants (FURTADO in
Gard. Bull. Straits Settl. IX, 1937, pp. 244-249 and 296-
299). Under more recent interpretations of the rule of
nomina ambigua (Art. 62) it would be possible to demand
the rejection of the binomial! C. Calaba L. as an impriorable
name on the grounds that its use in different senses has
become a source of confusion; but such a demand has been
argued by me to be contrary to the very fundamentals of
the Nomenclatural Rules (FuRTADO in Gard. Bull. Straits
Settl. XI, 1939, pp. 7 and 28).

Gardens Bulletin, S.S. Vol. XI. (1941).

PUBLICATIONS OF. THE BOTANIC GARDENS, cr eee

SINGAPORE

<4

1. The Agricultural Bulletin of the Malay Peninsula (oa ee

Series} nos. 1-9, 1891-1900: (out of print).

2. The Agricultural Bulletin of. the Straits sea FMS. - z,

[Second Series, monthly issues] Vols. 1-10, 1901—_

1911. ° Most numbers. are available, price 50 centes sae

each or $5:per volume.
3. The Gardens Bulletin, Straits Settlements.

Vol. 1 nos. 1-5, January to May 1912 (as Agitlede 3
tural Bulletin of the Straits and F.M:S.,Third—

Series). . nos, 6—12, December 1913-—Mareh”
1917 (as The Gardens Bulletin, S.S.).>

Vol. 2 nos. 1-12, July 1918-=August 1921. ie
Vol. 3 nos. 1-12, August 1923—March 1925:

Vol. 4 nos, 1-12, June 1926—January 1929, < =

Vol. 5 nos. 1-12, August 1929—June 1932. Pee 2

Vol. 6 nos. 1-15, (Gssued as parts 1-3) December. RSS)
1925—October 1930. Sane

Vol. 7 parts. 1-3, September 1932—June 1934.
Vol. 8 parts 1-4, October 1984—Getober 1935.
Vol. 9 parts 1-4, December 1935—March 1933: ~~
Vol. 10 parts, J end 2; January, 1939—August, 1939...

The above former issues of the Gardens Bulletin may
be purchased from the Botanic Gardens, Singapore, at $5 —
per volume or 50 cents per number for vols: 1-5 (Vol. 6; ee
15 numbers, $7.50); in some cases two or more ni
were published together. Vols. 7-10 aré:published at $8.

per volume; the prices of parts vary. according to thei size>~

The parts are published at irregular intervals, as material
is available.

Subscriptions per volume, payable “in advance, ae in-
Malaya, $8.50 or. £1 outside Malaya,

Vas year |
Mixa? Ae

eqeoippeneees
peur neater’
Snes net tet ittts|

cba. He » ee eoene od!

avagiey ut ‘ i +0
per ea! . ( {8 pEbe po} si
ne eter erat 4

: : , r

gage hm “yee:
Lege ert 4 f Jt fe
ager dpeveuet papepreeyrs teers:
: eke

ye
ee
eer
‘atte t

7 } ne t) fi 4 4 ow
ilegent : 4 4 ae eerie Pe! te

. god she, i pl i “2

ay & 4

es ‘ Mehrpeieendatet ed
itsetac rnin
ai ara ae,
pra eel +
Ses hry tps"
Al
ae
ryan bonbsveuried:
yee ag be

”

o”
te eee ret a
elage tte aaeede le" - rhe
ate painng 8et Ty aie 4
ot cyarerer yt? ¥ wont “ he

‘ : y a ener Pry yee Peres

' ines . ane Season Pe ies
: myers i mee md . " Pate Peet terol
‘ ; a IRENE " , - dg de ee beg
’ sar ht d
oa ry arr 4

ane
hae aoe ernest ae eet %
ange 4
ei ee pat hae Pe :
pan eee eee eh
hye

hth RETA
We ed need Tey te aden .

oa be ie tee eee 1