26 1
A STUDY OF THE COCONUT FLOWER AND
IrS RELATION TO FRUIT PRODUCTION .
Introdu cti o n.
The litera.ture on the biology of pollina.tion and fruit production of coconuts is extremely sparse and what little is available
is di stributed ill numin'ous books and peri odi cals so that investigators have experienced great difficulti e in consulting them. An
attempt is made here to bring together ail the av·ailable studies on
t he subject, including ·my own obser vations, a part 'oj' which have
already been publishcd in the P oana Ag1'iclllt1l1'al Co llege Magazine
and the Ag" iwltnnLI J07LnuLl of l nd'ia ('0) .
I am indebted to Ur. L H. Burkill for the information quoted
in this paper from the German and Dutch books ancl to Ur. F. N . •
Cha sen for the -identification of some of the insect visitors to coconut
flowers.
The In flores cen ce.
Coconut inflorescences are formed in t he axils of every leaf of a
beaTing tree and not of cvery third leaf as some wri ters have
su pposed (8) & ( IT ) . It is tru e that some axils fail t o t hrow ou t
any inflorescences, but this is beca use these inflorescences have become abor tive; and even then these aborted inflorescences do not
bear any defini te relation to the other which grow so as to justify
the statement th at .t he infl orescences are produced in the axil of
every thirdlenf. A vel')' prolific tree will produce twelve or more
inflorescences per annum or approximately one per month. There
are records of t rees hav ing produced sixteen inflorescences per year
&; (17 ) .
n
As the flower appear in the axils of leaves, it will be worth
while to know that the leaves are arranged on the stem in the form
of a spiral so that ever y sixth leaf opens, nearly abo:ve the first one,
that is, each leaf: opens according to the calculation of Sampson
in I ndia (21), at an angle of approximately 142° round t he circumferen ce of the tree from the previou leaf. In Goa the coconut
harve tel' divides the cocon ut trees un der two classes, the right- and
left-handed ones, accordin g as t he spirals formed by the leaves and,
t herefor'e, by the inflorescences, are ri ght or left. Both Costa (8)
and Loyola (n) who maintain tha t t he coconut inflorescences arise
in the ftx ils of ever.v third leaf give con ect diagrams of t he ph.vll otaxis of the coconut inflorescences. They were probably misled in
t heir observations on the orien tation of t he coconut inflorescence
by their mistaken vi ew that t he· coconut leaves are arranged }n
concentric circles and not in a spiraL
The in florescences fi rst appears enclosed in a thick, fib rous
sheath called the spathe whi ch is aga in protected during i~s eaTly
life by on e more ~'eJl ow sheath of somewh at flat n ature ancl of softer
fibres. This outer' heath stops growing yery earlY in the life of
. the inner spathe w that the latter pun ctu res it with its hard point on
262
jts ventral sid~ (i.e. the side toW'a.r ds the subtencling leaf) and
<!omes out erect as a yellow som ewhat flattened cone, which later on
,as it grows, tnrns green, curves a little outwards and becomes more
round than flat. In course of time when the spathe is of full
grown, the deyelopment and distension of the inflorescence within
,causes a great pressure on the walls' of the spathe with the result
that it ruptures longitudinally along a groove usually on its ventral
.side nnd the flowering bran ch eventually emerges : sometimes, how,ever, the rupturil)g of the spnthe takes place on its dorsal side but
t hen t he spadix turns round till the inflorescence within falls out.
'l'he process of splitting is very slow, the slit which a,p pears at first
a.t a point about an inch and half from the apex, takes about
twenty-four or more hours to reach down and give egress to the
inflorescence. It is at fir$t yellowish white in colour, but later on
• it turns greenish and also inclines downwards from its vertical
p~tioo.
.
The cocount is essentially a monoecious plant, that is, stamens
and pistils are produ ced in separate flowers on the same tree and
in the case of coconuts in the same inflorescence, and, though some
coconut palms will at times show a very ma.rked tendency to produce
spadices with all florets male, spadices -ate not usually produced in
coconuts where all flowers are female and none male. The tend ency
to produ ce completely male inflorescences is particularly apparent
when the palm produces the spathe for the first t ime in its life.
There is 'a great variation in such trees: some will produce their
second or third inflorescence with female flowers, while others will
bear no female flowers even in the sixth inflorescence. If this
variation is due to hereditary qualities, then this factor has also to be
taken into consider'a tion in selecting seed-nuts. The tendency to
produce completely male inflorescences is at times manifested by
trees which '8re given rest after a prolongecl period of tapping for
loddy.
The inflorescence itself consists of many flower-bearing ramifications or spikelets situated on a flesby peduncle : hence the inflorescence is termed a spad ix. Its size varies from two and half
t o six feet in length from tbe tip to the base, depencling upon the
vigour aJ1Cl individuality of the paJm. Each brap ch is fringed with
n umerous male florets from tip downwards and lower clown bears
one or more fema le flowers, all the flowers being sessile or subsessile
as Alclaba calls it. At times, however, some of these Tamifications
become spathuIate and then partially or wholly steril e. Some of
t he bran ches in an infiorescence may produce secon dary branches.
As far as my observations stands, such inflorescences rarely produce
female flowers which may be produced even on the secondary
branches. :Most of these inflorescences were noticed on t rees heavily
manured with nitrogenous manures, but I was not able to ascertain
whether the manure wa s responsible fo r the branching and reduced
fertility of these spadices.
263
Th e Male Flowe r.
The male flowers always exceed the number of female flowers in
the same spadix and may vary from a few hlindreds to thousands,
-d epending upon the nuplber of ramifications in the spadix and the
]ength of the flow er-bearing regions in them. Each male floret has
six yell ow perianth leaves arranged in two whorls, the inner three
.alternating with others whi ch are about one-third of the former in
size. Enclosed in th is floral envelope there are six hammer-shaperl
.stamen s which yield large quantities of powdery yellow pollen.
Aldaba (') has estimateel that each male flower carries about
272, 358, 50-1 pollen grains. In the centre of each male floret there
is a rudim entary pi stil which divides at its apex into three teeth,
-each bearing a gland, the nectar of whi ch attracts auk he;s and
other creatures. Rarely this rudimentary pistil is absent CO) . A
<case has been noted where these abortive ovaries were stimulated to
grow so that t he coconut palm bore, "instead of the few ordiJHl ry
i ruit at the base of the spadix, great numbers of small, crowded,
narrow, qui te banan a-like fru its." e') Th e male florets start
oOpeni ng from the tip downwards Rndliberate pollen, though sometimes a fe w flowers may open out of order. This opening of
lllale fl owers and shedd ing of the pollen 1ft ts about a month, the
inflorescences wi th secondary branches tahing a little more than
t he mu·nl period.
T he Fe ma le F lowe r.
The female flower are comparatively extremely few, their
llumber in each padix varying from zero to over 300 and being
·dependen t upon the strain , treatment, ctc. Th ey are always prodnced towa rds t he basal portion s of the spi kelets. Many of the
.yellow-llut-producing varieties from the Konkan (West Coast of
India), for instml cc, are usually very. hy in bearin g, while th e
1.1warfish varieties of Goa (the H enauly eed) or dwarf varieties of
:Malaya en joy the r eputation of being heavy producers.
Prior to it. openi ng, the female flower is a smail spherical body
·of about half nn in ch in diam eter with a great resemblance to a
small nut. Th ese female flowers consist of six flora l leaves whi ch
are arranged as in t he male flowers an d which completely envelope
the pisti l ; but these aTe mu ch larger and stiffer than those in the
male florets, and t he outer three are almost equal in size to the
inner ones. Apart f rom these six perianth members, there are two
1l10re just at the poin t of attacllment of the flower to the stalk,
resembling the others in colour anel texture, but differing f rom them
.in that they are broader amI . horter. These two may be termed
l)rophyll s or bracteoles.
sually ther e i a male flower on each
side of t he female resting on the.same cushion on which t he femal e
]s seated. The pistil is a nnall whiti sh body consistin g principally
01 embryoni c tissue of hu sk. F rom its tip t here extend downward s
three ridges which make t he whole ovary look globosely three-sided,
<cach ide being provided at its tip with a groove. Th ese three
grooves m eet in the centre ancl are the parts of the stigma. It will
264
be seen by cross-sectioning the pistil that just above the thalamus
there is the ovary (embryonic nut) with its three carpels, two of
which normally become abortive -even at this ea.r ly stage. Sometimes, however , all t hree ovules get developed and when fertilised
produce a trilocular-nut which on germination gives a tree commonly mistaken for a bran ched coconut pa}m.
Pollination and Fertilization.
Before proceeding any further a distinct ion between Pollination and F ertilization may be .made with advantage so as to avoid
confu sion on th is matter. Among the fr uit culturists, the t erm
pollination is often applied to designate all the iJlfiu€llces concerned
in the setting of fruits; and the term fertil ization is ofte n given the·
same significance. In botanical usage, however, pollination means
simply transferen ce of pollen grain s to the .s.t jgma, while fertilization is th e f usion of the male element f rom the pollen with the
female elemen t in t he ovu le, an d t herefore, conveys the idea that,
prior to this f usion, t he stigma bas to be poll inated and the male
element mu st reach the ovum through the pores in the stigma.
Cases, however, may occur where these two stages previous to the
feTtilization may take pla ce and stiillUlate the pistil to grow and
yet the actual f usIon may not eventuate proclucing seeds destitute
of any of embryo. The coconut fr uits known as "barren " "im..
perfect," " mal e," " man " or "seedless" Goconuts are probably a
result of such a phenomenon . S uch nuts have been observed by
me in India and Burma, have been r ecorded from Jamaica ('0) and
British Guinea (3), and pTobably occur in most places where coconuts aTe grown . They can usually be distinguishecl from others by
being narrower, and in sid e have , hell-substan ce and a cavity and
sometimes even a diminutive nut with or without .ome kern el,
but no embryo. Apparently the st' mulus of pollen is not even
necessary for such a development in coconuts, for Bailey «) has
written that "Coconu ts, like many otheT b'uits, often g row to a
considerable size without pollination, anel then perish." It must
not be ignored, however, that a fru it without an embryo may be the
co nseqnen ce of actual fer ti lization, but that it does not contain any
embryo, beca use the Elllbryo ceased to gl'OW soon after the f usio n
of male and female clemen t in the ovule without thereby arresting
the de.veloplllell t of the f ruit. A comparative study of: these abnormal coconuts and normal ones f rom J amaica maele by Kupfer
('0), indica ted t ha t the substance which usuall y goes to the for mation of seed was, in the case of seed less n uts, elevoteel to incr easing the bulk of hu sk. "Since no trace of: fungus, insect, or bacterial
activity coul d 1:e fo und, n o elirect evidence as to the cause of the
condition of the defective fruits could be proc1uced . The probability
is, however, tha t the r esponsibility for this state of the fr uiting
orO'Hl1S is to 1:e lai cl again st none or these agents, but lS t he r esult .
otthe failure on the par t of flower t o effect pollination."
Some coconut t rees produce " mal e" ll utS habitually in all
seasons of the year and others in certain seasons only of e~eTy year
265
()r after a. number of years; and still others do not produce male
nu ts a~ all or .produce onl:' thi s kind of fruit all tb eir life. Might
110t tIll S qualIty be associated wIth hereditary fa ctors ? In mmly
plants, for in stance, the production of sterile pollen grains and
ovules is due to an inhereJlt fa ctor, though influ en ces such as
·climate ma y modify these qualities to a slight extent. On the other
hand, there is the po sibility that the inability of ova being fert ilizedli eo in the f act that the nut rition is defective ; f or it has been
s hown by experimen ts tha.t weak and poorly nourished orchard
trees often produ ce ineffective pollen, or unfavourable weather conditiollS cau <e great losses by IlTeventing th e proper mat urity of the
JJolIen or pi stil. Aldaba (' ) ha s shown that desiccating influ ences
redu ce t he Yi tnli ty of pollen, whil e it is a matter of common belief
:a mong pl an ters tha t heavy clOWllpOurS of rain s excess ive cold, heat
.and wind s, or prolonged drought interfere wi th setting of nuts.
\Yh en it 1egan to l'e reali ed by coconut plantcrs that ordinarily
fertilization was neccss:ll'Y to produce coconut f ruits, it wa s custom.ary to ·on s' der all t he fe mal e flowers, even when they were not ripe
to receive pollen, a, fe rtilized or at least pollinated flowers. Thu s
in 1898 th e late J. M. de Sa ('0), then a District Admin istrator of
th e Yillage Associa tion s of Goa , wrote in hi s book" 0 Coqueiro"
that he h ad seen pollinated or fertilized flowers even in Ullopaned
spath es : an d t he same id ea wa s repeated about 14 years later hy
L. C. Brown (' ) , la te IllRpector of Coconut Estates in the
F. M. ,'. in a communi cation made to Mr. R . H. Smith, th e
senior author of the ('0 118018 of til e East. On thi s view Freclholm
(0 ), ,rho, as fa r as 1 am able to make ou t, was the fir st man to
correct the Yi ew, remark ed thu s : -" Bu t when )"OU state t hat
llollinati on takes r 1a r-e to a certaiu exten t before the actual
full openin g of the sheath, so that yOUJl g imit, whi ch mayor may
not properly set, i ob 'erved haH-formed as soon as the flower-spike
'comes into view, then you are wrong. You have evidently mistaken
the female fl o"\\'er bud for th e young fru it (the ferti]j zed ovum ) .
In thi s plaJJt in-breeding is so exceptionall y well gu arded against
that it is well-nigh impossible, the pollen grains and the ovules of
"on e and th e same infl orescen ce never arriving at maturity simult an eously. On thi s point 1 wri te, in my articles, as follow :-InlJreedin g or close-breeding is guarded aga in t as mu ch as possible
in na ture. It is prevented, in the case of the coconut palm, by fI
-d ifferen ce in the t ime of expansion of the male and female flo,vers
'on the same spadix, and as a palm rarely ha s more than one inflorescence at a t im e with open flowers, the pollination of the female
flower s is generally brought about by ]lollen from the staminate
flower s of another palm . Thu s cross-pollination is the rule. The
pistillate flower s do not expand before th e staminates of the same
'spadix have shed th eir pollen and fallen out.
nt il that time
the gynnecium r emain s complet ely covered by the perianth leaves."
Ohenation s mad e by YaI'iou s other investigator such as
P eteh (18 ) in C ~'lon, Sampson (21) in India, and Jepson (13) in
:Fiji, confirmed those of Fredholm's . . But though tbe conclusions
266
of Fredholm appear to hold good .i n most of the countries wherecoconuts are grown, t hey are not uni versally true. Not only is:
there the possibility of exceptions occurring in places where normally the female flowers open when all of male flolirers have been
shed ('0), but iu certain places as in the warm, humid lowlands of
Malaya « th e female phase not only begin s, but mOst frequently'
ends before, or at the same t ime as, the male phase, thus rendering'
self-pollination the rule instead of being an occasional chance occurren ce (-" ) ." Messrs. Jack and Sands from whom the above
quotation is taken have sncceeded in obtaining fruits by bagging'
an unopened COCOllut inflorescen ce, and my observations on the
dwarf variet ies in the Botanic Gardens, Singapore, lend support to,
their conclusions. Th is behaviour of th e palms in the F . M. S. may
be due t o some hereditary qualities ; but is also possible to attribute·
this variation in t he anthesis of coconut flowers to climatic con-·
elitions and th e fact that Van der Walk (24 ) has shown that the·
ripening of female flowers is hasten ed by covering them with black
paper so as t o redu ce li ght and warmth, may be invoked in support
of this
The mode in whi ch the coconut fl ower presents its stigma for
the reception of the pollen is quite different from that of most of
our gard en fl owers. I n the latter the petals become loose as the·
stigma matures, open out exposing the stigma and then after a·
time with er and drop off. S uch however, is not th e case with t he
members of the perianth of the coconut flower . They n ever dro p off .
unless the ovary or the f rui t itself is detached and when young t hey'
form a yery ti ght case in which the pistil is protected. The perianth
leaves grow extremely slowly attaining th e fi nal length of about
two to three centimetres, wherea s the pist il inside grows com-·
paratively at an enorm ous rate so that it forces them apart and
extrudes the stigma-bearing region, on the ripening of which thestigma tic grooves become exposed to receive pollen. There is a'
secret.ion of nectar both f rom these grooves as well as from the
r egion snlTound ing them. The period during whi ch th e female
flower remain s receptive varies in differen t places. At Ab.J 'ab (,0)
and at P erad eniya ('"), for instan ce, it is about 24 hours, while in
Los Banos (' ) and Sin gapore it is about 2 to 3 days.
Biology of Pollen.
POllell grain s of the coconut are sph erical und sinooth, without·
any asperi ties, but all exposure of a few Eeconds they turn ellipsoidal
" 'ii h a single meridian groove or suture which, accoreling to K erner ·
(H), is chara cteri stic of palms. On wetting the grains resume their"
original shape, the longitudinal fold cl isa.ppearillg. This groove·
seems to point out that coconut pollen belongs to the type adapted
to be tran srorted hy insects rather than to the type ensily wafteel"
by winds. Th ere are two kind s of pollen grain s, fer tile and infertile, the latteT are about half the size of the former. Aldaba's,
(') countings show that infertile [-allen grains vary in the'
2G7
Philippines from 3 to 33 pel' cent, but in Singapore t he abortive
grains appear to be very-f€w.
It is the general belief t hat pollen grains of palms, when kept
in dry condition, reta in their ferti li zing properties unimpaired for
a very long time so that they can be exported to distant countries
for the purpose of pollinating certa in varieties which are desired
to be crossed, and Kerner (H) quotes a tradition whi ch says that
the pollen of Da te-palms together with that of H emp and Maize,
can be used effectively for artificial pollination even after a lapse of
eighteen ~'eH.rs. H ell ce I made pollen culture in cane sugar solutions to ascertain whether there was any possibility of female flowers
being ferti li zed in nature with the pollen f rom the same inflore cence. At the time I started my studies I had come across only a
few exceptions at Akya b where female flowers ripened before the
male fl owers in the same inflorescence had finished shedding their
pollen, but had not seen the paper by Alc1aba on the subj ect, nor did
I 1."110W of the studies made by J ack and Sands which show that in
Malaya self-pollin ation is the rule rather than the exception. Hence
I dupli cated many of the results obtained by others. In these
studies I obtained th e best res ults 'llrith 2<0 % cane-sugar solution,
and it was found that every clay more and more pollen grains lost
t hei r vitality when kept under ordinary conditions till on the
seventh day only 3'0/< showed uny germination and after that pc riocl
no graius were seen germin atin g. Aldaba's (') find in gs show t hat
in Los Banos, 25 to 3,0 % are the best ca ne-sugar solutions for
effecting germination of pollen grains of the coconut, that pollen.
remains viable for b,'o to nine days, and that pollen grains from
different trees do not maintain their viability for the same length
of time.
Now we have seen above that in many places the staminate
flowers fa ll off before the stigmas of the female flowers in the same
inflorescence become receptive and it is usual among the planters to
argue that where thi s OCCUl"S in-breeding or fertilization by the
pollen from the same inflorescence is impossible and that emasculation of the inflorescence of which female flowers only are to be
used is Ullllecessary. That thi s way of arguing is falla cious is
shown by the above results which show that pollen may, under
ordinary conditions, retain its vitality even for nine days. Added
to this there is the danger of female flowers ripening earlier than
usual anc1, thereby, of their getting self-pollinated, thanks to the·
reduction of light and warmth caused by tIle bags U SE 1 to protect
the fl owers f rom foreign pollen; for in t he above referred experiments with black paper Yan del' Wolk (2') was able to secure
self-pollinated nuts from trees where under ordinary conditions selfpollination was impossible.
.
Wh en, instead of being kept exposed to ordinary atmospheric
conditions, the pollen grains were preserved in celluloid capsule
such as are used in administering quinine powder to pat.ients, and
the capsules were coatecl with melted tallow, a greater percentage
of pollen grains were fo und to rema in viable. Sampson (2'}
268
W1'i es in his Coconut Palm that coconut pollen can be preser ved
for several days in hermetically sealed tubes without losiug its
vitality. But f urther investigations in this matter are needed so
that a system of artificial pollination may be evolved which ,,·ill
insure the rapid improvement of so impor tant a crop as t he coconut.
It should also be such as t o render it easy for planters to know
J10t only the materni ty but also the patern ity of t he seeds chosen.
U If pollen grains are wetted," writes Sampson (2'), "they at on ce
assume a rounded shape and commence to disintegrate within the
space of two or three hours. It is thu s eviden t that, in t he moist
t ropical clim ate whi ch favours the growt h of the coconut palm,
there is no chance of the shed pollen grains remaining dormant
till the female flowers are open and r eceptive." We have seen
that when exposed to ordinary condition s in a laboratory in
Singapore pollen r emain e(l viable for even seven days. But to te~t
JlOW long the vitali ty of pollen may r emain when it· is exposed
to an atmosphere sahuated with moisture, some was dusted on
to a slide kept on a cell and put into a closed petri di h partly
fi lled with water. It was found that had after 61 hours exposure
to ,meh saturated atmosphere the pollen had not lost its vitali ty,
but after 12 hours exposure to such conditions more thau 75%
<>i the ~raills had lost their vitality.
Po llin at ing Agen t s.
In most countri es, as has been explained above, ther~ is very
little chance of the fcmale flowers bein g fertili zed by the pollen
from the ~p. me inflore cence. This means, ther efore, that they have
to depen d for pollen upon other 'inflorescences, either from the same
t ree or from others. The chan ces of obtaining pollen from an inflorescence on the same t ree are very much reduced by the fact that
it is on ly occasionally that a f resh inflorescence opens before the
previous one ha finished flowering, and th is in spite of the vigour
a nd prolificnes of ome trees. This mean s that a large number of
fema le flowers have to depend for their pollen on other trees. This
explains why in most countries there is so much variation in the
seecUinO"s raised from the nuts of the same tree or even f rom the
sa me inflorescence, when eedlings raised from dwarf cocpnuts of
Malaya where cross-pollination is an occasional chance occurrence
behave so like their parent palms (12) .
Since the stamens and pistils are botne in separate flowers, the
pollination in nature can only take place with the pollen brought by
winds, or by iJ1Sects and other creatures that are attractecl to them
beca use of their peculiar cent, colour, nectar etc. Knuth e') remarks that the coconu t is pollinated through t he agenc), of wind,
but quotes Fr. Dahl who noticed the birds ClumnosYllct sltbpZctcens
Scl.,Cinny?·is f?·ena/ct S. Mull. an d C. co,,.·innn Salvad. as the freque;lt
vi itors of coconut flowers in the Bismarck Archipelago. According
-. to Petch (' S), pollination is effected chi efl y by bees and hornets in
Ceylon, though from the stru cture of the flower, he admits that the
wind may be aIso responsible for the tran sterence to a grea t
extent. Hunger states that the coconut is pollinated by wind as
well as by insects, and among these figure wasp (w8spen) , bee
(begin ), fly (v liegen), beetle (kevers), and ant (11t'ie'l"en) . Aldaba
(') working in the Philippines fo und so little pollen carried by wind
from one tree to another that he attaches very li ttle importance to
cross-pollination by this agent. The principal insects observed by
him as probable pollinating agents are the house fly (Musca
d01nestiw Linn. ) several species of lAwilia (Di ptera) , 11 espa
luc/mosa Sauss" S(l1'coplt agn sp, Bhynchiu11t (tt1'U1n Sauss., l1pis
indic{t, T'l"igon{t bVl'Oi (Hymenoptera) and P,..ionece'I"'us we1"ltleipennis Perty (Coleoptera) . Samp son iTom the peculiar structure of the flower and the honey glands infers that nature has
intended that the coconut flowers shoulcl be fertilised by the aid
if insects, Burkill (ti) ha s noted Ap'is dOl"sni{t and A . indic{t on
coconuts in Singapore, but remarks t hat this genus is often found
in the Malay Peninsu la on "palms ,overwhelmingly "on male
flowers, or on flow ers in thcir male stage, obtainillg food without
giving what would seem to be an adequate return" and that only
Apis ind'icn ha s been seen behaving in that manner ill Siugapore.
The observations of J ack 'a nd Sands ("J, on thc pollination of
coconu ts i~l the Mal ay Peninsula are of uniqu e interest, "In three
unopened inflorescences which were baggcd ill muslin bags, selfpollination was effected naturally ancl fruits were formed, while in
three other inflorcscences which were emasculated immediately on
opening, no pollination took JJlace ancl no fru its wcre formed,
though ' the female flowers beha ved normally and although ma.]e
flowers on adjacent trees were in full bloom. In a similar connection, it has been observed that odd isolated coconut trees growing
even under bad conditions (produce fruit so that self-pollination
mu st take place. When coconu t !towers are in full bloom, at auout
1:0 a.m" when the dew has chied up and when the gentle breezes
f requ ently begin, clouds of pollen can be seen floating away in
sunlight. In a very slight breeze these pollen cloud s do not travel
far owing to the weight of t he pollen but it is highly probable that
with the streugthening of the breeze as the day advances the pollen
clOllds are ca1'1'iecl to considerable clistance and thus cross-pollinatioll is effected." From this it would appear t hat insects play an
unimportant part in t he pollination or rather cross-pollination of
coconut flowers in the Malay Peninsula. However, as said above,
in most countries where coconuts are grown t he coconut flow ers
behave difl'erently, Aldaba's (' ) results support t he view held
by 'm any planters outside the Peninsula that an isolated tree tloes
not ' berur fruit if male flowers in the succeeding cluster do not shed
pollen during the period when the stigmas of t he female fioweTS
below are yet in a receptive condition, and that a tree in a grove
Imilcr -the sa me condition s bears fruit. J epson (13) who paid a
special attention to the insects beneficial or otherwise to COCOllutS,
after saying that pollination OT coconnts in Fiji is dependent on
wind and insects, among which he noticecl bees and some black
hymenoptera, attributes the c]ropping in many districts of Fiji of
female flowers in large numbers, resulting in poor yield, to the
270
gl'eat scarcity of insect life in the vicinity of an open inflorescence.
H e cori'oborates his view by the observat ion that, on estates where
bees are present in large numbers owing t o artificial rearing or
otherwi se, t he yielcl of nuts is very remarkable high. On these
grounds he advises the planters in Fiji to in troduce bees on thair
cocon ut estates with t he view of increasing their crops.
My' own studies on t his subj ect have not been veTY extensive,
but they t hrow some fur ther light on the various points raised uy
the previous investigatora. Hegarding t he ant as pollintor .Petch
(18i) writes : "In considerin g the l)otentia.l insect vistors to
flowers in the Tropics one has always to t ake in to consideration t he
ubiquitous ant. At first sight it seems possible that this insect
may take part in the con veyance of pollen from male to t he female
flower, especially when the periods of t he inflorescences overlap.
In t hat case they migh t COil vey 1)01len from one inflo~'escence t o
another on the same tree. But it is improbable that they should
convey pollen f rom one tree to another, because the journeys .of
this species, as a rille, do not ex tend to two trees.
"There is, 'however, a special provi&ion on the female flower
of the coconut whi ch more or less effectually excludes ants from
t he work of pollination. The region below the stigma, almost the
whole of the area which is exposed when the female flower opens,
bears a large number of pores. When the flower is ripe these exude
a qu antity of moi sture w'hich, at least in fine weather, forms a ring
of liquid round the stigma and prevents the ants reaching the latter.
It is not uncommon to see a crowd of black ants congregated round
the edge 0:£ t his ring. It is proba.ble that, as is usual in cases of
this kin d, the liquid contains some sugar, so that the ants obtain
what they want without robbing the stigma. In ·a ny case, it keeps
the ants away from the stigma. The position of these water pores
can be clearly seen on the young fruit where they are indicated by
small whiti sh spots. These spots 0we their colour to masses of
minute crystal whi ch are deposited by the liquid."
In the Botanic Gardens, Singapore, ;r have not found the
secretion is in sufficient quantities as to exclude the ant from
the stigmas. Hunger ha s included ants among the pollinators of
coconut flow ers, and Aldaba admits the possibility of pollination
by ants . when 11e presumes the conveyance of pollen by ants as the
prob!Lble reason for the development of a nut in an emasculated
bagged inflorescence. However it may be t hat the climate of
Peradeniya is favourable for the accumulation of the liquid in
such large quant ities. In places, therefore, li]i:e Singapore, where
secretion is not in sufficient quantities as to exclude the ant from
the work of pollination, it will play an important part particillarly
with coconuts where male and female flowers ripen together so as
to insure self-pollination.
The insects that seem to do yeoman service in pollination or
cross-pollination of coconuts in Singapore are some species of
M elipona (the dumma1' bees ), Apis dorsata, and some Muscidae
271
principally M~,sca very near nebulo (the common, Oriental housefly), Lucilia sp., and Pycnosorna sp. These were seen visiting freely
both the male and female flowers, though the flies seem to engage
themselves more in sipping the honey from the female flowers
than in feeding on the pollen or honey in the male flowers. Apis
indica was a rare visitor to the male flowers, but this bee, according
to Burkill, does not give an adequate return for the food it obtains
from the flowers of palms. Wasps do not appear so useful as the
bees in pollin ation as they visit the flowers mostly for the purpose
of preying on the bees and other i.nsects which are usually beneficial
to t he pollin ation of palms. VespCt cinctCt was frequently observed
hunting insects and only on one occasion it was observed
alighting on male flowers. Cy,·tostorltus pectomlis liors., was seen
but once sipping the honey from the female flower s of a coconut
and according to Mr. Chasen of" the Baffles Museum, Singa.pore,
Anth"eptes m.Ct/a.ccensis is the sun bird which is almost invariably
associated with cocouuts in t he Malay Peninsula. Variou s other
bircls were also seen in the I'iciuity ·of coconnt inflorescences, but
it must be borne in mind that even those hirds which possess especial
adaptations in their beak for extracting nectar from various kinds
of fl owers, often visit the flowers for the plll'pose of capturing
insects for thei r prey and hence the utility of birds in places where
there are insects such as bees to pollinate the flowers, is, like that
of wasps, doubtful.
My emascul ation experiments on dwarf varieties gave results
wh ich differ from those obtained by Jack and Sands in the F. M. S.
in that I have been able to obtain nuts even though the nearest
tree from which pollen could be brought was about 50 yards away
f rom t he t r ee, and t hough the infl orescences looked sickly and the
nectar secretion was reducecl because of the injury resulting from
the emasculatiOll. Further I have seen nuts developing on those
few iJlflorescences where female flowers became receptive only after
th'e falling of t he male flowers. It makes me wonder therefore,
whether the failure in F . 1\1. S. to obtain fruit from emasculated inflorescen ces was not due to the injury resulting from the emasculation or to the absence or scarcity of insecf life useful in the pollination of coconut flowers. By keeping unpollin ated emasculated inflorescences side by side 'with the emasculated but artificially
pollina.ted ones it WOl1ld be possible t o throw considerable light on
this point.
Influence of Manures.
It is a known fact in horticulture that application of too much
nitrogenous manUTes causes the plants to vegetate at the expense
oi flowers and fru its, while lime and ph osphates tend to divert
this energy in the opposite direction, namely to the production of
flowers and fruits. Does the same thing occur in coconuts? What
manures exert beneficial influence on the flower production in
coconuts nud wha.t others act detrimentally? Unfortunatelly coco)1,* have r eceived very little attention in this lin e from the in~
•
27 2
vestigators and ' hen ce our kn owledge is ·at present very limited.
We only kn ow that t illage and manures increase the yield and that
coconuts require more potash, and probably salt also, than many
ot her fru it crops. However, it mll st be understood that to study
the influ ence of ma11l11'es on flower production is n ot an easy matter
i n the case of coconu ts especially becau se the immedi ate effect of
mnn ures on estate i. often to i ncrease t he vegetative growth and
t o Tedu ce t he yield of crops. In making such experiments due
consideration ha s to be given to the fact that t he yield in coconuts
can be in creased in a number of ways, the following being the
pT i ncipa l ones :1. By the in Cl:ease in the number of female fl owers in each
spadix ;
2. By cau sing to grow the spad ices that otherwi se would have
been dornia n t or aboTti ve ;
3. By t he m pi d production of leaves and inflorescences, due
to inerease ill t he number of. leaves produced during a given time.
4. By increasing the abi lity of the ovary to be fer tilized even
under ad verse cond itions or in creasing the vigoUl' of pollen so that
it may be effective ill ferti lizing t he ovum.
5. By l'educin g the fall of immatul'e nuts due to maltnutl'i bion of the plant .
All such poin ts have to be considel'ed in a study of the effect
of manures befol'e anything can be definitely said about theil'
influence on the prod uction of flowel's. Copeland (') has shown
t ha t in the P hilipp.ines t he leaves take one and a half years from
t heil' fil'st appearan ce until th eir fu ll development, and th at another
one and hal f: years aTe required for fTuits to mature in their axils.
Th is peri od, no doub t, will vary accol'cbn g t o the local conditions;
but, at any rate it gives us a~ idea how long th e effects of a tl'eat men t wi ll last considering that t he pl'esent t reatment will, to a great
extent, determin e t he nat.ure of t he emhl'yonic leaves and infloTe 'cences and, ther€iiore, t he j'uture crop. Let 115 assum e b'y wa.y
of illustrat i'on, th at we ha.ve st arted ·cultivating a ver.v n eglected
cocon ut tree. The first effect of bhis treatment may be that the
suppres ion of many inflor escences an d a considerable improvement
in the general aspect of the tree. The next symptom of thi s treatmen t may be the redll cti'on in droppin g of immature nu ts, lat er on
the anIlual J1Um bel' 01 leaves produced may increase and with them
the number of infl orescences, then the number of abortive spadices
may gradually be lessened an cllastly it may occm that t he inflorescen ces produ ced may be longer, beaTin g a greater number of female
flowers. Many of these flower s 'may a.t first chop because the palms
are too weak to prod uce a sufficient quantity of good pollen. The
vario lls changes may not t ake place in t he order mentioned above;
bu t the illustration will, 1 think, show the n ecessity of makin g. •
ve rI' careful record s for a number of year s and herein lies the
chi ef clifficulty of experiments. Jud ging, however, from indirect
eyi Cle]J ce it call be said 'tha.t good tTeat ment does increase th ~
273
!l llm ber 01' i nfl orescences on a tree and of: femal e fl owers in each
i!l floJ'Pscence . :M an." of the villager s' holdin gs in c:o,~ a re sit uated
at the foot of later i te hills, fa r away fr om any human habitat ions and
th ey do !lot. usually r eceive any attention fr om t heir owuer s. The
cocon u ts t rees in s lIch hold ings have, as a rule, ma.ny spatlices
a borted, t hose t hat ar e p rodu ced have a stun ted g rowt h, ea·ch bearing
ordinariII' not m ore th a,n :tOlli' 0 1' si x female fl owel'S; while 0'11
hold ings which a rc und er a more caref ul cultiva tor or neal' a well
or r'a t tie h\'l'e t he tr ees produ ce mor e pathes and mOre fem ale
lI ower s in ear·h s pad ix. SmnJ.lson (" ) is also 0:1' opinion t hat
ma llur illg does in d uce to all apl" 'cciable extent, the ["'pid developmen t 01' li e\\' bU ll ches of {lower s, drawing a basis for his asser t iOll
'l'l'Om t he better behu, iour of coconu t tr ees neal' a dwelling t han
f'IHlt of t hose g rowing at a cii stan ce ,'r om wher e plan t food j ,
plel1ti I'ul in t he soil.
C. X.
FUR'I.'ADO.
Lite rature Quoted.
1.
AJdaM, V. C.
rf he P ollin ation of Coconut- Philip. A.g·l'icul-
t !l1'ist, X. H)21, pp.
2.
:~ .
,I.
.i .
(i .
'I.
R.
n.
1 9~ -2Q17 .
A nonymou . A I'{e P alm rl1'ica. Th is booklet on coconut cult ivation \YR ,~ eil'cu /.atecl in tJhe 111anuscr i pt by the olcl J esnits
from Goa. In 1814 loJlg- after the exp uls ion 01' t he J esui t;;
'l'rom t he Port ug uese tel'l'i to l'i es, th e Goa Govern men t
P ress pr interl 'i t rOT t he prlbJic. Sin ce th er e has I,een
many ed it ion s OJ' imp ressionR. T qnote 'l'rom B. F. (Ia
C.osta' ('rl iti on of 1872.
A nonV1110U;;. :\1" Ie or Ser(ll c"" Cocon uts-A .'/1'7. .\ ' MUS, 'Barbados, VT, ] 907, p. 87.
B~,il e~' . L . H.
S/a.l7da.?'f1 r :lJr./oped.ia of HO?·ticultu1·e, II, 1917,
p. 813 .
B r owll , L. C. Quote d bv Smit h and P ape. J1 p. 590-59 1.
Hurkill, T. H. Some .'\ otrs on t he Pollination or Flower in
t he MalHI' P enin sula. ()U l'rle1l s' 13nll p{ ;1I , TT, 1919, ]Jp. 165176.
Copeland , K H. Phys iolo,gy 0'1' t he C'ocon nt-P7IIili p. A.'/ I' im/.lItm·is/ {f ll rl P OI'I'S {PI' , vol. T, 1911, pp. 44-50.
Costa, H. F. rI a. J{(lIlwil I'm{·i co do A.'/ I·iw lto'l' Tnd';n?1o,
Vol. I, L isboa. 1872.
Fredholm , A. Qu oted bv Sm ith eml Pape, pp. 59 1-1)93.
10.
F lll'tac1o. C. X. 'Phe Cocon ut Tnfl orescem:es-- Poonn 11..'1"7.
r oll. Mag. , XTI'. 1923 pp. ~ 1 3 -22 1 , a nd Ripening 01'
COC0111l t Fl owers . 1.'/ 1'7 . ,J onl'//'. i ndia, XVTTT. 192il,.,u.. 56 1.
1 1.
12.
B unger, T. W. F. Co('os 'llt/cifel'(/', Amste rc1am, 1920.
J ack, H . W. & I\' . .'\ . Sand s- T he DwaTf Coconu t in Ma laya Jra,layam A,ql'l. .'0111'11. X . 1923, pp. 4-1 2.
274
13 .
l~· .
I :' .
10.
17.
18.
In.
2,n.
2 1.
~..~.
~1.
2., .
.J epson, F. P. Depart. ngric, l<' ij i, Pall/pillet r,'o. ](j .
p. 3, Suva 1915.
'Kern er I"on Maurilaul1, A- The Na.lm·uZ History of P7,MltS.
Rnglish Eel . Yol. n , 1.onelol1 1895. pp. 9(i & 99.
I,"uth, B. fla11 dbllch de'r lilldcnbiolo.r;ie. BIl. III, Lei p'l,ig.
19M, pp. 59 and ') 8 .
1\1IJlI'er , J£lsie M. , Imperfec-t Cor·onl1t ~-./ o'l/./"Ii . .\" e'/l) J"01·7t; Hal .
(:(trf/. .. 11 , 1901 , PI'. 69-7 1.
I,oyola, J. I. rl e. ('jili 'ILI'as I lIi{;ia'lI(Js, Orlim , Cioa, 1806.
I'etch, 'I'. 'l'he Flowe r 0'1" ihe Cocon ut - Trop . .'I.r;rirll/.I1I.1·i.,/ .
XU , 19 1B. pp. ~4!) - -1·5,.5 .
Ilnngachari. K.- JI .l;a:n1l1J7 of NlclI1en/al'l/ lio/II/I,I/ /01' Iidin.
2 nd ErlitOo Marlra s. p. ~· 1 5.
~a , .'1. ~<f. de . 0 ('ol/1U'iro, :\ov>I-Uoa . 1 ~98 .
Sampson . 1-1. COo The ('OCOIl1!/ 1'01111. I,ondon, '192:, .
Smith. If. fl. & F. A. fl . 1II1'e, .('O/,O/llUS: fhe ('OIiSO/.~ of II, I'
ga.,I, 2nd E(l . London, 191,3 . .
\'an (ler I\·olrl . Quoted h~' I-hlll ger , PI'. ?,:L
\\'orsd ell , \\'. COo {'rili cip/es
{'lal1/-T I'I'II/olo(J!J. \'01. I I,
Londol1 191 G, pp. 191-102.
at
BRANCHED COCONUT PALMS AND THEIR
FERTILITY .
Apropos of the note on the Fel'till:tY, of Bmnchcri ('oco'lltil
Pltllns by Mr. Burkill, published on page 1-2 of Vol. III of t hi s
Rltllelin, the following may not he without in tere. t to reaelers.
Xorma ll), one cocon ut fruit gives ri se to . one shoot an el this
in its turn to one stem. It is not infr equently, howevel', thnt one
meets with wiele deviations from this normal p henomen on. A nu t,
for in tan ce, may on germination, give ri se to more than one shoot,
each arising from a separate carpel in th e nu t. Th e writer has 110t
~onle across an in stance wher e a nut had more tlian two fer tile
cHI'pel s: but it mu st be r emembered that t he coconut has a triloPlllar ovary whereill normally two of the locules become abortive.
('a ses, the;·efore. ma.'· occur ;vhere all t hree ca rpels ma.,I' be stimulated i'o become fertile a.' in som e spec·ieR Tela1'ed to Cocos 1Mwifem.
POI'heR reports o·f. "a nut w ith t hree cell s separated hI' leathery
"',11k" 8aml'son (I") appears to ha ve seen cases where t he septa
separntill g the ovules were hard amI not lea.thery. 'POl·bes. in his
ll1'tic-le above referred t o write : " 1 ha ve seen al 0 nuts with cell~
l'angilll( from 1'0111' to eight f;lld ten . r ~end VOll a rough ouLline
sketrh 01' a tree whir-h ha s eome up from a nut of fOlll'teell celk
all of \\'h ieh germinating. prorlu('in,!); a h'pe with fomtecn stems
•