senake bandaranayake

CHAPrERONE

Senake Bandaranayake

Sri Lankais perhaps theonly country inSouth Asia where we still have substan-tial archaeological remains of fonnally-laid out royal and monastic gardensdating from a period before A.D. 1000.They belong to a tradition of gardenarchitecture and planning that is well-documented from the late I st milleniumB. C. onwards.Literary referencesThe royal and monastic gardens of theEarly and Middle Historical Period (3rdcentury B.C. to 13th century A.D.) arereferred to in the B uddhist chronicles ofSri Lanka from as early as the 3rdcentury B.C. The chronicles them-selves, of course, were written betweenthe 3rd and Sth century A.D. from ear-lier written and oral sources. Whateverthe actual history of Sri Lankan garden-ing may be, the Sri Lankan Buddhistsinhe.rited and developed two conceptsof the early Indian tradition, which havea direct bearing on the art of site selectionand landscaping. The first of these isthe concept of the urban or suburbanpark or grove in which kings, noblesand merchants, as weIl as ascetics andreligious teachers, took refuge from theheat and dust of towns and cities.lndianBuddhist literature bas many referencesto such parks and gardens. The otherconcept was the distant forest grove ormountain or cave retreat, which ascet-ics and sages frequented in order to

concentrate on religious or philosophi-cal pursuits.

The Sri Lankan chronicles echoBuddhist canonicalliterature in refer-ring to royal and suburban parks andwoods donated by the flrst Buddhistkings as sites for the early monasterles{Mahavamsa XV, 1-25). This is con-firmed by the archaeological evidencewhich shows the city of Anùradhapuraringed by well-planned monastic com- ~plexes in which parkland, trees andwater clearly played an important role(Silva, 1972; Bandaranayake, 1974: 33

fr.).The alternative monastery type to

the park orgrovemonastery (or 'ararna ')was what has been called the 'girl' ormountain monastery (Basnayake, 1983 ;see also Bandaranayake, 1974: 33,46,47). Here, a rocky mountain peak orslope was selected and caves or rockshelters fashioned from the sides ofmassive boulders. Hundreds of suchsites with inscriptions dating from aboutthe 3rd century B.C. to the lst centuryA.D., including Sigiriya itself, appearin the Sri Lankan archaeological record.

A third concept of water festivalsand water sportsfrom both popular androyal contexts, and probably pre-Bud-dhist in origin, are also featured in theearly texts. ln the Later Historical Pe-riod from the 13th century onwards, thedescription of royal water sports, often

:v 3

''l '

~

;:~f

, ,l '

\ 11i !

, 1" Ii' ",

Fig.l

Fig.2

Fig.3

Fig.4

~4

~

9Pu

I ~-.=-.::[Ii

~ I; }10.

V'.:::::"

G':;~ ~

b.\-~'

~> J"-, .'"

~~

~~..,~

J;tJ" ~

~-., ' "-'-'

..~

,..;J

"-/

~

""'"'1(\

j

-..,~ -oou..~ .~..u .-~~'.NN. " -.,- .-T--'-

.~H~"

-, .-

()oc

~

r0--

.:i~J 1!ii!1r,ir11;"!,c~~",~cf

!;:, ~{j;f'i"

- +.

ç

Fig. 5 'Boulder Gar-

den' monastery,

Maligatanna.Date uncertain.

Fig. 6 Early Buddhist

rock-shelter

monastery,Dambulla, 3rd-1 st century B. C.

o+~

.

w", \1\ vr:~ 5


set ting and surround for the monasticbuildings. One of the most striking ex-amples of this cao be seen in the south-west quadrant of the Abhayagiriviharaat Anuradhapura (see Bandaranayake1974: 55). Secular gardens in this sec-ond cate'gory and from aslightly laterperiod than the Sigiriya, Anuradhapuraand Mihintale examples, are the ves-tigial gardens of the royal precincts atPolonnaruva, which include two strik-ing and well-preserved water-palaces,and the uninvestigated urban and pal-ace complex around the excavated andconserved royal bath at Galabadda inRuhuna.

The first scholar to draw our atten-tion to site selection, landscaping andgarden forms in Sri Lanka was SenaratParanavitana, one of the founding fa-thers of Sri Lankan archaeology and thediscoverer and first systematic excava-tor of the Water Gardensat Sigiriya. lnamonograph on Polonnaruv~ publishedin bis retirement in 1954, to mark the8th centenary of the consecration ofKing Parakramabahu I (A. D. 1153-1186), Paranavitana bad this to say ofthe landscaping of the monasteries ofPolonnaruva.

'The architects of old have given asmuch consideration to the aesthetic ef-fects created by a whole ensemble ofrelated buildings as they have to thedesign of an individu al edifice. Theskill with which the architects employedby Parakrarnabahu have tackled theproblems of site planning cao best bestudied in that extensive complex ofmonastic buildings, known as theAlahana Parivena, of which (the)Lamkatilaka (temple) was the centre.A slight elevation of the ground hadbeen made use of in order to locate onit the principal edifices intended forreligious worship or for ecclesiastical

combined with erotics, become a stand-ard feature of' epic ' and courtly poetry .

Royal parks and gardens have also beendescribed in detail in relatively lateworks such as the Kotte Vittipota and,folk poems such as the Mala Raja UruDanaya, which has an extensive de-scription "of the plants and shrubs grownin a royal garden" (Peiris 1978: 120-

121).Site, architecture and gardenAt the same time, the cornbination ofarchitecture and garden space; ofbuild-ings, trees, pathways, water and openareas; the fusion of symmetrical andasymmetrical elements; the use ofvarying 1evels and of axial and radialplanning are all inherent aspects of theSri Lankan architectural tradition, inboth religious and secular contexts.

As far as secular gardens are con-cemed, at least two well,-preserved ex-amples exist of what we might calI'pure' garden complexes. By this ismeant a garden which has an autono-mous or semi-autonomous existence,independent of a building complex, orwhich is a clearly demarcated part of alarger architectural whole or even asituation where substantial buildingsexist but form an integral though subor-dinate element in the garden layout.Two clear examples of this type are theroyal gardens at Sigiriya, which are adistinct part of the royal and urbancomplex at this site, and the RoyalGoldfish Park below the bund of theTisavava at Anuradhapura. A rare ex-ample of a monastic garden which mayalso be included in this category is theminiatllre-lake-and-island garden ofthe Kaludiya Pokuna monastery atMihintale.

Elsewhere, several monastic gardensexist where the garden forms an inte-gral part of the architecturallayout -a

6

~

-{\l1""\i"""""\;+\\~*""\\1"~~~""\\\\I "1\\\III\\\~\1\1\\"I\\"\1\\\\

lAKe\\\\\"!\\\I\\H\ti" 1I\\1\\~\"1 "\~I\ I" I tl\"\\"\\\~\\"\\\" "\\"f++\\\~\~~'!\\\

.Rock

llMeat9 1~Om

~4

5t~

<

~:

}v,~

~7

8

Sigiriya-city, palace and garden. Sth century A.D.

The 'Greater Sigiriya' area.

,1

/". ,.,~

..' :

.

i'iâfj,'.~

T~

j1;--:') 7


by the slopes of the hill on which itstands, and two rectangular precincts toeast and west, fortified by moats and

ramparts.The gardens at Sigiriya consist of

three distinct but interlinked sections:the symmetrical or geometricallyplanned water gardens; the asyrnrnetri-cal or organic cave and boulder garden,the stepped or terraced garden encir-cling the rock, the (miniature) watergarden and the palace gardens on thesummit of the rock.The Water GardensOf these, the water gardens are, per-haps, the most extensive and intricate,occupying the central section of thewestern precinct. The water gardensdivide into several distinct units. As canbe seen in the plan, the three principalgardens lie along the central east-west-axis. The largest of these gardens, Gar-den 1, consists of a central island sur-rounded by water and linked to themain precinct by cardinally-orientedcauseways. This plan anticipatesAngkor, on the one band, and the charbagh of the Mughal garciens, on theother, both of which Sigiriya predates.

The central island was originally al-most entirely occupied by a large hall orpavilion. The water-retaining structures,separated into four L-shaped pools ineach quadrant, are connected by under-ground water conduits at varying depths,suggesting an attempt at providing dif-ferential water levels. One of the twoexcavated pools is that in the south-west. It shows a subdivision into a largebathing pool with a corbelled tunneland steps leading down into it, and asmaller pool with a central boulder onwhich was placed a brickbuilt pavilion,whose scanty remains still survive. Thesides and base ofboth pools are made ofseveral courses of brickwork and were

purposes-the image-house, the stupaand the chapter-house. The slopingground on aIl sides of this central erni-nence has been made intp tenaces, andon them were located the stupas oflesser importance, the residentialquarters for the monks, Iheirrefectories,libraries and other adjuncts necessaryfor the life of ahighly organised religiousfraternity ...Ponds and sheets of waterreflected the forms of the main build-ings and the flowering trees which wereplanted at intervals must have contrib-uted to the serenity' of the scene. Natu-raI boulders were made to hannonisewith the scheme of building and thevarious edifices no doubt sat easily intheirsunoundings' (Paranavitana 1954:

29).SigiriyaThe present paper is principaIly con-cerned with the gardens at Sigiriya,which provide excellent material for acase-study of Sri Lankan garden his-

tory.The 5th century AD site at Sigiriya

provides a unique and little-known ex-ample of what is one of the oldestgarden forms in the world, whoseskeletaIlayout and significant featuresare still in a fair state of preservation.The Sigiriya gardens are the survivaIsof a fairly recently identified Sri Lankantradition of garden-art, of which thereare a few other surviving examples,some historicaI and literary documen-tation and traces and fragments at nearlyevery site of the historical period.

The gardens at Sigiriya are a con-crete and mature expression -on a grandscaIe -of these various strands andtraditions, which we see at other sitesand in literary descriptions, in a frag-mentary form. The highly systematicplan of the Sigiriya complex consists inessence of the central rock, sunounded

8

Fig. 9 Water Gardens, Sigiriya.

Fig. 10 Moats. Water Gardens and rock,

Sigiriya.

9

heavily plastered and polished.

The entire garden is a walled enclo-sure with gateways placed at the head

of each causeway. The largest of the segateways to the west, has a triple en-

trance. The cavity left by the massive

timber door posts indicates that it was

an elaborate gatehouse of timber andbrick masonry with tiled roofs, no doubtoriginally somewhat resembling thegopurams of the Kerala temples or the

torii of Japan.To the north and south of this walled

enclosure were two rectangular exten-sions, containing shallow reflectingpools, water-surrounded pavilions anddeep cisterns with a rectangular planand battered walls. Bach of these shal-low pools, as well as the water pavil-ions, which have a shallow depression

around them, and the deep ci stern, wereoriginally paved in polished marble.

The much weathered limestone slabs ofthis paving still remain in .s-itu in sev-eral places. What is distinctive aboutthis area is the precise geometry of itslayout and the application of the prin-ciple of symmetrical repetition or 'echo ,

planning. Thus, the northern and south-ern precincts are almost exactly alike,

while the eastern and western halves ofeach precinct are duplicates of each

other.Garden 2 (the 'Fountain Garden') is

a narrow precinct on two levels. Thewestern half has two long and deep

pools with stepped cross-sections.Draining into these pools are shallowserpentine 'streams' paved with marbleslabs afid defined by marble kerbs. Theseserpentines are punctuated by fountains.

The fountains consist of circular lime-

stone plates with symmetrical perfora-

tions. They are fed by underground

water conduits and operate on a simpleprinciple of gravit y and pressure. Wilh

the cleaning and repair of the under-

ground conduits, in rainy wealher thefountains operate even today.

Two relatively shallow limestone

cistems are placed on opposite sides of

the garden. Square in plan, and care-

fully constructed, they may weil havefunctioned as storage or pressurechambers for the serpentine and thefountain. The upper half of this garden,

which is raised above the lower westernsection, has few distinctive features. A

serpentine stream and a pavilion with alimestone throne are almost aIl that arevisible today.

Garden 3 is again on a higher leveland forms an extensive area ofterraces

and halls. To the north is a large octago-nal pond at the base of a toweringboulder. A raised podium and a

dripledge for a lean-to roof form theremains of a 'bathing pavilion' on thefar side of lhe pond. The pond is sur-rounded by a wide terrace which fol-lows its octagonal plan. The pond andthe high boulder to which it is joined,form a dramatic juxtaposition of rockand water at the very point at which thewater garden and the boulder garden

meet.

Adjoining the octagonal pond is anoctagonal pavilion and a circular struc-ture, both of which may be of some-what later date than the main garden

itself. Matching the octagonal pond inthe southem sector is a reclangular pool.

Recent excavations have shown thatthis pool has no structural features otherthan an earthen embankment but, as it

appears in early maps of Sigiriya, itmay weil be an untinished teature froma much later period than the 5lh cen-

tury .The central feature of Garden 3 is

a segmented. L-shaped pool. a halved

char bagh, lying on eilher side of the

central axis and fronting the entrance to


the inner citadel and the boulder gar-

den.The eastern limit ,of Garden 3 is

marked by the wide entrance andmassive brick and stone wall of theinner citadel. The citadel wall forms,as it were, a dramatic backdrop tothe wate;r gardens -echoing the evenmore dramatic backdrop of the greatrock and the palace on the summitfurther east. This backdrop extendsfrom the towering boulder near theoctagonal pond ln the north-east ofGarden 1, to the matching 'bastion' onthe s outh-east, which bas the rec-tangular pond and other buriedstructures at its base. The bastionitself is formed of wide brick wallslinking up a series of boulders, sur-rounding a cave pavilion housing arock-cut throne.

These gardens (Gardens 1,2 and 3)form a dominant series of rectangularenclosures of varying size and charac-ter,joined together along a central east-west axis. Moving away from this to thewider conception of the western pre-cinct as a whole, we see that its otherdominant feature is a sequence of fourlarge moated islands, arranged in anorth-south oriented crescent, cuttingacro~s the central ax:is of the watergarden. These, once again, follow theprinciple of symmetrical repetition, thetwo inner islands, on the one band, andthe two outer islands, on the other,

forming. pairs.The two inner islands closely abut-

ting the Fountain Garden on either side,are partially built up on surfacing bed-rock. They are surrounded by high rub-ble walls and wide moats. The flattenedsurface of the island was occupied by'summer palaces' (Sinhalese: sitalamaliga or cool palaces) or water pavil-ions (Hindi; jal mahal). Bridges, built

12

or cut into the surface rock, provideaccess tothese water palaces. Furtherto the north and south, almost abuttingthe ramparts, are the two other moatedislands, still unexcavated but clearlydisplaying the quartered or char bagh

plan.In the far north-eastem corner of the

western sector is a large and deep tankpopularly known today as the 'HabaVila', while near the southern gatewayare the scarcely discernible signs ofanother pool shown clearly on 19th

century maps.Intricately connected with the water

retaining structures of the western pre-cinct are the double moat that surroundsit and the great artificiallake that ex-tends southwards from the Sigiriyarock.Excavations have revealed that the poolswere interconnected by a network ofunderground conduits, fed initially bythe Sigiriya Lake and probably con~nected at various points with the sur-rounding moats.The hydraulic systemStanding as they do, at the mid point ofthe development of one of the world'smost sophisticated hydraulic technolo-gies (Needham 1971: 365- 378), thewater gardens at Sigiriya seem to havebeen the playground not only of thecourt but also of the ancient engineers,who applied here on a micro-scale theprinciples of the macro-hydraulicswhich formed the essential technologi-cal basis of the Sri Lankan civilizationduring the Early and Middle HistoricalPeriod.

The hydraulic system at Sigiriya stillneeds to be researched, but it is possi-ble, even in our present state ofknowl-edge, to say that its essential charac-teristic was the interconnection ofmacro~ and micro~hydraulics. The irri-gation works, conduits, pools and cis-

~


Fig. 13 Southern water palace, Sigiriya. !

Fig. 14 Inner rampart and moat, Sigiriya. I

Fig. 15 'Vertical drain', Boulder Garden, Sigiriya.1.

Fig. 16 Water inlet, Garden 1, Sigiriya. ~

Fig. 17 Fountain, Garden 2,Sigiriya. 1

Fig. 18 Drainandsilttrap, BoulderGarden, SigiriyQ;J

Fig. 19 Cobbled water course, Miniature Waterl

Garden, Sigiriya. j

14 ~

~~=

MINIATURE WATER GARCEN

r---

I v"'

i'lill

'""!iI l

;""I i

1IIi;: +"'

!ii!jl

;:ii[!i

1II11i! !\1' , ' , ...; , +

, f f~ -" :ii ? - "'c ,

", c , ..t c c .r .

, ' ~-..+ ~ ..

, .:cj I

A',~C~-."" --

.' '

~

It l i:!frl:il,; '1"

1,Iiirl

i jl l : I, 'i

II t , i

I,

il;

l 'i

l "

I '! l' t i,it

!11:iiJji

:i!il'I

I !' I '

[ i! i: Il i,1

!!'!II!'

l ll ";'

1 ';'" , !!'

I li !, !! ,

li i i 1111/il

!!!;lij:'

111'i',! I '! !"1 , , , , c

iii i'I!',1:1! I! , , ,";':1! 1 '

il !'!!" 1 '

I1 "'"

III '" :

1!,'i' ; , I, , 1,11

1' 1 "

1: '"i ,

ii;

t l,!lli

lil

~

N

T,

~q!

~~

'l ii'l! , l , I : " 1,1"" , ""

l '

I i I: , iI!

~

i

lI!

~~~

Fig.20 'Miniature' Water Garden,

Sigiriya.

Fig.21 'Miniature' Water Garden,

Sigiriya~lter excavation and

alter conservation.

16

~~!

rden.

rden,

'I and

tems sirnultaneously serve a number offunctions such as:(a) Essential water supply for domes-

tic, horticultural and agriculturalpurposes, including the omamen-tal horticulture of the gardens;

(b ) Surface drainage water control toprevent erosion of terraces and gar -

dens during the rainy season, whena substantial water-flow is createddown the slopes of the rock and thehill;

(c) Omamental and recreational watercourses and water-retaining struc-tures as an essential aspect of theroyal pleasure gardens (this in-cludes swimrning pools, bathingpools, serpentines, fountains, re-flecting ponds and pools for oma-mental flora and fauna).

(d) Cooling systems, such as a varietyof water-surrounded pavilions andthe generalized presence of waterin parklands and courtyards.

Again, at least four water systems,two macro and two micro, are con-nected together at Sigiriya, thus:(a) The lake, the Sigiri Vava, which

was much larger in the past than itis today, with a 12-krn long earthdam extending southwards fromthe base of the Sigiriya rock;

(b) A series ofmoats, two on the westand one on the east, fed directlyfrom this lake.

( c) The main water-retaining structuresof the water garden in the westernprecinct, which were intercon-nected by underground conduits,and were also -presumably -fed bythe lake and the moats (the lake is ata slightly higher elevation than thewater gardens);

(d) The water control and water-re-taining systems, which begin withthe rock-cut ponds and brick-lined

cisterns on the summit of the rockand on the various terraces below,and end in the water gardens. Theseincludes a series of horizontal andvertical' drains' cut in the rock andunderground conduits made of CY-lindrical, terracotta pipes. A deepdrain, large enough for a man towalk along it, is cut on the edge ofthe rock running along its westernperiphery , preventing water fromflowing over the western rock faceon which were depicted the paint-ings, while a vertical drain or 'gut-ter' carries this water down to anenormous cistern located on thehighest southern terrace. A numberof such vertical drains can be seenin the boulder garden.

The total conception involves theknitting together of a number of hy-draulic structures of varied scale andcharacter in a single intricate network -a complex masterpiece of irrigationengineering design that formed the hy-draulic skeleton of the landscaped gar-den (see also Ellepola 1990).The 'Miniature' Water Garden

Excavations through the mid-1980sexposed a garden unit of a hithertounknown type, in the previouslyunexcavated southwest corner of themain water gardens, in an area whichwas covered by turf and showed littleor no signs of buried structures. Theexcavation of this garden has revealedan elaborate network of structures -water -pavilions, pools, cistern, court-yards, conduits and water-courses.There are five distinct units in this gar-den, aIl combining buildings and pavil-ions of brick and limestone with paved,water-retaining structures and windingwater-courses.

The two units at the northern andsouthern extremities of the garden are

~ 17

'~

J:~~,

Senake Ban~ayake

badly eroded but as the present planshows, the generallayout of the majorportion of the garden and of the threecentral units is clear .

One of the most interesting of thesehas a shallow pool or water-surroundwith an intricately -recessed, U-shapedplan, running between a series ofbrick-built pavilions and platforms.The floor of this pool is finely pavedwith large, quartz pebbles, bringingto mind a garden detail usually associ-

atedwithJapanesegardens. Originally,these smoothly rounded pebbles layglistening under a shallow layer ofmoving water. A covered brick-builtdrain led water into this pool from aslightly higher level, while two well-preserved limestone conduits main-tained the water at a constant levelabove the pebbles, the whole concep-tion animated by a gentle continuousflow.

The water from the conduits wascarried out into a sharply-dipping wa-ter-course which, atalower level, formsalimestone fountain and an inspectiQnchamber. The sides of the pebble poolitself, have limestone mouldings of anarchaic, quarter-round type, typical ofthe Kasyapan period at Sigiriya. Pol-ished marble walkways and 'bridges'provide access across the water to thecentral pavilion.

In another unit, the entire floor of thewater-surround on three sides of a cen-tral pavilion or bathhouse is paved withmarble slabs, as elsewhere in Sigiriya.These slabs are now badly weathered,but in their original form were probablyas finely polished as modem terrazzoflooring. ln this building small gullieswith raised brims lead surplus wateraway from the water-surround into adeep, bricklined ci stem with rampedwalls.

A striking feature of this 'miniature'garden -it is in fact about 90 rn long and30 rn wide -is the use of these water-surrounds with pebbled or rnarbledfloors, covered by shallow, slowlyrnoving water. These, no doubt, servedas a coolingdevice and at the sarne tirnebad great aesthetic appeal, creating in-teresting visual and sound effects.

Another distinctive aspect is thegeornetrical intricacy of the garden lay-out. While displaying the symmetryand 'echo-planning', characteristic ofthe water-gardens as a whole, this mini-ature garden bas a far more cornplexinterplay of tile-roofed buildings, wa-ter-retaining structures and water-courses than is seen elsewhere inSigiriya -even more intricate, in fact,than the beautiful 'Pountain Garden',which lies along the central axis of themain water-garden system.

This newly-discovered garden be-longs to more than one phase of con-struction. Originally laid out as an ex-tension and 'miniaturized' refinementof the Kasyapan macro-plan in the lastquarter of the fifth century, it wasadded to later, remodelled, abandonedand again partially built over in the lastphases of the post-Kasyapan period,between the tenth and thirteenth centu-ries.

It seems very likely that a similargarden lies buried beneath the lawns ofthe unexcavated parallel sector in thenorthern half of the water-gardens- an'echo' or 'twill' of the presentgardeninthe south. In keeping with the conserva-tion policy of the Sigiriya Project, thisnorthern garden will be leftunexcavated.The Boulder Garden

If the water gardens at Sigiriya re-flect in miniature the advanced tech-nology and planning of the Sth century ,

lR

-~.E~--- ~---

~,

;~

,: 4Senake Bandaran~~e ~

20

Senake Bandaranayake~ --

above, there are also a few unusualconceptions such as the 'Cistern Rock'impluvium, taking its name from alarge cistern from massive slabs ofgranite, and the' Audience Hall Rock'which has a flattened summit and alarge 5 rn long throne carved out of theliving rock. The honeycomb of postholes and the flattened ledges of the'Preaching Rock' are another' .Whileconsiderable excavation will have to bedone before we can recover the originalpathways of the boulder garden, at least

two distinct markers are provided bytwo 'boulder arches' and limestonestaircases as well as various flights ofsteps and passageways constructed ofpoli shed marble blocks and slabs. Thevertical 'drains' cutinthesidesof rocksin a few places indicate, as discussedabove, that water courses and control-led water movement formed part of thegarden architecture in this area too.

ln its total design, therefore, theboulder garden forms a type of land-scape architecture whose principalcharacter is a number of rock clusterswith tiled roofed buildings on them and'cave' pavilions below, all elaboratelypainted and decorated, and linked to-gether by winding pathways and paved

passages and stairways.Located on the westerly slopes of the

Sigiriya hill, its western boundary de-marcated by the inner citadel wall, onthe east and the south the boulder gar-den merges with the terraces.The Terrace Gardens

The third garden form at Sigiriya,the terrace gardens, has been fashionedout of the natural hill at the base of theSigiriya rock. The terraces have beenformed by the construction of a seriesof rubbled retaining walls, each terracerising above the other and running in aroughly concentric plan around the rock.

,.,

the boulder garden looks back to amuch earlier and entirely different his-

torical environment. Derived originallyfrom the early giri monasteries, it isalso related to a much later type of

monastery known as a padhanagharaparivena or meditation monasterywhich often had buildings constructedon top of rocky outcrops, surroundedby moats (Bandaranayake, 1974: 102-

133).The boulder garden :uea is in

marked contrast to the symmetry andgeometry of the water gardens. It is anentirely organic or asymmetrical con-ceprion, consisring of a number ofwinding pathways -now mostly ob-scured and often incorrectly reconsti-tuted -which link together several clus-

ters of large natural boulders extendingfromthe southem slo~s of the Sigiriyahill to the northem slopes below the

plateau of the lion staircase.One of the most striking features of

this boulder garden is the way in whichalmost every rock and bclulder had abuilding or pavilion set upon it. Whatseem to us today like steps and drains ora honey-comb of holes on the sides ortops of boulder, are in fact the founda-rions or footings of ancient brick wallsand oftimber columns and beams. Thebuildings or pavilions were constructedon the summit of the boulders leavingthe natural form of the rock at its base.The buildings had brick walls, tim-bered superstructures and tiled roofs,none of which, of course, survives to-

day.Many of the boulders alsohaverock-

shelters beneath them, orginally fash-ioned by the early Buddhistmonks, butlater plastered and painted as another

, feature of the royal boulder garden.While most of the pavilions in this

garden area are of the type described

")")


on the summit revealed even here anelaborate underground drainage SYS-tem.Historical origins and correspond-ences

The historical importance of the gar-dens at Sigiriya lie essentially in twofactors: one, their antiquity and degreeof preservation; the other, their ingen-ious combination on a grand scale, of atleast three traditions of ancient land-scape gardening. The date of the Sigiriyagardens is clearly established. TheCulavamsa records that King Kasyapa(477-495 A.D.) built the palace on therock, the 'Lion Staircase-house', theramparts and the gardens. Past and re-cent archaeological excavations haveconfirmed that the gardens substan-tially date from the 5th century .Theyhave also shown that some elements ofthe garden and mainly the water gardenwere maintained during several post-Kasyapan phases extending from the~th to about the 13th century, and alsothat certain clearly discemible addi-tions had been made to the gardensduring the later periods of Occupation.

However, there is no doubt that thebasic features and many details of thegardens date from the period of theiroriginal layout. Successive construc-tions of some features in the gardenssuch as 'flower beds' noticed by Bell,probably relate to reconstructions dur-ing the original Kasyapan buildingphase between 477 and 495 A.D. Else-where it is clear that some of the origi-nal features of the Kasyapan gardenswere covered up or buried under aban-doned debris and a much laterconstruc-tion placed over it. The recent excava-tions have also shown indications oflater period underground earthen con-duits at a much higher level than theKasyapan water courses, and probably

Their present condition varies consid-erably. Some sections of the rubblewalling have been preserved intact,others have collapsed or have beeneroded or covered with deposits of soiland debris, while still others have beenexcavated and restored or reconstitutedin modern times. Of the original archi-tectural character of the terracing, verylittle is known, except in some of thebroader areas where the remains ofbuildings still survive on the terraces. Acolossal rubble-walled and brick-Iinedcistern on the uppermost terrace of thesouth seems to have been fed by water,channelled from the rock -summit above,forming once again part of a watercontrol system, probably related to thewater courses of the boulder garden.

The great brick-built staircases withlimestone steps traverse the terracedgardens on the west, connecting thepathways of the boulder garden to theprecipitous sides of the main Sigiriyarock itself. From here, a covered ambu-latory or gallery provides access to thebelly of the rock to what is in effect theuppermost terrace, the 'Lion Staircase'plateau, with its chambered buildingsand pavilions and the great lion itself. Athird stairway led across the terracesfrom the northem ramparts (de Silva1976:5). At the south-west, an avenueleads down towards the lake, while asteep set of modern steps connects thistoday to the southern rarnparts.The Palace Garden

A fourth garden element at Sigiriyaon a much smaller scale than the othersare the terraces and the rock -cut poolsof the summit which formed, so to say,the domestic garden of the palace itself.These terraces are found on the easternhalf of the summit -i. e. the outer palacearea -and especially on the slopingterrain to the south. Early excavations

24

emphasis on foreign influences onSigiriya or Sigiriya's influence on for-eign traditions. The clearest parallels tothe water gardens at Sigiriya are to befound in the much more ancient geo-metrical gardens of Egypt or the 'para-dise gardens' of ancientPersia, whileSigiriya' s chronological successors arethe weIl-preserved ex amples ofMughalgardening and the geometrical gardensof Renaissance Europe. Of the ancientgardens of the world that survive in anarchaelogical sense, the only signifi-cant examples that predate Sigiriya andare equally weIl preserved are those ofthe Romans such as the private andpublic gardens of Pompe ii andHerculaneum and the imperial gardensof Hadrian at Tivoli.

The subsequent developments in theAsian water gardening tradition, that isrepresented at such an early date at ~Sigiriya, are found on a majestic scaleabout four or five centuries later atAngkor in Cambodia and even muchlater in the exquisite gardens of theMughals. In both instances, we havemonuments and complexes in a muchbetter state of preservation than in anyof the Roman or Sri Lankan sites. TheMughal tradition is furtherdocumentedin miniature paintings which often havea garden set ting or actually depict gar-dens and garden activities. The rela-tionships between Sigiriya and Persian,Mughal or Cambodian traditions arenot necessarily direct ones but emergefrom an interrelated matrix of cultures.On the other hand, the fact that Sigiriyahad Roman connections is amply testi-fied to by the discovery of large num-bers of Roman and Indo-Roman coinsin the gardens at Sigiriya. At the sametime, a tenuous link is also evidencedby the discovery of a smaIl arnount ofcoin and ceramic material of probably

~ 25

~

belonging to more than one phase of

activity (Bandaral;layake 1984).The successive post-Kasyapan

monasteries at Sigiriya seem to havehad their ritual centre at the site justinside the western entrance of the innercitadel -that is to say on the westernedge of the boulder garden. At leastsome of the rock shelters in the bouldergarden and some monuments in WaterGarden 1 and Water Garden 3 showclear signs of later period construction.

As we have seen, three distinct tradi-tions of landscape gardening are foundat Sigiriya -the highly symmetricalplan of the water gardens, the organicboulder garden area and the successivestepped tiers or hanging gardens of theterraced area around the rock. Bach ofthese bas clear antecedents and succes-sors within the Sri Lankan traditionitself. Sites such as Vessagiriya,Isurumuniya, (both of wbich are asso-ciated with Kasyapa); the monasterycomplexes at Anuradhapura, and espe-cially the south western sector of theAbbayagiriya Vihara; the KaludiyaPokuna complex at Mihintale; theAlahana Parivena at Polonnaruva; sitessucb as Situlpahuva in the south orSalgala and Maligatanna in the west,are some of the many outstanding ex-amples of site selection, site modellingand landscape gardening produced bythe Sri Lanka tradition. The royal pleas-ure gardens at .o\nuradhapura thel~eside gardens and water palaces atPolonnaruva, and the royal palace andlake complex at Kandy are some of theexamples of secular landscaping andwater gardening tbat still remain in askeletal or fragmentary form.

At the same time the internationalparallels and correspondences with theSigiriya gardens are worth recalling.while taking care not to place too much

~

~~~-.r


the ancient world from the ziggurats ofMesopotamia and the 'Hanging Gar-dens' ofBabylon to the prehistoric ritualtenaces (Quaritch Wales 1961) andstepped stupas of Southeast Asia. Theyare often encountered in early lndiansites and are a conscious element inChinese landscape gardening and ar-chitectural planning. Their most obvi-ous correspondences, however, are withthe tenaced rice fields and other ter-

raced hillside agricultural systemswhich are extensively found in SriLanka, as in most parts of tropical Asia.The diversity of paralle1s and corre-spondences that we see in the gardens atSigiriya serve, in the end, on1y to under-1ine the uniqueness of this 5th centurycreation of the Sri Lankan master bui1d-

ers.TheGoldfishParkat AnuradhapuraHowever grand and comp1ex the gar-dens at Sigiriya, the more 'modest'royal garden at Anuradhapura is a 'puregar(len form ' in the fu11est meaning of

the term, in that it seems to have existedindependently of any architecturalcomplex whatsoever. Within the con-fines of the present paper it is not pos-sible to do more than to describe it

briefly.The 'Goldfish Park' or Ranmasu-

uyana at Anuradhapura lies just belowthe great earth embankment of theTisavava, one of the three western lakesproviding water for the city ofAnuradhapura, and to the north of thescenic -lsurumuniya monastery .It

consists of an elongated rectangulararea of about 15 ha demarcated by aboundary wall to north, east and south,with the towering bund of the Tisavavato the west. At the northem extremity of

the garden is a small moated precinctcontaining buildings, inc1uding two oc-

tagonal pavilions.

Sassanian origin. Sri Lanka' s knownrelationship with Carnbodia dates from

much later than the Angkorean period,but Indonesian connections go back tothe 7th century (de Casparis 1961;Sirisena 1978). The links between Per-sian and MughaI gardens, on the onehand, and Sigiriya, on the other, ismuch too distant in time to be of anyhistorical significance, apart fromconnections of Sassanian date. Butwhatever links there mayor may notbe, what is important is that arche-typaI elements common to both tradi-tions help us to understand the watergardens at Sigiriya better .

The boulder gardens at Sigiriya onthe other hand have eastern rather thanwestern correspondence. The closestparailel to the Sigiriya gardens are thegardens of China, Korea and J apan. SriLanka' s connection with China, is arich and long standing one (Werake1983; 1990) and while there is no basisto suggest any degree of influence ofSri Lanka gardening on the Far Easterntraditions, it is illuminating to comparethe two traditions. In this conection, wemay usefully quote Boyd's remarks on'The combination of symmetry and

asymmetry in. Chinese architecturalgarden planning'. 'The house and thecity were ruled by Confucian ideas:formality , symmetry , straight lines, ahierachy of importance, clarity, con-ventionaIity, a man-made order. Thegarden and the landscape were formed

by typicaily Taoist conceptions, irregu-larity , asymmetry , curvilinear , undu-

lating, and zigzag forms, mystery,originality, the imitation of nature ,

(Boyd 1962: III).The third garden form at Sigiriya,

the terraced gardens, are so basic andarchetypal in character that parallelforms exist in many diverse cultures of

26

r

I

c

Fig. 29 Ranmasu Uyana, royal pleasure gar-den.s', and the monasteries atlsurdmulliya and Vessagiriya-belowthe dam of the Tisavava lake,Anuradhapura. 3rd century B.C-13thcentury A.D.

~ 27

~~~

~~o..o

o....,-.

...,

-()-

,;

:~l'':

i "'."".:","

'."'Itt

D

t~

~

@J

c:

o

o

~

Fig.30 Ranmasu Uyana (Goldfish Park),

AnuradhapuTa. 9th-1Othcentury A.D.

Fig. 31 lsurumuniya monastery, Anuradhapura.Circa 7th-1Oth century A.D. (conjecturalreconstruction of outer moat).

Fig. 32 Vessagiriyamonastery,Anuradhapura. Jrdcentury B.C.-IOth century A.D.

29

~I

"~ci, ,

",; "

!,

:l~:,:;;:ii]

30 ~


In the southem half of the garden aretwo clusters of natural boulders, whilein the east are three large elongatedpools. Around the central cluster ofboulders are two highly omarnentedstone baths, or swimming pools. Thesebaths are elaborately designed with re-lief sculptures of elephants sportingarnongst lotuses, an artificial waterfallwith a rock-cut charnber placed behindit, water storage cistems and complexwater systems controlling the flow ofwater into the baths. The water systemshere, dating not later than the 10th cen-tury A.D. are weIl preserved. Althoughon a smaller scale, they are as, or moreelaborate, than the water systems in thegardens at Sigiriya, from a period fivehundred years earlier, or those in theMughal gardens of India and Pakistan,more than five hundred years later .Isurumuniya and Vessagiriya

The Ranmasu Uyana appears todayas an isolated garden but it is likely thatit was once connected with theIsurumuniya complex lying immedi-"ately next to its southem boundary.Isurumuniya today is one of the mostbeautiful monasteries in Anuradhapuraunfortunately somewhat modemized inrecent times, with a collection of inter-esting sculpture of a largely secularcharacter, including the well-known'Isurumuniya Lovers'. The rock-and-pool garden plan at Isurumuniya isreminiscent of Sigiriya. The juxtaposi-tion of the Isurumuniya and RanmasuUyana formations provides us with agarden plan of extraordinary beautyand complexity. Apart from the in-scriptions which identify and date theRanmasu Uyana, the history of thesetwo complexes is not clear and it islikely, as in the case of Sigiriya, thatthey carne under royal and monasticuse collectively or separately at various

times in their long history .This history is further complicated

by the affiliation of these two com-plexes with the Vessagiriya monasterylying to the f~ south of the Tisavavadam. Several scholars have suggestedthat the rock monastery at Vessagiriya,itself an outstanding example of SriLankan landscape architecture, datingfrom at least the 3rd century B.C. is infact the ancient Issarasamana (orIsurumuniya) monastery , thus alsoopening up the possibility that therewas once a large monastery --in fact thesouthemmost of the monasteries ofAnuradhapura --which extended fromVessagiriya northwards to the presentIsurumuniya and the boulders of theRanmasu Uyana, along the great earthdam of the Tisavava. That Vessagiriya,itself, bas conne.ctions with Sigiriya inthe Sth century A.D. bas also beensuggested on the basis of its architec-tural detailing and painting remains.The Kaludiya Pokuna monastery,Mihintale

The Kaludiya Pokuna or 'Dark Wa-ter Monastery' is located at thesouthem extremes of Mihintale, an an-cient monastic centre lying eight milesto the west of Anuradhapura, and asso-ciated with the conversion to Buddhismof the first Sri Lankan King, Devanam-piya Tissa (2S0-210 B. C.) bythe arahatMahinda, son of the Indian emperorAsoka. The monastery in its ruinedform today consists of five distinct ele-ments: an entrance pathway windingthrough boulders and trees (i.e.a Boul-der Garden, a geometrically laid outmonastic complex ofbrick-and-timberstructures, now vestigially preserved;buildings in stone (and also originallywith brick -and-timberelements ) locatedon boulders, elevated terraces, or inrock-shelters; natural boulders inte-

32

f:1;;.~

~enake Bandaranayake

J

Gooneratne, Sita Padmini. 1983. Water Gar-dens and Water Spots in Sinhala Literature.Proceedings of 'Seminar on Sigiriya'. 15October 1983. Colombo: Central CulturalFund and Sri Lanka Foundation lnstitute

(Mimeo)The Mahavamsa, or the Great Chronicle of

Ceylon. (Translated by Wilhelm Geiger, as-sisted by Mabel Bode,with an addendum byG.C. Mendis). 1960 (reprint). Colombo:Ceylon Government Information Depart-ment.

Needham, Joseph. 1971. Science and Civiliza-tion in China, 4 (3) Cambridge: CambridgeUniversity Press.

Paranavitana, S. 1954 Art and Architecture ofCeylon; PolonnalUva Period. Colombo: Arts

Council of Ceylon.Peins, E. 1978. Studies Historical and Cultural,

Colombo: Catholic Press.Quaritch Wales, H.G" 1961. The Making of

Greater India, London.De Silva, R.H. 1976. Sigiriya, Colombo: De-

partment of Archaeology.Silva, Roland. 1972. LessonsofTown Planning

from Ancient Ceylon: Institute of ArchitectsJoumall(I), July-Dec. 1972.

Sinsena, W.M.1978.SriLAnkaandSouthAsia:Political, Religious and Cultural Relationsfrom A.D... c. 1000 to c. 1500. Leiden: E.J~

Bnll.Werake,Mahinda. 1984. TheAbhayagiriVihara

and Sino-Sri Lanka Buddhistic Relationsduring the Anuradhapura Period, TheQuarterly of the Cultural Triangle 2(3): 23-

28.Werake, Mahinda. 1990. Sino-Sn Lankan Rela-

tions during Pre-Colonial Times. Sri LAnkaand the Silk Road of the Sea, (eds. S.Bandaranayake et.al.) Colombo: Sri LankaNational Commission for UNESCO andCentral Cultural Fund: 221-231.

Notes1. Earlier versions of the present article were

published in Proceedings of 'Seminar onSigiriya'15 October 1983 (mimeo) andAn-cient Ceylon No.10, 1990.

grated often dramaticaUy, into the

building or water system; and finally,

the 'Dark Water' lake itself extendingeastwards along the eastern flank of themonastery , surrounded by the outerprecincts of the monastery and a

boundary wall. The entire monasterynestles at the foot of the A Ilaikutikanda,the Southernmost peak of the severalhills which form the Mihintale com-

plex. Together with the IsurumuniyaVihara, it is perhaps one of the two best

preserved water-and-boulder gardenmonasteries of the Earl:f and Middle

flistoric Period in Sri Lanka.

These monastic gardens togetherwith the garden forms from

Polonnaruva, Sigiriya and the Ranmasu-uyana at Anuradhapura, provide uswith unique data and beautiful and dra-

matic examples of Asian garden designunrepresented in any other South Asian

tradition from such an early period.

ReferencesASCAR Archaeological SuTV,ey of Ceylon, An-

nual Report.Bandaranayake, Senake. 197"~. Sinhalese Mo-

nastic Architecture, Leiden: E.J. Brin.Bandaranayake, Senake. 1983. The Architec-

tura! and Spatial Organization of the Sigiriyacomplex. Proceedings of 'Seminar onSigiriya'15 October 1983. Colombo: Cen-tral Cultural Fund and Sri Lanka Foundationlnstitute (Mimeo)

Basnayake, H.T. 1983. Th(~ Giri Monastery,Samskrti. 2.

Boyd. Andrew, 1962. Chinese Architecture andTown Planning, 1500 B.C.. toA.D... 1911.London: Alec Tiranti.

Culavamsa; Being the More Recent Part oftheMahavamsa. (Translated by Wilhelm Gei-ger and C. Mabel Rickmers. 1953 (reprint).Colombo: The Ceylon Govemment Infor-mation Department.

De Casparis, J.G. 1961. New Evidence on Cul-tural Relations between Java and Ceylon inAncientTimes. Artibus Asiae 25: -241-248.

Enepola, Chandana, 1990. Conjectured Hy-draulics of Sigiriya. Ancient Ceylon No. Il:169-227.

34

~nake Bandaranayake

AcknowledgementsThe photographs in this al1icle are by Maya Upanada of the Central Cultural Fund's $igiriya Cultural

Triangle Project, with the followillg exceptions: Fig. 3: Archaeological Departrnent; Fig. 10: Robert $wartheInc., Los Angeles; Figs. 19, 22, 25, 26: $arath Athulathmudali of the $igiriya Cultural Triangle Project; Fig. 24:$enake Bandaranayake; Fig. 37: I. $. Madanayake of the Postgraduate Institute of Archaeology (PGIAR),University of Kelaniya.

The plans are by the $igiriya Cultural Triangle Project and the PGIAR Cartographic Unit, with thefollowing exceptions: Fig. 5: Mallendra Adhikari, Ravindra Athukorala, Ganga Munasinghe and $hamilaJayathilaka, under the direction ojr c. Anjalendran; Fig. 6: Institute for Photogrammetry and Cartography,Fachhochschule. Karlsruhe and tJle Dambulla Cultural Triangle Project; Figs. 29 to 32: ArchaeologicalDepartrnent and adapted or revised by the PGIAR Cartographic Unit; Figs. 33-35: Archaeological Department;Sources Unknown: Figs. 33, 38;

The author wishes to thank Maya Upananda, I. $ $. Madanayake, G. F. de Alwis and Asoka Perera forpreparing and editing the photographs and illustrations and Robert $warthe for so kindly permit ting the use ofmaterial from bis $igiriya aerial pl1OtO collection for articles and books on $igiriya by the Cultural Triangle

Project.

Taoist Valley.

36

senake bandaranayake

Documents