Gudrun Bühnemann - Mandalas and Yantras in The Hindu Traditions

BRILLS
INDOLOGICAL
LIBRARY
E D I T E D BY
JOHANNES RRONKHORST
IN C O - O P E R A T I O N W ITI-i
R IC H A R D G O M B R IG Ii O S K A R V O N H IN U B E R
K A TSU M I M IM A K I ARVIND SMARM A
V O L U M E 18
MANDALAS AND YANTRAS

IN THE HINDU TRADITIONS
BY
G U D R U N BU H N EM A N N
with contributions by H. Brunner,
M.W. Mcister, A. Padoux, M. Rastelli
and J. Torzsok
B R IL L
L E ID E N B O S T O N
2003
T his book is printed on acid-free paper.
L ib r a r y o f C o n g r e s s C a ta lo g in g -m -P u b lica tio n D a ta
B iih n e m a n n , G u d r u n .
M a n d a la s a n d Y a n lra s in th e H in d u tra d itio n s / G u d r u n B ith n e m a n n .
p. cm . (B rills In d o lo g ic a l lib ra ry ; v 18)
In c lu d e s b ib lio g ra p h ic a l referen c es a n d in d ex .
IS B N 9 0 -0 4 -1 2 9 0 2 -2
1. M a n d a la . 2. Y an tras. 3. H in d u sy m b o lism . I. T itle. II. Series.
B L 2 0 1 5 .M 3 B 8 5 2 0 0 2
2 9 4 .5 3 7 ~ d c2 1
2002043732
IS S N
IS B N
0 9 2 5 -2 9 1 6
90 04 12902 2
Copyright 2 0 0 3 by Koninklijke B rill NV, I A d en , T he .Netherlands
A ll rights reserved. Jsro part o f this publication may be reproduced, translated, stored
in a retrieval system, or transmitted m any form or by any means, electronic,
mechanical, photocopying, recording or othemrise, without prior written
permission from the publisher.
Authorization to photocopy items for internal or personal
use is granted by Brill provided that
the appropriate fees are paid directly to 7 he Copyright
Clearance Center, 2 2 2 Rosewood Drive, Suite 9 ! 0
Danvers, M A 0 1 9 2 3 , USA.
Pees are subject to change.
P R IN T E D IN T il K NETHERLANDS
CONTENTS
Notes on Contributors ...................................................................

List o f I llu s tra tio n s.....................................................
v [\
jx
In tro d u c tio n .......................................................................

Gudrun Biilm em ann
Mandala, Yantra and Cakra:
Gudrun Buhnem ann
Some O b s e rv a tio n s ..........................
13
Mandalas and Yantras in Smarta R it u a l...........................................

Gudrun Biilm em ann
Part I. Selected Mandala-like Structures, M andalas and
Yantras
Part 11, Bhadramandalas
57
The Use o f Mandalas and Yantras in the Pancaratra Tradition ...

M arion R astelli
Mandala and Yantra in the Siddhanta School o f Saivism:
Definitions, Description and Ritual U s e ......................................
H elene B runner
Icons o f Inclusivism: Mandalas in Some Early Saiva T a n tr a s ....
Judit Torzsok
Mandalas in A bhinavaguptas Tantraloka ...........................
A n d re Padoux
The Sricakra according to the First Chapter o f the
Y o g im h rd a y a ....................................................
A ndre Padoux
119
j 53
179
225
239
Vastupurusamandalas: Planning in the Image of Man ................

M ichael W. M e is ter
251
Bibliography and Abbreviations.........................................................
27 1
I n d e x ..........................................................................................................
291
NOTES ON CO NTRIBU TO RS
H E L E N E B R U N N E R [-LACHAUX] was a m em b er o f the Centre

national de la recherche scientifique (Section o f Oriental Studies),
Pans, from 1963 to 1985, She retired in 1985, but continues working
in hei field, that is Agamic (Saiddhantika) Saivism, as represented in
mediaeval texts, A gam as or paddhatis. Her m ajor publications in
clude an edition and French translation o f the Som asanibhupaddhati
(four volumes, Pondichery: Institut Frangais d lndologie 1963-1998)
and a French translation o f the kriyapada and caryapada o f the M rgendragama (Pondichery: Institut Frangais d lndologie, 1985).
G U D R U N B U H N E M A N N is a Professor at the Departm ent o f L an
guages and C ultures o f A sia o f the U niversity o f W isconsinM adison. Her m ajor publications include B u d h a -K a u sika s Ranraia ksa sto tia : A C ontribution to the S lu d y o f Sanskrit D evo tio n a l
Poetry (Vienna: Institut fur Indologie, Universitat Wien, 1983); Puja:
A S tu d y m Smarta R itual (Vienna: Institut fur Indologie, Universitat
Wien, 1988); The W orship o f M ahaganapati according to the N ityotsava (Wichtrach: Institut fur Indologie, 1988); F orm s o f Ganesa: A
S tu d y based on the Vidyarnavalantra (Wichtrach: Institut fur Indo
logie, 1989); and The Iconography o f Hindu Tantric Deities. Volume
I: The Pantheon o f the M antramahodadhi. Volume II: The Pantheons
o f the Piapahcasara and the Saradatilaka (G roningen- Egbert Forsten
2000- 2001).
M I C H A E L W . MEISTER holds the W. Norman Brown Chair o f South

Asian Studies at the University o f Pennsylvania. He has edited the
Encyclopaedia o f Indian T em ple A rchitecture, D iscourses on Siva,
M a kin g Things in South A sia , C o o m a ra s w a m y s E ssays in E arly
Indian A rchitecture and E ssays in A rchitectural Theory, C ooking for
the G ods and Ethnography and Personhood.
is Dirccteur de recherche honoraire at the Centre

national de la recherche scientifique, Paris. His publications include
Vac. The C oncept o f the W ord in S elected H indu Tantras (Albany,
ANDRE PADOUX
viii
NOTES ON CONTRI BUTORS
N ew York: State University o f N ew York Press, 1990); L e coeur de

la YoginJ. Yoginlhrdaya, avec le com m entaire DIpika d Am rtananda
(Paris: De Boccard, 1994); and (with Lilian Silburn) Abhinavagupta.
La lum iere sur les tantras, chapitres 1 a 5 du Tantraloka, traduits el
com m entes (Paris: De Boccard, 1998).
M A R IO N R A STELLI received her Ph.D. in Classical Indian Studies
(Indology) in 1998 from the University o f Vienna. She is presently a

research fellow at the Institute for the Cultural and Intellectual
History o f Asia o f the Austrian A cadem y o f Sciences, Vienna. Her
research focuses on the teachings and ritual o f Pancaratra. Her most
im portant publications are: P hilosophisch-theologische G rundanschauungen der Jayakhyasaniliita. M il einer D arstellung des taglichen R itu a ls (Wien: V erlag der O sterreichischen A kadem ie der
W issenschaften, 1999) and The R elig io u s Practice o f the Sadhaka
A c co rd in g to the Jayakhyasam hita (Indo-lranian Journal 43, 2000:
319-395).
JUDITTORZSOK is Lecturer (Martre de conferences) in Sanskrit at
the Universite Charles-de-Gaulle-Lille 3 (France). She completed her
D.Phil. thesis in Oxford in 1999 on a Tantra entitled the Siddhayogesvarlmata, belonging to the tradition o f the Trika. Her research
focuses on the early Saiva Tantric traditions. Since 2001 she has
been contributing to the Hindu T antric D ictionary (Tantrikabhid h a n a k o s a ) and p a r tic ip a te s in the S k a n d a -P u ra n a p r o je c t
(Groningen, the Netherlands).
LIST OF ILLUSTRATIONS
Illustrations to M andala, Yantra and Cakra: S o m e O bservations b y

Gudrun Biilm em ann:
1. A yantra for attraction (akarsanayantra), described in M antram ah o d ad h i 2 0 .8 4 c d -8 5 ; reproduced from the M a n tra m a h o d a d h i
(edited by R. Prasada, Lucknow: Sm arahim sakadatta Press, 1872,
appendix). The nam e D evadatta (to be replaced with the intended
p ersons name) appears in the centre o f the pericarp o f the lotus, and
the krodhabija (that is, the syllable h u m ) in the four lotus petals. The
yantra should be drawn on a leaf o f birch-bark using a mixture o f red
sandalwood paste and one s own blood. After the puja the yantra is
normally soaked in ghee.
2. A supporting mandala for the vardliani vessel in Tantric puja;
reproduced from Buhnem ann 1988b, Illustration 36.
3. A ca. 17th-century yantra with the R am araksastotra inscribed
on textile (33 x 34 cm); photograph by G. Buhnem ann. In 1979 this
yantra was part o f the collection o f Sarabhai N aw ab (Ahmedabad).
4. A yantra assigned to stanza 55 o f the Saundaryalahari; repro
duced from Saundaryalahari 1957: 87. The syllable ya m (w hich is
usually the seed [blja] syllable o f the elem ent w ind) is inscribed
thrice in the yantra. A ccording to the instructions (jjrayoga), the
yantra should be draw n on a golden plate or on fish-bone. One
should repeat the stanza 20,000 times a day for 45 days to secure
freedom from bondage.
5. A yantra for subjugating o n e s master, described in D am o d aras
Y an tracin tam an i 3 . 2 0 -2 6 (T urstig 1988: 2 1 ); re p ro d u c e d from
Turstig 1988, appendix, yantra no. 3. The nam e D evadatta (to be re
placed with the intended p e rso n s name) appears in the centre o f the
pericarp o f the lotus prefixed by the syllables om srlm and suffixed
by srlm om. O n the lotus petals the syllables srim and ksah alternate.
The yantra should be draw n on a le a f o f birch-bark using yellow
pigm ent. It should then be placed into a vessel and burnt, and its
ashes consumed.
LIST OF I LLUSTRATIONS
6.
A pujayantra o f M ahaganapati, reproduced and adapted from
Buhnemann 1988b, Illustration 40. The yantra features a downwardpointing triangle inside a hexagram, surrounded by an eight-petalled
lotus and a square with four gates.
Illustrations to M andalas and Yantras in Smarta R itu a l b y Gudrun

Buhnem ann:
Part I. Selected M andala-like Structures, Mandalas and Yantras
Following a com m on South Asian tradition o f depicting mandalalike structures, the east (and not the north) is shown on top o f the
diagrams.
1. A diagram showing the deities o f the baliharanacakra, or baliharanamandala, reproduced from K ane 1968-1977, volum e 2: 747.
W ith m inor variations, this diagram is found in a num ber o f con
temporary texts, such as the Rgvedlyabrahmakarmasamuccaya.
2. A table showing the arrangement o f the five deities in (dom e
stic) pancayatana shrines as prescribed by Bopadeva.
3. A m andala o f the heavenly bodies (grahadevatanvandala or
navagrahamandala); a contemporary print reproduced from the ritual
manual Rgvedlyabrahmakarmasamuccaya.
4. The (durga)saptasatlmahayantra; a contemporary print reprodu
ced from the manual Rgvedlyabrahmakarmasamuccaya.
5. The rudraplthamahayantra; a contem porary print reproduced
from the manual Rgvedlyabrahmakarmasamuccaya.
Part II, Bhadramandalas

Table: Constituent Parts o f the Bhadramandalas.
Black and White Prints o f Mandalas
The black and white prints o f mandalas which are listed below are
reproduced from the Bhadramartanda (BM) edition. The edition does
not illustrate all mandalas described in the BM. In addition, some
m andala draw ings are incom plete and have not been reproduced
LIST OF I LLUSTRATI ONS
xi
here. Since the original numbers o f the illustrations as printed in the

BM have been retained, some numbers are m issing.1 A com plete list
o f m andalas described in the BM is provided in the appendix to the
article. The diagrams use the following scheme to indicate colours
other than black and white: one dot in the centre o f a square yellow; two dots - red; and three dots - green.
1. Sarvatobhadra, type 1
3. Sarvatobhadra, type 3 (= astadalamandala)
5. Ekalirigatobhadra laghugauritilaka
6 . Caturlingatobhadra brhadgauritilaka
7. Gaurltilaka
8 . Caturlingatobhadra, type 1
9. Caturlingatobhadra, type 2
12. Astalihgatobhadra, type 1
13. Astalihgatobhadra, type 2
14. Dvadasalihgatobhadra, type 1
( hariharatmakadvadasalingatobhadra)
( hariharatmakadvadasalinga tobhadra)
2 0 . Dvadasalihgatobhadra, type 7
(.hariharatmakadvadasalingatobhadra)
2 1 . Dvadasalihgatobhadra, type 8
(hanharatmaka/[hanhara]dvadasalihgatobhadra)
2 2 . Dvadasalingatobhadra, type 9 (latalihgalobhadra)
23. Dvadasalihgatobhadra, type 10 (lihgasvasdkabhadra)
26. Sodasalihgatobhadra, type 1
27. Sodasalihgatobhadra, type 2
( sodasaUngodbhavahariharamandala)
28. Saptadasalingatobhadra, type 1
1T h e se arc the num b ers 2 4 - 2 5 , 36, 3 8 - 4 0 , 5 0-51 and 5 3 - 6 5 .
xii
29. Saptadasaliiigatobhadra, type 2

31. Caturvimsatilingatobhadra
32. AstavimSatiliiigatobhadra
33. Pancavimsatilingatobhadra, type 1
35. Astottarasatalingatobhadra, type 2
37. Ekavimsottarasatalingatobhadra, type 1
41. Ekamudraramatobhadra
42. Caturmudraramalirigatobhadra (see also Colour Plate 7)
43. Astamudraramatobhadra, type 1
44. Astamudraramatobhadra, type 2
( astam udraramalihgatobhadra)
45. Navamudraramatobhadra
46. Dvadasamudraramalingatobhadra
47. Trayodasamudrararnatobhadra
48. Laghusodasaramamudraramaiingatobhadra
49. Sodasamudraramatobhadra
52. Astottarasataramalingatobhadra
6 6 . Pancabhadra
67. Suryabhadra, type 1 (see also Colour Plate 10)
6 8 . Suryabhadra, type 2
69. Ganapatibhadra vighnamarda
70. Cakraravindamandala
71. Svastikamandala
72. Trayodasalirigasamudbhavamandala
73. Caturdasalingatobhadra
74. Vimsatilingatobhadra
75. Catvarimsallingatobhadra
76. Sastilihgatobhadra
Illustration to The Use o f M andalas and Yantras in the Pahcaratra

Tradition b y M arion Rastelli:
1,
Ca. 17th-century bronze statue o f the 16-armed SudarSanacakrapurusa in the Sri-Kalamekaperumal Temple, Tirumohur, M adu
rai District; photo courtesy o f the Institut Franpais de Pondichery/
Ecole Frangaise d Extreme-Orient.
LIST OF I LLUS TRATI ONS
xiii
Illustration to M andala and Yantra in the Siddhanta S c h o o l o f

S a iv ism : D e fin itio n s, D e sc rip tio n and R itu a l U s e b y H e le n e
Brunner:
1. The sarvatobhadramandala reconstructed according to Saradatilalca 3.106 131 ab and R aghavabhattas commentary.
Illustration to Icons o f Inclusivism : M andalas in S o m e E arly Saiva

Tantras b y Judit Torzsok:
1.
The outline and construction o f the srlmandala according to the
Netratantra with an illustration o f some technical terms (see A p p e n
dix 1 for a description); drawing by Paul Coatalen.
(In the illustration the bottom is the w estern direction, for the
disciple would enter and see the m andala from the west, facing the
auspicious eastern direction.)
Illustrations to M andalas in A b h in a va g u p ta s T antraloka b y A n d re

Padoux:
1. Outline o f the m andala o f the three tridents and (seven) lotuses
(tritrisulabjamandala) p rescrib ed by the T rik asa d b h a v a ta n tra ; see
Tantraloka 31.10
41 b; drawn by and reproduced w ith the kind p e r
mission o f Stephanie Sanderson.
2. Outline o f the m andala o f the trident and lotuses ( Irisulabjamandala) prescribed by M alinivijayottaratantra 9.6-31 (= Tantraloka
31,6285b); drawn by and reproduced with the kind perm ission o f
Stephanie Sanderson.
3. The m andala throne and the three goddesses enthroned upon it,
as v isualized along the axis o f internal sensation during internal
worship; see T antraloka 15.295c-328b; draw n by and reproduced
with the kind permission o f Stephanie Sanderson,
X IV
LIST OF ILLUSTRATIONS
Illustrations to The Sricakra according to the First Chapter o f the

Yoginlhrdaya b y A ndre Padoux:
1. The sricakra, reconstructed and drawn by Gerard lluet and re
produced with his kind permission.
2. The bodily cakras according to the first chapter o f A m rtan a n d a s com m entary on the Y oginlhrdaya; adapted from Padoux
1994: 126.
Illustrations to Vastupurusam andalas: Planning in the Im age o f

M a n b y M ichael W. M eister:
1. The geometric construction o f a cardinally oriented square loca
ting an altar as defined in 6 ulba-Sutra texts o f the third-fourth
centuries B.C.
2. Vastupurusam andala o f 81 squares, as described in the BrhatSamhita (Apte/Supekar 1983: VPM-4).
3. Ground plans and constructing mandalas:
A. B h aratesv ara tem ple, B hubaneshw ar, Orissa, ca. 6 0 0 -6 5 0
century A.D. The walls, in thickness, are h alf the breadth o f the
sanctum. Entries project beyond the square m andala (karnavyasa
system).
B. Siva temple no. 2, Mahua, M adhya Pradesh, ca. 650-675 A.D.
M easurem ents at the hoof-m oulding o f the vedlbandha demonstrate
that projections on the walls directly express the inner space o f the
sanctum and the central place for B ra h m a n ( brahmasthana) o f the
vastumandala.
C. M ah ad ev a tem ple, A m rol, M adhya Pradesh, ca. 700 A.D.
Diagram to demonstrate extension o f the brahmasthana and sanctum
to demarcate the central offsets o f the outer walls.
D. Naktim ata temple, Bhavanipur, Rajasthan, ca. 875 A.D. In the
ninth century, the m andala diagram is used in a different way, still
dem arcating the central offsets, but pulling them within the c o n
structing grid (bhadravyasa system).
4. BrhadlSvara temple, Garigaikondacolapuram , Tainilnadu, ca.
1035 A.D. Plan with odd-numbered grid (after Pichard 1995: 47).
LIST OF I LLUSTRATI ONS
xv
5. Khajuraho, M adhya Pradesh:

A. V iSvanatha te m p le , 1002 A .D . A n e v e n -n u m b e re d grid
surrounds the sanctum at the level o f the vedlbandha base m ouldings
and extends from there.
B. L aksm ana tem ple, socle and vedlbandha m ou ld in g s (after
Kramrisch 1946: 259).
C. K andariya M ahadeva tem ple, ca. 1 0 2 5 -1 0 5 0 A.D, An evennum bered grid extends from the sanctum , seem ing to overlap a
parallel grid centred on the mandapa that shelters the worshipper.
6 . A. Tabari temple, Kharod, M adhya Pradesh, ca. seventh cen
tury. P lan and e m b e d d e d octagon c o n s tru c te d u s in g an oddnumbered grid.
B.
Gargaj M ahadeva tem ple, Indor, M a d h y a Pradesh, ca. 750
A.D. Turned-square plan producing 12 bhadras: A -D are Siva and
his family, 1-8 are the eight dikpalas (guardians o f the directions),
7. C om parison o f the constructing geom etry, and the continuing
regulative function o f odd-num bered grids, in plans based on three
and six turned squares: left, Gargaj M a h a d e v a tem p le, Indor,
M adhya Pradesh, ca, 750 A.D.; right, U dayesvara M ahadeva temple,
Udayapur, M adhya Pradesh, ca. 1081 A.D.
C olour Plates
1. A yantra o f G uhyakali from the Book o f Pictures Containing
Images and Y a n tra s ,5 preserved in the B harat K ala B havan, V a ra
nasi, no. 10054, dated 1764/65 A.D.
2. The baliharana o f the vaisvadeva rite perform ed by a Rg-Vedin;
photographed by G. B uhnem ann in Pune, Maharastra.
3. A dom estic sivapahcayatana in Pune, M aharastra; photograph
by G. Buhnemann.
4. A rudrapithamahayantra in w hich areca nuts re p re s e n tin g
deities have been placed and w hich serves as a support for a vessel
with the icon o f Rudra/Siva during the rudrayaga; Pune, M aharastra;
photograph by G. Buhnemann.
5. A sarvatobhadra; a contem porary print reproduced from the
manual Rgvedlyabrahmakarmasamuccaya.
xvi
6 . A ramalihgatobhadra with 26 ramamudras and 28 Jihgas; pain

ting preserved in the S.R.C. M useum o f Indology, Jaipur; photograph
by G. Buhnemann.
7. A caturmudraramalihgatobhadra with 4 ramainudrtK and 8 lihga s and a sarvatobhadra in the centre; painting from Rajasthan repro
duced from Stadtner 1998: 350, no. 353, with the kind permission o f
Robert Clark, Barcelona. This bhadra corresponds to the bhadra
reproduced as Illustration 42 from the Bhadramartanda (see Illustra
tions to M andalas and Yantras in Smarta Ritual [Part II. B hadra
mandalas] by Gudrun Buhnemann).
8 . A ganesabhadra with five icons o f Ganesa reprinted from Ganeskos (edited by A. GadgTI, Pune: SrTram Book Agency, 1981 [in
Marathi]): 477.
9. A ganesabhadra with 21 icons o f Ganesa; painting preserved in
the Bhandarkar Oriental Research Institute, Pune, Maharastra.
10. A suryabhadra with 12 icons o f the sun; painting from R aja
sthan reproduced from Stadtner 1998: 350, no. 352, with the kind
permission o f Robert Clark, Barcelona. This bhadra corresponds to
the bhadra reproduced as Illustration 67 from the Bhadram artanda
(see Illustrations to M andalas and Yantras in Smarta Ritual [Part
II. Bhadramandalas] by Gudrun Buhnemann).
11. The construction o f a dvadasalihgatobhadra with a sarvato
bhadra in the centre; photographed by G. B uhnem ann in Pune,
Maharastra.
12. The invocation o f deities into areca nuts placed on a sarvato
bhadra during a vratodyapana rite; photographed by G. Buhnemann
in Pune, Maharastra.
13. A vessel o f plenty placed on a sarvatobhadra during a vratod
yapana rite; photographed by G. Buhnemann in Pune, Maharastra.
14. The cakrabjam andala reproduced from the colour print in
serted in the back o f volum e 1 o f P a d m a n a b h a n s edition o f the
Padma-Samhita.
15. The navapadm aniandala reproduced from the colour print pre
fixed to the foreword of the first edition o f the Jayakhya-Sam hita by
K rishnam acharya (1931). Apte 1973: 505 points to the fact that the
draw ing is not in accordance with the textual description in every
detail. The nine lotuses should have the same size and should be
adjacent to one another.
LI ST OF I LLUS TRATI ONS
X V ll
16. The M m andala o f the N etratantra follow ing K s e m a ra ja s

com m entary (see T o rz s o k s A ppendix 1 for a description); illustra
tion by Paul Coatalen.
(In Colour Plates 16-19 the bottom is the w estern direction, for
the disciple would enter and see the m andala from the west, facing
the auspicious eastern direction.)
17. An alternative structure o f the srunandala o f the N etratantra
(see T o rz s o k s A ppendix 1 for a description); illustration by Paul
Coatalen.
18. The mandala o f the nine lotuses (navanabham andaJa) a cco r
ding to the Svacchandatantra and K s e m a ra ja s co m m e n ta ry (see
T o r z s o k s A p p en d ix 2 for a d e scrip tio n ); illu stra tio n by Paul
Coatalen.
19. A tentative reconstruction o f the trid en t m an d ala o f the
SiddhayogeSvarimata (long recension) according to the T antraloka
(see T o r z s o k s A p p e n d ix 3 for a d esc rip tio n ); illu stra tio n by
Guillaume Coatalen.
IN TRO DU CTIO N
Gudrun Buhnem ann

General R em arks
In recent years mandalas have attracted much interest am ong a wider
public. The main focus o f such interest has been directed tow ard
Tibetan mandalas, specimens o f which have been included in n u m e
rous publications. But mandalas are found across a w ide spectrum o f
South Asian religious traditions, including those o f the Hindus and
Jains. M andalas are also part o f East Asian Buddhist traditions.
In South Asia, mandalas have been used mainly in occasional rites
o f worship. In these rites deities are invoked into m andalas with the
aid o f mantras. The construction o f a m andala is specially important
in Tantric initiation ( dlksa) rites, In esoteric teaching, a m andala may
be visualized as present in the practitioners body by correlating the
cosmic symbolism o f the m andala with the practitioners body parts.
M andala patterns have had other far-reaching influences. They have,
for example, had an impact on ancient tow n-planning. The use o f
mandalas is also documented in alchemy.'
The South Asian tradition o f preparing and w orshipping mandalas
and yantras continues up to the present. O n the level o f folk art the
kohbar mandalas, which decorate the walls o f the nuptial cham ber in
the M ithila region o f north B ihar (India) and N epal, are a good
exam ple o f this, So are the auspicious floor designs prepared with
rice flour or coloured pow ders and reg io n ally k n o w n as rahgoll,
alpana, m u g g u lu or kolam , w hich have been influenced by m andala
and yantra patterns.
Y antras have been em ployed especially in rites o f magic. Their
use has b een reco m m en d ed in astrology and, to som e extent, in
A yur-Veda. T he yantra o f a deity is custom arily placed u n d er the
d eity s statue at the tim e o f its installation in a tem ple. Patterns o f
1
For a detailed ex am in a tio n o f the use o f m antras, yan tras and m a n d a la s in A y u r
V eda and in alchem y, sec R o?u 1986a an d 1986b.
G UD RU N B UHN EMA NN
yantras, like those o f mandalas, have had widespread influence. In

the citrabandha compositions in Sanskrit, for example, text can be
arranged in yantra-like shapes.2
Like mandalas, yantras continue to be worshipped in South Asia.
The sricakra or sriyantra, which is a configuration o f a ccntral point
and sets o f triangles surrounded by lotus petals, circles and a square,
is widely worshipped in contemporary India and Nepal. It is installed
and w orshipped, am ong other places, in the Srngcri m atha, which
claims to uphold Sam karas tradition. In Nepal, it decorates roofs o f
shrines. The sricakra is now also sold as a pendant to be worn around
the neck, and is printed on popular wall calendars. A numerical
yantra, the visoyantra ,3 is currently worshipped in Ambaji, G ujarat.4
Popular books promote yantras for miscellaneous mundane purposes,
including safe driving. C opper yantras from India can easily be
purchased over the Internet for similar purposes.
Patterns typical o f m andalas and yantras have inspired m odern
Indian architecture, art and dance, The Mumbai-based contemporary
architect Charles Correa has been guided by mandala designs in his
layout o f buildings, such as the new State A ssem b ly (V idhan
Bhavan) in Bhopal. Inspired by a navagrahamandala pattern, Correa
designed the Jaw ah ar Kala K endra, a cultural centre in Jaipur.
C orreas Surya Kund in Delhi is said to be based on a mandala plan
featuring the sricakra in its centre.5 Inspired mainly by the sricakra,
the 2 0 th -c en tu ry Indian artist N irad M a ju m d a r created his ink
draw ing Yantra .h The contem porary dancer C handralekha acknow
2 S o m e authorities do not rccognize these com positions as poetry. F or an e x h a u s
tive tre a tm e n t o f this topic, s e e R u d ra d e v T r i p a t h i s stu dy, S a m sk rt-sa h ity a m c m
sa b d a la h k a r (Di111: SrllalbahaduraSastri K cndrlya S am skrt V idyapith, 1972 [in H in
di]).
.
.
3 T h is yantra is rcp ro d u ccd in B u n c c 2001: 53, w ho labels it erro n eo u sly A m ba
M atta Y a n t r a in stead o f A m b a M ata Y a n t r a . It is also k n o w n as bisonyantra
( P r a n a v a n a n d a < 1 9 7 7 > : 52), w h ile C h a w d h ri 1992: 53, 2 0 2 -2 1 1 c lassifies it as
bccsiyanlra.
4 F o r c o n te m p o ra ry yantra w o rsh ip in G u jarat, see the d iscu ssio n in P a d m a ja
1985,
5 F o r pictures o f the V id h an B havan, see Khan 1987: 1 3 4 -1 3 9 ; for the J a w a h a r
K ala K endra, see K han 1987: 142 -1 4 3 and for the Surya Kund, see Khan 1987: 105,
159.
6 N ira d M a j u m d a r s Y a n tra is rc p ro d u c c d in C h a k r a v o r ty S p iv ak 1999: 193,
Figure 2. N u m e ro u s m odern m an d alas have been creatcd by both Asian and W estern
artists: see, for ex a m p le , the oil pain tin g by the Nepali artist Sharda Man Shrestha
( r e p ro d u c e d in S ingh 2 000: 85, Plate X I) and the m a n d a la s b y the G e rm a n artist
L ore B ert (reproduced in Singh 2000: 87, Plate XII).
I NTRODUCTI ON
ledges the influence o f the Saundaryalahari attributed to Sam kara on

her dance piece Yantra: Dance D ia g ra m s, a w o rk in w hich g e o
metrical figures are created by dancers.
S o m e Problem s
W hile a body o f literatu re is grow ing in w h ic h m a n d a la -lik e
structures o f different cultures are com pared with one another and
their use in therapy is explored, not m uch solid research has been
done on mandalas in the Hindu traditions, and indeed no systematic
study has as yet emerged. Descriptions o f mandalas in ancient texts
are barely studied, and usually left untranslated. Descriptions o f them
in popular books often appear to be confused, since m any authors
apply the same terminology to what appear to be som ew hat sim ilar
structures without differentiating between traditions. Psychoanalysts
and psychologists endeavour to interpret the m andala by applying
their own categories, T hese approaches are o f lim ited value for an
understanding o f the structures and functions o f m an d alas in the
context o f South Asian traditions. Since m andalas are not objects o f
art p e r se but are em bedded in a ritual context, a purely art-historical
approach to the subject will not do justice to them either.
T hanks to advances in the study o f T antric texts over the past
decades and the increased availability o f objects from South Asia,
new m aterials have b e c o m e available w h ich put us in a b ette r
position than previous scholars to carry out research on m andalas
and yantras. But m useum s are usually not the places to look for
m andalas and yantras, since the latter are ritual rath er than art
objects, and so ex e c u te d by c ra fts m e n ra th e r th a n artists. An
exception is the collection o f about 60 copper yantras from Bengal in
the M useum fiir indische Kunst, Berlin. The private collection o f
yantras and mandalas o f R obert Clark, Barcelona, is docum ented in
Stadtner 1998.
D rawings o f yantras are often found in South A sian m anuscripts
and printed books dealing with magical and Tantric rituals, and in art
catalogues as w e ll.7 The yantra designs found in these sources are
7
See, fo r e x a m p le , S o t h e b y s L o n d o n : C a ta lo g u e o f Islam ic, In d ian , T ib e ta n ,
N e p a le s e and S o u th -E a s t A s ia n D e c o r a tiv e and O t h e r W o r k s o f A rt, also A n t i
quities. D a y s o f Sale: M o n d a y , 16th F e b ru a ry 1981, 2 p m , T u e s d a y , 17th F e b ru a ry
1981, 10.30 am and 2 pm.
GUDRUN B UHNEMANN
often repetitive. One problem is the authenticity o f mandala and

yantra designs. Yantras are frequently executed on copperplates as
ordered by a practitioner. They are copied from draw ings in
manuscripts, sketchbooks (Nepal) or printed books. Their structures
and the m antras inscribed in them often contain errors that go
unnoticed due to the ignorance o f craftsmen, copyists and practi
tioners. Pranavananda <1977>: 7 5 -7 9 examines nearly 200 tiricakras
from various parts o f India and concludes that most o f them show
major or m inor flaws in their designs or other irregularities, and so
do not tally with the descriptions in ancient texts. According to this
author (Pranavananda <1977>: 4, 109), certain changes were made to
the structure o f the sricakra early on and these errors have been
perpetuated blindly by tradition. Artists in popular tourist spots in
Rajasthan and N epal paint m ostly for the tourist industry. They
freely mix elements from different traditions and copy designs from
books and m useum catalogues printed in the West. Their products
often do not represent a continuation o f ancient traditions. During a
recent visit to Bhaktapur in Nepal I interviewed a painter about the
use o f the sarvatobhadras and lingatobhadras in his country. He had
not seen these m andalas, and indeed eagerly pho to co p ied my
diagrams. I w ould not be surprised if painted bhadramandalas, are
soon being sold in the shops o f Bhaktapur as traditional Nepalese
m andalas. C u sto m ers will then use them as wall decorations,
although such mandalas were never intended to be hung on the wall
but were traditionally prepared on the ground from powders or grains
as supports for deities invoked into them. The Indigo Gallery in
Kathmandu was already recently displaying a painted lingatobhadra
which, along with another mandala, had been copied from M adhu
K hannas book Yantra: The Tantric Symbol o f Cosmic Unity.
Previous Scholarship on H indu M andalas and Yantras
A m ong the early studies o f m andalas and yantras the works o f H.
Zim m er and P.H. Pott should be mentioned. Based on works by Sir
J. W oodroffe (alias A. Avalon) (1865-1936) and his collaborators,8
H. Z im m er (1890-1943) published his influential book Kunstform
und Y oga im indischen K ultbild in 1926. The work contains two
BF or rcccnt research on J. W o o d ro ffe and his learn o f collaborators, sec T a y lo r
2001: 203ff.
IN T R O D U C T IO N
large sections, dealing w ith mandalas and yantras, w hich influenced

C.G, Jung (1875-1961), the originator o f analytical psychology, in
his interpretation o f the m an d ala .9 Z im m e rs book, w h ich did not
target an academic readership, endeavours to interpret m andalas and
yantras based on both H indu and Buddhist texts and m onum ents.
Z im m er argues that icons o f deities or figurative sacred im a g e s
(pratim a) can be subsumed under the category yantra, and in fact are
essentially and fu n ctio n a lly identical w ith yantras, cakras and
mandalas (1984: 28-29). P.H. Potts Y oga and Y antra (1946) takes
a different approach. Even though he recurs to his predecessors J.
W oodroffe and H, Zim m er, P o tts goal is to explain the function
yantras have within the context o f Tantric Yoga. Like Zim m er, Pott
refers to both B uddhist and H indu texts th roughout his work. A
classic work is G. T u c c is The Theory and Practice o f the Mandala,
W ith S p ecia l R e fe re n c e to the M odern P s y c h o lo g y o f the S u b
c o n sc io u s, published in 1949 in Italian but translated into English
only in 1961. The b o o k s main em phasis is on the sym bolism o f
Buddhist m andalas, although the sricakra and H indu parallels are
considered.
Comparatively recent publications for a general readership include
the book on yantras by M. K hanna entitled Yantra: The Tantric
Sym bol o f C osm ic U n ity , published in 1979, and S.K.R. R a o s
small w o rk The Y a n tra s , w hich ap p eared in 1988. R ao also
authored a tw o -v o lu m e b o o k on m andalas entitled M andalas in
Temple W orship (1988-1990).
In 1986 A. P adoux edited M antras et diagram m es rituels dans
l hindouism e, w hich is a collection o f scholarly articles on mantras,
m an d alas and y a n tra s em p lo y e d in the H in d u tra d itio n s. The
contributions are based on lectures presented at the co nference
L H indouism e-textes, doctrines, pratiq u e s o f the research team
no. 249 o f the Centre national de la recherche scientifique (CN RS)
organized by A. Padoux in Paris in June 1984. These lectures some
o f w hich have been abbreviated or thoroughly revised are p u b li
shed along with a sum m ary o f the discussion that follow ed their
presentation. 10
9 C.G. J u n g s r e m a r k s and ob serv atio n s on the sy m b o lis m o f th e m a n d a la ap p e a r
in se v e ra l sectio n s o f his C o lle c te d W orks; see, for e x a m p le , J un g 1950 a n d J u n g
1964.
10 P a d o u x s v o lu m e c o n ta in s th e f o llo w in g p a p e r s th a t sp e c if ic a lly f o c u s on
m a n d a la s and yantras: M an d a la e t y a n tra d a n s lc SivaTsme a g a m iq u e . D e fin itio n ,
G UDRUN BUHNEMANN
Several authors have studied individual mandalas and yantras.

Kramrisch 1946, volum e 1: 46-63 was the first scholar to analyze
two main types o f the v a s tu p u r u s a m a n d a la a mandala employed in
the construction o f buildings. She was followed by others,12 inclu
ding A pte/Supekar 1983 and Apte 1986 and 1987. Apte also con
ducts research on mandalas in the Pancaratra tradition. Apte 1973
focuses on m andalas described in the Jayakhya-Sam hita. In the
introduction to his edition and translation o f the Pauskara-Samhita
(Part 1, 1991), the same author analyzes one o f four sels o f mandalas
described in this Samhita. This set comprises 25 mandalas, which
Apte attempts to reconstruct.1'1 The well-known Sricakra, which is
em ployed in the ritual worship o f Tripurasiindari, is the subject o f
several publications.14
B u n ces recent volum e on yantras (2001) examines the relation
ship between numbers and yantras, The book is based on secondary
materials, with Johari 1986 as one major source.
d escrip tio n , u s a g e (H. B ru n n e r) [the revised and e n larg ed version o f this article
appears in this v o lu m e in English translation]; Q u elq u cs rc m a rq u c s sur l usag e du
m a n d ala ct du yantra dans la vallec de K athm andu, N e p a l (A. Vcrgati); Panjara ct
yantra: le d ia g ra m rn c de 1im ag e s a c r c e (B. Biiumcr); Dc Ic ffic ic n c c p sy c h a g o g iq u c dcs m a n tr a s c i dcs yantras' (F, Clienet); La vision dc la divinitc d a n s les
d ia g ra m m c s selon 1c v ish n o u ism c vaikhfm asa (G. Colas); Lcs d ia g n im m c s co sm o g o niqucs scion le S vacehandatantra: Perspectives p h ilo s o p h iq u c s (C. Conio); M a n
tra et yantra cn m e d c c in e cl a lch im ic in d ie n n e s (A. R o?u); Lc SrT-cakra d a n s la
S a u n d a r y a - L a h a r i (T. M ichacl); M a n d a la and A g a m ic Identity in the T rik a o f
K a s h m ir (A. Sanderson).
11 For an a ssessm en t o f K r a m r is c h s w ork on the vastupurusam andala, see Bafna
2000: 3 0 - 3 1 .
12 See, for exam ple, Kulkarni 1979; see also the interpretation o f the vastupiirusam a n d a la given in D a n ic lo u 1977: 2 8 - 3 5 (2001: 3 9 - 4 1 ) . T h e re la tio n sh ip o f the
vastupurusam andala to a rc h ite c tu re is the topic o f M c is tc r s co n trib u tio n to this
volume.
13 A p t e s set o f re c o n stru c te d m an d alas is re p ro d u ced in c o lo u r in: Prakrti: T h e
Integral V isio n ( V o lu m e 3: T h e A g a m ic T ra d itio n and the Arts, e d ite d by B,
Biiumcr, N e w Delhi: D.K. Printw orld (P) Ltd., 1995): 193+, Illustrations P.P. Apte
1 .1 - 1 .2 5 , H o w e v e r, d u e to an error, the plates arc a p p en d ed to a n o th er article by
Apte included in this volum e. Black and white d raw ings o f the sam e set o f m a n d a la s
arc in clu d ed in A p t c s edition and translation o f the P a u s k a ra -S a m h ita (P art 1),
appendix, pp. i-x x i.
14 See, for ex a m p le , R ao 1 9 1 4 -1 9 1 6 , volu m e 1: 3 3 0 - 3 3 2 , Z im m e r (19 2 6 ) 1984:
1 5 8 -1 8 0 , Pott (19 4 6 ) 1966: 4 0 - 4 4 , B o lto n /M a c lc o d 1977, P ra n a v a n a n d a < 1 9 7 7 > ,
K u laich ev 1984, M ichacl 1986, F o n se c a 1986, K hanna 1986, K u la icliev /R am en d ic
1989, R a o 1990, B rooks 1992: 115-146, 189-199, Rao 1998 and W ilke 2003.
The Scope o f this B o o k

My interest in mandalas goes back to a period in the 1980s w hen I
conducted research in Pune, M aharastra. The plan to publish a book
on mandalas and yantras in the Hindu traditions took shape over time
as I observed the grow ing popular interest in T ib etan B u d d h ist
mandalas. Unlike the many Tibetan mandalas which include pictorial
representations o f multiple deities, m ost published m andalas in the
Hindu traditions appear to be sim pler and m ore abstract in design.
H ow ever, H indu m andalas, especially from N epal and Rajasthan,
often include painted images o f deities.15 Com plex m andalas are also
described in texts, and the practitioner is instructed to visualize
multiple deities in the mandalas, although these deities may not be
represented. This volum e reproduces several m andala designs, some
o f which have been reconstructed from texts. Since texts often do not
specify all details o f the mandalas, such reconstructions necessarily
remain tentative.
W ith the exception o f the gricakra, w h ich has attracted c o n
siderable interest, adequate attention has not been devoted to m an d a
las and yantras in the Hindu traditions and their multiple uses. Unlike
the approaches o f earlier books, w hich indiscrim inately deal w ith
Buddhist and H indu m andalas and which often arrive at generalized
conclusions, this book attempts to clarify im portant aspects o f m a n
dalas and yantras in specific Hindu traditions through investigations
by specialists. In the present state o f research it is best to avoid
generalizations and broad com parisons across traditions that rarely
take into account existing differences, and often turn out on closer
examination to be inaccurate. The com plex B uddhist m andalas for
their part merit a separate study. N evertheless I hope that this book
will indirectly contribute to a better understanding o f the m andala in
other South Asian traditions, and will lay the foundation for future
inquiries.
T he essays in this book explore some aspects o f m andalas and
yantras in the Smarta, Pancaratra, Saiva and &akta traditions. An
essay on the vastupurusamandala and its relationship to architecture
is also included. It would have been useful to have essays on the use
15 In Nepal, even the sricakra o ccasio n ally includes p a in ted im a g e s o f deities; see,
for exam ple, Illustration 43 in K re ijg e r 1999: 1 1 0 -1 1 1 . F o r a 19 th -c e n tu ry m a n d a la
fro m R ajasth an w ith icons o f goddesses, sec, for e x am p le, Pal 1997: 215, 337.
GUDRUN BUHNEMANN
o f yantras in Indian medical systems, astrology or folk traditions, or

on geographical space as a mandala. It was, however, not possible to
find qualified authors who could write these essays within the given
time frame. Thus this book is a contribution to the study o f an area o f
South Asian culture which has hardly been researched, but it is not
an exhaustive treatment. This would have been an unrealistic goal,
given the extant mass of material on the topic.
In secondary sources, mandalas (and yantras) have been described
too uniform ly as aids to meditation or visualization.16 W hile they
certainly function as meditational devices in some traditions (as, for
instance, the sricakra frequently does), this use o f mandalas is but
one aspect o f a larger picture. In this regard I I. Brunners paper in
this volum e is significant, since she sets out to exam ine some
popular notions about m andalas critically and to em phasize other
uses o f m andalas in ritual. In architecture, the notion o f an allgoverning m a n d a la o f symbolically significant dim ensions which
underlies all buildings is frequently met with in the literature, and
has recently been challenged by Bafna 2000: 4 2 - 4 3 ,17
The first essay in this book is designed as an introduction to the
topic. Referring to H. Brunner and others, I discuss the meanings o f
m andala, yantra and cakra, and suggest distinctions am ong these
terms. This is followed by a treatm ent o f different categories o f
mandalas, yantras and cakras and their constituent parts.
In the next essay, I focus on mandala-like structures and actual
mandalas and yantras currently em ployed in the ritual practice in
M aharastra. In its first part, I discuss mandala-like arrangem ents,
such as the baljliaranacakra and pancayatana shrines, along with the
navagrahamandala as an exam ple o f a mandala with a lotus design.
A description o f two yantra structures follows. Together with the
previous essay, this section is intended to introduce the reader to
basic concepts and m andala designs in the Hindu traditions. The
second part o f the essay focuses on a specific category o f mandala
called bhadramandalas. These are square-shaped mandalas employed
mainly in concluding ceremonies o f religious observances (vrata).
16 For a critical exam in atio n o f the claim that m an d alas in S hingon B u d d h ism arc
aids or s u p p o r ts for visualization practiccs, see S h a r f 2 0 0 1.
17 See M. M e i s t c r s p a p e r in this v o lu m e for a critical a s s e s s m e n t o f B a f n a s
position.
I NTRODUCTI ON
M arion R astellis essay focuses on the use o f m andalas and y a n

tras in the V aisnava Pancaratra tradition as based on original p as
sages from the Samhitas, It describes the selection, purification and
ritual acquisition o f the m andala site, guidelines and m aterials used
for drawing mandalas, and the types o f m andalas found in the texts.
She then discusses the multiple functions o f mandalas in Pancaratra
rituals, The choice o f a m andala for a rite is guided by the desire to
achieve specific results. It depends on the suitability o f a mandala for
a certain rite and the main deity worshipped in it. The use o f m a n
dalas in initiations (diksa) is treated elaborately. Some details o f the
ritual, such as the casting o f a flower onto a m andala by the blind
folded initiand, have parallels in Buddhist Tantric initiation rituals.18
The deity is made to be present in a mandala by imposing the deity s
mantras on the m andala structure. Tw o im portant m andalas in the
Pancaratra tradition are the cakrabjamandala and the navapadm amandala. The Pancaratra Samhitas consider the m andala a represen
tation o f the deity s body, and o f the universe as well. A ccording to
some Sam hitas, em an cip atio n is only possible th ro u g h m andala
worship. Rastelli further discusses the significance o f yantras in the
Pan caratra tradition. She focuses esp ecially on the saudarsanayantra'9 which is considered so pow erful that the person w ho w ears
it requires another yantra, the yantra o f the w e a re r (dharakayantra),
to keep its pow er in check. As in the case o f m andalas, the material
from which yantras are made is considered essential for the efficacy
o f the rite. D ifferent m aterials are believ ed to produce different
results, (The texts o f the Saiva Siddhanta that Brunner exam ines em
phasize the varying efficacy o f the materials from w hich m andalas
are constructed, from precious stones on downwards.)
T h e three fo llo w in g essay s deal w ith aspects o f the Saiva
traditions. Helene B ru n n er has been researching S aiv ag am as for
more than thirty years. M ost o f her w ork is w ritten in French and
therefore accessible to a more limited readership. F o r this volum e,
her French paper, originally published in P a d o u x s edited volum e,
M antras et diagram m es rituels dans 1h indouism e, (1986: 11-35),
was translated into English by R. Prevereau, M .A ,, and com pletely
T h is topic h as b e e n d ealt w ith rep e a te d ly ; fo r a d e s c rip tio n o f th e d i s c i p l e s
e n tra n c e into the m a n d a la and his casting o f a flow er, see, for e x a m p le , W a y m a n
1974.
19 T h e saudarsanayantra a p p e a r s to b e identical w ith th e sudarsanayantra (se e
section 2.3.3 o f the follow ing essay).
10
GUDRUN B U HN EM ANN
revised and enlarged, B runners essay is divided into two parts. The
first part attempts to clarify the m eaning and use o f the term s
m andala, yantra and cakra. Her classification o f different types o f
m andalas based on their ritual application is o f special interest. (I
have taken up B r u n n e rs discussion o f the different types o f
mandalas in the following essay [section 1.2], as has Torzsok in her
own.) The second part describes the use o f mandalas in the ritual
w orship o f Siva. B runner reconstructs the sarvatobhadramandaki
described in chapter 3 o f the Saradatilaka, which is used in an
initiation ( diksa) ritual, and analyzes its structure in detail. Finally,
she discusses the signi ficance o f mandalas in the Siddhanta School.
Judit Torzsok examines pre-1 Ith-century Saiva mandalas as icons
which express a relationship between certain branches o f Saivism
and betw een Saiva and non-Saiva groups. In the first part o f her
paper she deals with the uses o f the terms mandala and cakra, a topic
also taken up by Brunner. This leads into a discussion on how the
circles (cakra) o f deities are present in a mandala. T orzsok then
focuses on tw o kinds o f m andalas: mandalas used in initiations
( diksa) and mandalas (and yantras) for the acquisition o f supernatural
powers ( siddhi). Giving examples from the Svacchandatantra, she
shows how m andalas can visually represent doctrines o f other Saiva
groups and teachings o f non-Saivas. Torzsok. specifies three major
strategies (specialization, expansion and substitution) which are
em ployed to ad ap t m andalas to a specific purpose, such as the
acquisition o f supernatural pow ers, In the A ppendices, T orzsok
attempts to reconstruct four mandalas from textual descriptions. The
reconstruction o f two mandalas (see Colour Plates 18-19) is tentative
and does not show the outer boundaries that are characteristic o f
mandala designs. These boundaries are not specifically mentioned in
the texts, but are likely to have been assumed.
Andre P ad o u x s first essay in this volum e examines descriptions
o f mandalas and their use in Abhinavaguptas Tantraloka (early 11th
century). Basing him self mainly on material from various sections o f
the text, Padoux portrays the uses o f such mandalas as the tri$ulabjamandala and trilris u la bja ina n da hr in rites, including the different
forms o f the initiation ( diksa) and the practitioners daily ritual w or
ship, in which the mandala is visualized as being present in his body.
Andre P adouxs second essay deals with the sricakra as described
in the first chapter o f the (most likely) 1'lth-century Yoginlhrdaya.
11
This chapter offers a description o f the d e s c e n t (avatara) o f the

sricakra as a cosm ic process and m anifestation o f divine power,
which the practitioner visualizes and experiences in his body. The
cakra is portrayed here as a cosm ic rather than a ritual diagram ,
whose contemplation has a visual/spatial as well as a phonic/mantric
dim ension and leads to an identification o f the Y ogin w ith the
supreme level o f the word (vac).
M ichael W . M eister measured a large num ber o f ancient temples
in the course o f extensive research in India. His drawings o f groundplans o f tem ples show h o w the vastupurusamandala was used in
practice. M eisters contribution to this volum e is concerned with the
vastum andala as described in V araham ihiras Brhat-Samhita and its
application in temple architecture.
This book contains only one bibliography, in order to avoid
repetition o f references and to allow the interested reader to find
relevant literature on mandalas, yantras and cakras in one place.
The title o f this volum e contains the m uch-debated w ord Hindu,
which has been the focus o f some controversy. I will not discuss the
p ro b le m s asso ciated w ith this term h e r e .20 I h av e d ecid ed , for
pragm atic reasons, to use it rather than choices such as Brahm anical, a word which would indicate to some that the subject matter is
c o n c e r n e d only w ith th e B ra h m in c o m m u n ity . T h e eq u a lly
problematic terms Tantrism and Tantric21 are also used in this book
for practical reasons and without further discussion.
R em arks on the Transliteration
It is difficult to avoid inconsistencies w hen transliterating words
from different Indian languages. For the names o f m any places and
temples, popular transliterations are already in circulation w hich may
not conform to scholarly standards. I have in m any instances retained
the popular transliteration o f such words in order to avoid burdening
the reader with unusual spellings o f nam es. The transliteration o f
words from Nevari poses its own problems, since there is often more
than one current spelling o f a word. I am aware o f minor inconsisten-
211 For a d isc u ssio n o f the p ro b le m s a sso c ia te d w ith the term H in d u is m , see, for
instance, Smith 1987.
21 P a d o u x 1987b, V e ra rd i 1994: 5 2 - 5 3 and U r b a n 1999, a m o n g o th e rs, h a v e
discussed these p ro b lem atic terms.
12
G UDRUN BUHNEMANN
cies in spellings o f words from Indian languages used by the diffe

rent authors, and also their divergent treatment o f parentheses. It is
difficult to avoid such inconsistencies without interfering loo much
with the style of the individual contributions.
A cknow lcdgcm en is
M y research for the volum e extended over a prolonged period o f
time. 1 would like to thank the University o f W isconsin-Madison and
the Lumbini International Research Institute for support at different
stages o f my research. I am indebted to Professor K.S. A rjunw adkar
and Dr. R.P, Goswami for valuable suggestions on earlier drafts o f
my chapters in this book. 1 am also grateful to G. M evissen, M .A.,
for some bibliographical references; to R.S. Green, M .A., M .W .
Dennis, M.A., and S. W eier for help with computer-related issues; to
T. Cowall and P. Radder o f E.J. Brill publishers for editorial support,
and to Professor J. Bronkhorst for including this volume in B r ills
Indological Library. I would like to express my thanks to P, Pierce,
M .A ., for editing my contributions for style. Finally 1 thank R.
Prevereau, M .A., for preparing the translation o f 11. B runners paper
from the French.
M A N D A L A , Y A N T R A A N D C A K R A : S O M E O B S E R V A T IO N S
Gudrun B uhnem ann

1 M andala
1.1 The Term Mandala
In its most general use, the word mandala refers to something that is
round or circular, such as a ring or circle, further, a region, terrestrial
division, dom ain, assem bly or a g ro u p .1 The term is used in Kautily as Arthasastra, book 6 , in the sense o f a spatial configuration o f
neighbouring states from the view point o f a king. In Tantric tra
ditions, the term mandala often refers to a space with a special struc
ture that is enclosed and delimited by a circumferential line and into
which a deity or deities are invited by means o f mantras. This space
is often a circle, but m ay also appear as a square, a triangle or
another shape.2 The various shapes and structures o f m andalas are
based on the traditions o f the.different schools, ritual applications,
the deities worshipped and the practitioners qualification and goal.
M andalas themselves are prepared from various materials, including
1 F o r a d is c u s s io n o f the u n c e rta in e ty m o lo g y o f th e w ord m a n d a la , see M ayrh o fer 1 9 8 6 - 2 0 0 1 , v o lu m e 2: 294, A re lig io u s e ty m o lo g y o f the w o r d a p p e a rs in

K ularnava-T antra 17.59:
m angalatvac ca dakinya yoginiganasarnsrayat /

lalitatvac ca dcvesi m andalam parikirtitam l/
O m istress o f the gods, it is called m a n d a la be c a u se it is a u s p ic io u s (nrangalatva),
b e c a u se it is the a b o d e o f the g ro u p o f Y o g in is o f the D akini, and b e c a u s e o f (its)
beauty (lalitatvi i),
F o r an e t y m o l o g y o f th e w o r d m a n d a la , w h ic h d iv id e s t h e w o r d in to th e
c o m p o n e n ts manda ( e x p lain ed as sara [essence]) and la (fro m th e v e rb a l ro o t la [to
take]), see T an tra lo k a 37.21 w ith J a y a r a th a s c o m m e n ta ry , re fe rred to in P a d o u x , p.
227); B u d d h is t texts also divide the w o rd m a n d a la into these tw o c o m p o n e n ts , b u t
different in te rp re ta tio n s are giv en to th e m ; cf. the d is c u s s io n s in W a y m a n 1999,
L e ssin g /W a y m a n 1978: 270, note 1, T o g a n o o 1971: 1 5 0 -1 6 0 , R a m b e lli 1991: 9 - 1 3
and T rib e 1994: 127.
2 B runner, p. 157, note 5 and T o rzso k , p . 208 also refer to sem i-circu lar m a n d a la s
and m andalas having the (triangular) shape o f a vulva ( y o n i ), am o n g others.
14
G U D RU N
BUHNEM ANN
coloured powders, precious stones, fruits and leaves, and fragrant

substances.3 It must be emphasized, however, that the mandala is not
merely a physical structure with a specific design. A mandala is the
place in which the practitioner beholds the deities who have been
invoked into it and so have become an integral part o f the structure.4
M andalas figure among the places into which deities can be invoked.
These include statues, vessels and lire.5
M andalas are required in occasional (not daily) rituals, such as
festivals or religious observances (vrntii) and more im portantly
Tantric initiation (diksa) rites, in which latter the viewing o f the
m andala is an essential clem ent/' At the time o f initiation the m an
dala structure functions as a place in which the deities b ccom e
visible to the initiate for the first time, thereby confirm ing the
in itia te s new identity (Torzsok, pp. 183 -1 8 4 , 189, 190). T he
m a n d a la stru c tu re can function as an im p o rtan t d e v ic e for
representing the pantheon o f deities in a system or school, and
expressing the hierarchy o f deities within the system. This hierarchy
can even include deities o f o ther system s as pari o f a low er
rev elatio n , and can indicate a cosmic order as well (Torzsok, p.
196). Further, mandalas, like yantras, are used in rituals leading to
the attainment o f supernatura l powers ( siddhi).7
While most mandalas follow the common pattern o f a concentric
arrangem ent o f deities in order to express a hierarchy, the trident
m andala o f the Trika also features a vertical ascent. The m andalas
trident is seen as rising three-dimensionally from a central lotus, as if
com ing out o f the m andalas surface (Torzsok, p. 196). We do not
know w hether three-dimensional m andalas were actually construc
ted. Such m andalas are known from Buddhist texts and traditions.
The V isnu-Sam hita (cf. Rastelli, p, 123) instructs the practitioner to
make the lines o f a mandala in varying thicknesses, with the centre
3 F o r m a te r ia ls listed in th e P a n c a r a lr a SamhitFis, se e R astelli, p. 123; for
m a n d a la s m ad e from fragrant substances (gandhainandalu), sec Padoux, p. 226.
4 See the discussion in T o rzso k , pp. 1 8 3 -1 8 4 for m ore details.
5 Rastelli, p. 126 discusscs the w o rsh ip o f the deity in four pla c e s ( catuhstlulna )
attested to in the y o u n g e r P ancaralra Sam hitas. T h e s e placcs arc a m andala, a vessel,
fire and a statue. T o rzso k , p. 193, note 60, quoting the T antraloka, lists 11 sup p o rts
o f external worship, including a rosary, m anuscript and mirror.
6 Cf. R a ste lli, pp. 13 0 f f ., T o rz s o k , pp. 185ff. and P a d o u x , pp. 227ff. for this
aspect o f m andalas.
1
See T o rz so k , pp. 2 0 1 - 2 0 9 for a d escription o f such use o f m a n d a la s in early
Saiva Tantras.
M ANDALA, YANTRA AND CAKRA
15
o f the m andala its m ost elevated part, w hich could be taken to

presuppose the concept o f three-dim ensionality. Three-dim ensional
yantras are not uncom m on in the Hindu traditions, and are described
below in section 2 . 1.
D ifferent theological interpretations have been applied to m a n
dalas, the structural parts and deities being correlated with doctrines
o f different systems. Interpretations are extremely varied, and even
one text may provide more than one interpretation o f the parts o f a
mandala.
Patterns exhibited by m andalas have had w idespread influence.
M andala patterns o f cities have frequently been d escrib ed .8 H o w
ever, it often remains unclear w hat the connection betw een a m a n
dala and a city or tem ple really means, as B afna 2000: 26 notes.
P roblem s arise w hen one attem pts to correlate m andala structures
and actual building plans. Gutschow 1982: 179, 185 argues that con
temporary drawings o f mandalas o f cities, such as the m andala o f the
city o f B haktapur in Nepal, usually do not reflect ancient guidelines
for tow n-planning b u t rather represent a specific interpretation o f
existing urban conditions. A m andala pattern is thus projected onto
the city by establishing connections betw een already existing b u il
dings. These connections may not be immediately intelligible to the
outside observer, and are indeed open to interpretation.
The terms cakra and yantra are sometimes used as synonyms for
mandala, and all three terms are often translated indiscriminately as
(mystical) d iag ram s. The fact that the geom etric designs o f m a n
dalas, yantras and cakras are similar contributes to confusion am ong
the three. N ot only W estern authors confuse the terms, even later
Sanskrit texts often use m a n d a la and y a n tra rather loosely as
synonyms. Occasionally metrical considerations and constraints may
have played a role in the choice o f a word, as when a text uses the
word pura (c ity ), for example, as a synonym for m a n d a la.9 O ther
s See, fo r e x am p le, G u ts c h o w /K o lv e r 1975, w h e re the a u th o rs describe the layout

o f the city o f B h a k t a p u r in N e p a l; s e e also Z a n e n 1986: 1 4 8 - 1 5 0 , relatin g to th e
N e v a r to w n S a n k h u , F o r a c ritical a p p ro a c h to a m a n d a l a as a c o n c e p t said to
u n d e rlie to w n - p la n n in g , see R o y 1977, w h o d is c u s s e s the la y o u t o f J a ip u r, and
T illotson 1987: 8 1 - 8 3 , w h o fo cu ses on the p alaces o f B u n d e lk h a n d .
9
F o r th e u s e o f the w o r d pura ( c i t y ) as a s y n o n y m f o r m a n d a l a , s e e th e
d isc u ssio n a m o n g T. G o u d riaan , H. B r u n n e r and P. Filliozat r e p ro d u c e d in P a d o u x
1986: 32, and also Rastelli 2 0 0 0 b : 3 7 5 , n o te 57.
16
GUDRUN 1301-IN EM ANN
synonym s o f m andala found in the literature are yHgn,H) blmvana/

bhuvana, vesm an and, in a metaphoric sense, pitha."
Various definitions o f the term m andala have been proposed.
Kramrisch 1946, volum e 1: 11 defines m andala as a yantra when
she writes about the vastupurusamandala: The Vastupurusamandala,
the diagram o f the temple, is a Yantra.... Lieberl 1976: 168 does the
same when she defines the word as the name of a kind o f y a n tr a '2
Sim ilarly, R enou/F illiozat 1947-1953, volum e 1: 568 state that
yantras in w hich a more or less decorated circle predom inates are
called cakra or mandala. In addition, some authors assum e that
yantras are the counterparts o f mandalas in the Hindu traditions. This
erroneously implies that m andalas arc rarely part o f the H indu
traditions and that yantras are not found in the Buddhist traditions. 13
Thus Tucci (1949) 1961: 46 states: in Hinduism, however, yantras,
purely linear designs expressing the same principles, are usually
substituted for m andalas ...; 14 and Eliade 1969: 219 writes: The
simplest mandala is the yantra, employed by Hinduism....
Several scholars have attempted to establish semantic distinctions
am ong the three terms m andala, yantra and c a k ra .15 One approach
attempts to establish distinctions on the basis o f the structure and
constituent parts o f these objects. Rao 1914-1916, volum e 1: 330
states that a cakra is defined in the Tantras as a figure consisting o f
10 Rastclli, p. 1 19, note 1, reports that the P auskara-S am hitn uses the w ord y fig a
sy n o n y m o u sly w ith m andala.
11 F o r a dis cussion o f the term s b h a va m t/b h u viw a , vesm a n and p ith n a s s y n o n y m s
for m an d ala, sec T orzsok, p. 182.
12 A s im ila r s ta te m e n t is f o u n d in B e r n ie r 1979: 120: E v e ry m a n d a l a is
essentially a yantra...."
13 Bizot 1981 describes B u ddhist yantras in South-east Asia, especially C a m b o d ia
and T h a ila n d . Y an tra are also d e sc rib e d in B u d d h ist S a n sk rit texts, such as the
V im a la p ra b h a c o m m e n ta r y on the K alacakratantra, chapter 3 (V im a la p ra b h a tik a o f
Kalkin S rlp u n d a rik a on Srilagh u k alacak ratan lraraja by rimaiYjuriya&as, v o lu m e 2,
edited b y V. D w iv cd i/S ,S , B ahulkar, Sarnath, Varanasi: C entral Institute o f H ig h er
Tibetan Studies, 1994): 19, 15ff.
14 S ee also R en ou/Filliozat 1 9 4 7 -1 9 5 3 , volum e 1: 568 for a sim ilar statement,
15 T h e f o llo w in g sta te m e n t by H ocns (in G u p t a / H o c n s /G o u d r ia a n 1979: 113)
illustrates the co n fu sio n s u rro u n d in g the term s m andala and yantra: In the existing
literature y a n tr a and m a n d a la arc often c o n s id e re d to be s y n o n y m s . T h is is not
correct, b e c a u sc yantra in general m e a n s an instrument, an im plem ent, T h e yantra is
often th re e -d im e n sio n a l w h e re a s the m a n d ala a lw ay s is tw o -d im en sio n al. M an d ala
and y a n tra often h a v e the s a m e g e o m etrical form s, but the yantra m ay also h a v e
different forms..., T h e yantra is m ore w o rsh ip p ed than m editated upon. As far as the
aim s arc c o n e e rn c d o n e can say that the yantra is m o re used for w o rld ly p u rp o se s
than for liberation, w hereas the mandala is used for both purposes.
M A N D A L A , YANTRA AND CA KR A
17
angles and petal-like parts; that w hich consists o f angles alone is

called a yantra, It is unclear w hich text Rao cites here, but this state
ment can be identified in a quotation in the com m entary Saubhagyabhaskara by B haskararaya (18th century) on the Lalitasahasranam a,
p. 171, 4 - 7 . In this quotation the word angle (asra) is synonym ous
with the Sanskrit w ord corner (k o n a ) as used in the terms triangle
( trikona) or hex ag ram (satkona). The expression 'petal-like p a rts
renders the Sanskrit word patra. The above distinction betw een cakra
and yantra, how ever, appears to be purely theoretical and m ay be
applicable only in a specific tradition. It does not account for the
m any yantras w hich are com m only described as consisting o f petal
like parts. Z im m er (1926) 1984: 2 8 - 2 9 translates the three terms
cakra, m andala and yantra as circle-shape designs (cakra), rings h a p e d designs (mandala) and linear figures (yantra). It is not
explained, however, exactly w hat is meant by these terms and what
the differences between the circle-shaped and ring-shaped designs
would be. Gaeffke 1987: 155 notes that it has becom e customary to
call the simpler designs for daily worship yantras, and to reserve the
term mandala for the larger ones in public cerem onies w here the
w hole cosm os has to be present. A nother ap p ro ach attem pts to
establish distinctions betw een m andalas, yantras and cakras on the
basis o f the deities invoked into these objects. W o odroffe 1914,
volume 2: 285, note 13 makes a very generalized statement, w hich is
applicable only to few mandalas, when he asserts that the difference
between a M andala and a Yantra is that the former is used in the case
o f any Devata, whereas a Yantra is appropriate to a specific Devata
only. 16 The following formulation by Shankaranarayanan 1970: 9 is
a variation o f W o o d ro ffe s statement, and is equally problem atic:
The M andala is used in the case o f any deity w hile the C hakra is
specifically intended for a partic u la r d e ity . S h an k aran ara y an an
apparently replaced the word yantra in W o o d ro ffe s definition with
the word cakra. Schneider 1988: 100 attempts to make a distinction
betw een m andala and yantra on the basis o f the n u m b er o f deities
invoked. He suggests that a m andala represents the m icrocosm and
a c c o m m o d a te s a p a n th e o n o f deities w ho are p o sitio n e d in it
according to rank. A yantra, on the other hand, is the dom ain o f a
single deity, but may include that d e ity s retinue. This distinction
16
T h e sa m e, apparently w id esp rea d , d efin itio n a lso appears in Jhavery 1944: 7 1 ,
W o o d r o f fe 1956: 9 1 , note 2 and in K ane 1 9 6 8 - 1 9 7 7 , v o lu m e 5: 1135.
18
GUDRUNBUHNEMANN
appears to be based on a statement by Pott (1946) 1966: 71, w ho

describes a m andala as a co sm ic configuration in the centre o f
w hich is an im age or sym bolic substitute o f a pro m in en t god
surrounded by those o f a num ber o f deities o f lower rank ordered
hierarchically both among themselves and in relation to the c h ie f
figures, which configuration may be used as an aid to m editation and
in ritual as a receptacle for the g o d s. l ie adds thal a m andala is
distinguished from a yantra by a more graphic representat ion o f the
deities or o f their symbols and by a richer elaboration o f the details.
This last statement by Pott also takes the structure o f mandalas and
yantras into consideration and is somew hat more satisfactory than
the definitions o f his predecessors.
Yet another approach looks at the ritual use o f mandalas and y a n
tras. Thus Vergati 1986: 37, 4 4 -4 5 observes that mandalas are used
in secret as well as public ceremonies o f the Hindus and Buddhists in
N epal, w hereas the yantras, which always represent the goddess,
have more restricted uses. It has also been noted that mandalas are
usually objects for temporary ritual use. The deities are invoked into
them and dismissed at the end o f the ritual, after which the mandala
is dism antled. Yantras m ade o f perm anent materials into w hich a
deity has been invoked are usually kept in the temple or shrine for
continued worship (Sharma 1994: 423-424). It must be added, h o w
ever, that m any yantras are made for temporary use, like mandalas.
Rastelli, p. 144 notes yet another feature that sets m andalas and
yantras apart in the Pancaralra tradition. She states that mantras are
already inscribed on yantras at the time o f manufacturing the yantra.
The draw ing o f the lines o f the structure and the w riting o f the
mantras are a single process, which may indicate that a yantra repre
sents one integrated unit in which the deity is worshipped. Mandalas,
however, are constructed first and the deities are invoked into them
w ith m antras only later. It must be added, however, that later texts
enjoin that yantras be first prepared and then infused with life in a
special ritual, the pranapratistha, with the help o f mantras. It is not
possible to sum m arize all attempts at defining m an d ala, y a n tra
and cak ra in the literature. The use and functions o f these terms are
com plex and it will be im possible to arrive at a universally valid
definition. An in-depth study o f the use o f the terms in texts o f
different religious systems and time periods would be required to
M AN D AL A , Y AN TR A A N D CA KRA
19
determine how the terms have been em ployed by different authors

and how the use o f these terms has changed over time.
1.2 T ypes o f M andalas according to H. B runner
H. B ru n n ers contribution to this book describes uses o f the word
mandala based on her study o f p re -13th-century Saiva manuals. Even
th o u g h she confines h e r s e lf to an analysis o f the texts o f the
Siddhanta School o f Saivism , her observations on the use o f the
terms mandala, yantra and cakra appear to have a som ew hat w ider
application. B runner takes the term mandala to signify a limited, not
necessarily round, surface, and distinguishes four basic types o f
mandalas:
T y p e 1: L im ited surfaces w ithout a clear structure, w h ich are
com m only em ployed as seats for divinities, m en or objects during
ritual, such as m andalas o f cow-dung smeared on the ground. They
can be called seat-m andalas,
Type 2: Limited surfaces with geometrical designs prepared from
co lo u red po w d ers, w hich serve as su p p o rts for the regular or
occasional worship o f deities. These mandalas are for tem porary use,
being destroyed after the ritual. They are constructed in a ritual, with
close adherence to directional orientation. C om m only three, four or
five different colours are em ployed. T hese m andalas, often called
pow der m an d alas (rajom andala), may be large-sized and so allow
for the priest to enter through the doors and move around in streets.
A ccording to Brunner, such m andalas are tem porarily constructed
divine icons and can be called im a g e -m a n d a la s (the term is not
used in any texts).
Type 3: Limited surfaces divided into a certain num ber o f squares
or units called padas, dom ains into which divine or dem onic pow ers
are invoked to receive food offerings (ball). T h e ir co nstruction
usually does not involve the use o f colours. The best know n m andala
in this category is the vastum andala. B runner also includes in this
categ o ry g e o m etric al figures div id ed into b o x e s am o n g w h ich
objects are distributed. She refers to the mandalas in this category as
distributive diagram s.
Type 4. The term mandala is also used to designate the symbolic
shapes o f the five elem ents and the spheres/orbs o f the sun, m oon
20
G UDRUN BUHNEMANN
and fire. The shapes o f the elements are visualized, for example, in
the Tantric rite o f purification o f the elements ( bhutasuddhi) o f the
perform ers body. Since the shapes o f the elements and the spheres
o f the sun, moon and fire are neither concrete material objects nor
supports for w orship in the way that the previously discu ssed
m andalas are, they do not really fit the present context and are
therefore excluded from further discussion.
Concerning B runners first category o f mandalas, I would like to add
that in other traditions seat-m andala appears to be more com m only
used for ritual objects than for persons. The function o f these
m andalas is to protect ritual objects placed on them. Such supports,
made o f various materials, may feature simple geom etric patterns,
and can be referred to as mandalas or yantras for (establishing) a
foundation (stbapanayantra) (see 2.2 . 1).
T he nam e o f the second category, im a g e -m a n d a la , may be
som ew hat misleading, since it suggests the presence o f a pictorial
representation o f the deity in the m andala which is not intended.
W hat is meant is that the entire mandala is the principal support for
worship and is present as an image/icon for the duration o f the ritual.
These m andalas are also called pow der m an d alas (.rajomandala)
(but they may also be made from other materials, such as grains) and
can be characterized as supports into which deities axe invited in
order to receive worship.
B ru n n ers classification o f types o f mandalas and their ritual use
in the Saiva tradition is valuable. In all attempts at classification,
however, we need to be aware o f the fact that in both texts and ritual
practice the distinction am ong the types o f m andalas is not always
that clear. A n y classification can therefore only be o f limited
practical value, and is often applicable only within one particular
system.
1.3 S o m e Structural E lem ents o f Mandalas
M andalas display different shapes and patterns, and are made up o f
various constituent parts, depending on the tradition they come from.
In the follow ing I will describe two basic structural elem ents o f
mandalas, the lotus design and square grid. In the next essay, I will
provide concrete exam ples o f these structures from the Smarta tra-
M AN D ALA, Y ANTRA AND CAKRA
21
dition o f M aharastra. G eom etric figures like the triangle and hex a
gram, which occasionally also appear in mandalas, will be described
in section 2.3 in connection with yantras. In the following I will look
at m andala patterns o f different periods and traditions sim ultane
ously, without attempting to treat the topic historically.
1.3.1 Lotus Designs
Lotus designs appear com m only in Indian art as well as in mandalas
and in yantras. The lotus is a co m m o n South A sian sym bol o f
creation, purity, transcendence and the sphere o f the absolute, 17 but is
especially know n as a symbol o f the female reproductive organ. It
has also been connected with water symbolism since ancient times,
as already indicated by a statement in Satapatha-Brahm ana 7 .4 .1.8 :
The lotus is the w aters. Indeed, in descriptions o f the symbolic
shapes (mandala) o f the elem ents ( bhuta)]S the lotus represents the
element w ater.19
In m andalas and yantras o f lotus design, the central deity is
positioned in the pericarp (karnika), and the em anations or subordi
nate deities in the p e ta ls .20 A lotus design m ay have one ring or
several concentric rings o f petals. T h e petals o f an eight-petalled
lotus ideally point in the cardinal and intermediate directions, but we
find numerous specimens in books and coins in w hich it is the spaces
betw een two petals that are oriented to the points o f the compass.
This orientation m ay be due to the ignorance o f the craftsm en who
prepared the yantras. Bunce 2001: 28 explains that this latter orien
tation signifies pow er and the feminine element, but I am doubtful
w hether it is described in ancient texts. The eight-petalled lotus
whose petals do the pointing is a shape w hich is well suited for
positioning deities in their respective directions. This purpose is not
served when two petals point in each o f the cardinal directions and
none in the interm ediate directions. The relationship betw een direc
tions and lotus petals is borne out by a statem ent in M aitrayaniya17 For a recent and detailed d is c u ss io n o f the s y m b o l is m o f the lotus, se e Garzilli
2 0 0 0 ; for the lotus m o tiv e in architecture, s e e G u ts c h o w 1997: 2 4 8 ff.
18 T h e s y m b o l i c sh a p e s o f the e l e m e n t s are c l a s s if ie d as m a n d a la s o f ty p e 4 in
B runners a f o r e m e n t io n e d c la s s ific a t io n o f m andalas (s e e s e ctio n 1.2).
19 S e e , for e x a m p l e , S a ra d a tila ka 1 . 2 3 - 2 4 , w h e r e a lo t u s w ith a h a l f m o o n
represents water.
a) For a description o f the co n struction o f the various sh a p e s o f petals, se e B u n c e
2001:26.
22
GU DRU N
BOIINEM ANN
U panisad 6.2 which identifies the lotus (of the heart) with space
( akasa), and its eight petals with the four cardinal and interm ediate
directions. Eight-petal led lotus designs com m only appear in the
centre o f Buddhist mandalas, such as in the mandalas o f the eight
great B odhisattvas.21 They are also found on Nepalese coins o f the
M alla p e rio d 22 and on Indian23 coins. An eight-pointed star24 can
serve the same ritual function as the cight-petallcd lotus, but is less
common.
In addition to eight-petal led lotuses, lotuses with two, four, 10, 12,
16, 24, 32, 100, 1000 or more petals appear in mandalas and yantras.
The num ber o f petals is mostly even, but yantras with an odd num ber
o f petals (for exam ple, five) are also found, in which case their
directional orientation may not be o f any obvious relevance. A
special kind o f six-petalled lotus is the va/ra-lotus described in the
K ubjikam ata-Tantra. This is an eight-pet a lied lotus from w hich two
petals have been rem oved. Its shape resembles a vajra with three
peaks on either side.25 Nepalese coins o f the Malla period also depict
four-petalled,26 five-petalled27 and six-petalled2K lotuses.
Som e texts prescribe that the lotus petals should have different
shapes depending on the purpose o f the associated rite. Thus the
petals may be curved along their edges, and with or without pointed
tips, and so forth (Torzsok, p. 207).
The lotus pattern is commonly found in current ritual practice, for
example, in M aharastra. An eight-petal led (astudala) lotus, prepared
from grains or coloured powders, frequently functions as a support
for ritual vessels. Atop the vessel is a dish filled (with grains) (purnapatra), especially uncooked rice, that serves as the scat o f the main
deity o f the rite. In Nepal, lotus designs can also be found on stones
21 S e e , for e x a m p le, L eidy/Thurm an 1997: 2 6 - 2 8 .
72 S e e R h o d e s / G a b r i s c h / f della R o cc h c tta 1989, Plate 16, nos. 2 8 1 - 2 8 5 , 2 8 9
2 9 7 , P late 17, nos, 2 9 8 - 2 9 9 , 3 0 3 - 3 0 4 , 3 1 3 - 3 1 8 , 3 2 7 - 3 2 9 , 3 3 8 - 3 4 4 , Plate 18, nos.
3 4 5 - 3 4 6 , 3 6 1 , ctc.
23 S e e Sircar 1968, Plate xx iii, nos. 9 and 10 rev,
24 S e c R h o d e s / G a b r i s c h / t della R occhctta 1989, Plate 2 9 , no. 5 9 6 rev., Plate 30,
no. 6 5 6 rev., Plate 3 3 , no. 8 2 5 rev., Plate 35, no, 9 3 6 rev,, Plate 36, no, 9 7 8 rev.
25 For a descrip tion and a draw in g o f the va/ra-lotus, s e e H c ilijg e r s -S c c le n 1994:
131-132.
26 S e e R h o d e s / G a b r is c h / t della R o cchctta 1989, Plate 23, nos, 4 5 0 rev., 4 5 9 - 4 6 5
obv., Plate 2 4 , nos. 4 6 6 - 4 6 7 obv.
27 S ee R h o d e s /G a b r is c h / t della Rocchctta 1989, Plate 2 2 , nos. 4 3 8 rev., 4 4 4 rev.
28 S e e R h o d e s / G a b r i s c h / f della R o cch ctta 1989, Plate 2 1 , no. 4 1 8 rev., Plate 2 2 ,
nos. 4 4 5 obv., 4 4 6 ob v., Plate 35, no. 95 3 rev.
M A N D A L A , Y ANTRA A N D CAKRA
23
near thresholds, on roads or in public places. T hese stones, w hich

have been term ed 'lo tu s sto n e s by some authors (Auer/Gutschow
1974: 29, 32, 33, 124), serve special functions in the N evar religious
tradition.29
A pattern o f nine lotuses arranged in groups o f three placed one
above the other appears in several im portant mandalas. T hese in
clude the Pancaratra navapadm am andala (see C olour Plate 15), the
Saiva navanabhamandala (Colour Plate 18) and several versions o f
the Buddhist vajradhatumandala,30
In the context o f specific mandalas and yantras, different inter
pretations o f the lotus design and the lotus petals are given. The
(most likely seventh-century) G anesapurvatapaniya-U panisad, sec
tion 3 gives an interpretation o f the constituent parts o f a yantra o f
Ganesa. The y an tras innermost ring of eight lotus petals is taken to
represent the eight-syllabled gayatri; the adjacent ring o f 12 petals,
the 12 Adityas and the vowels; and the following ring o f 16 petals,
the purusa w ho consists o f 16 parts (kala), and the consonants, M isra
1959: 4 8 2 -4 8 3 interprets an unidentified Sakta yantra as represen
ting the process o f creation, and takes the eight petals o f its lotus to
signify the five elem ents, m a n a s, buddhi and ahamkara. Som e
29 W e k n o w o f se veral ty p e s o f s t o n e s with e n g ra v e d lo tus d e s ig n s . O n e type is
d e s c r ib e d as a g u a rd ia n s to n e ( N e v a r i pikhalakhu, s o m e t i m e s c o n s id e r e d s y n o
n y m o u s w ith N ev a r i chetrapala) in front o f thresholds (G u t s c h o w /K o l v e r /S h r e s t h a carya 1987: 3 5 , 5 4 - 5 5 , 9 2 , 120 and K o lv er/S h resth a ca ry a 1994; 1 0 1 , 2 0 9 ) . A c c o r
d in g to T o f f in 1999: 4 2 , the pikhalakhu p rotective sto n e is c o n s id e r e d the a b o d e o f
th e d e it y P i k h a l a k h u d y a h , w h o m b o th B u d d h i s t s and H i n d u s i d e n t i f y w ith
K umara/S kanda. T h is stone reportedly h a s different fu n ctio n s in ritual: o ffe r in g s are
d ep o sited on it; it r e c e iv e s w o rs h ip as part o f the m arriage c e r e m o n y , at w h i c h tim e
it m a y b e sm ea r ed w ith c o w - d u n g ; or e ls e a diagram m a y b e drawn on it (T o f fin
1999: 4 3 ) . For a photograph o f o n e su ch stone, s e e G u ts c h o w / K o lv e r /S h r e s th a c a r y a
1987: 120, no. 121. A n o t h e r ty p e o f s t o n e is c a l l e d chvasa ( G u t s e h o w / K o l v e r /
Shresthacarya 1987: 35 and K olver/Sh resth acarya 1994: 105). T h is is a d eifie d stone
f o u n d at c r o s s r o a d s on w h i c h ritu ally im pure ( ucchista ) o b je c t s are d iscarded
( G u t s c h o w 1982: 105). T h e g ra n d m o th e r-g o d d es s (ajima) is propitiated there. For a
photograph, s e e G u ts c h o w / K o lv e r /S h r e s t h a c a r y a 1987: 9 2 , no. 14. A third ty p e o f
s to n e is c a l le d mandalVmamdah ( N e v a r i ) b e c a u s e o f its m a n d a la -lik e d e s ig n (G uts c h o w /K 5 1 v e r/S h re s th a ca r y a 1987: 6 5 , 120 and K o lv e r / S h r e s th a c a r y a 1994: 2 5 8 ).
T h i s s t o n e m a y b e c o v e r e d w i t h b ra ss. F o r a p h o t o g r a p h , s e e G u t s c h o w /
K olver/Sh resth acarya 1987: 120, no. 122.
30 S e e , for e x a m p le , the central part o f the 1 2 th-century vajradhatumandala from
A l c h i , L ad akh ( L e i d y /T h u r m a n 1997: 4 0 , F ig u re 3 6 ) . T h e n i n e f o l d structure is
already se e n in the m a n d a la o f the e i g h t great B o d h i s a tt v a s in C a v e 12 in Ello ra,
Maharastra (late s e v en th to early eighth century) ( s e e F ig u re 21 in L e id v /T h u r m a n
1997:31).
24
G U D RU N
BU! IN IiM A N N
Pancaratra texts identify the mandala with the deity s body and its
constituent parts with the deitys body parts. Thus the Satvata-Samhita (Rastelli, p. 139) takes the lotus to represent the d eity s intellect
(dhi = buddhi). The Visnu-Samhita provides yet another interpreta
tion o f the lotus, equating it with the deitys heart,
The central lotuses in mandalas or yantras often have triangles and
h e x a g ra m s inscribed in their pericarps. In a tw o -d im e n s io n a l
structure, the lotuses are usually surrounded by a square enclosure,
often termed a seat or throne (plilui), adjacent to which may be a
corridor or passage (vlthl) for circumambulation (prndnksina). In a
three-dim ensional structure, the pitha would be the support o f the
lotus and project beyond it. Between one and three concentric circles
and a square (with often three nested lines) frequently surround the
central lotus on the outside. These geometrical structures will be
discussed separately in section 2.3 in the context o f yantras. In
m andala designs, lotuses also appear in com bination with S iv a s
trident(s).31 A central lotus in a mandala may be replaced by a wheel
(cakra)?2 The deities are then assigned to the hub and the spokes o f
the wheel. A wheel can also appear in com bination with a lotus
design.33
1.3,2 The Square Grid
A com m on structural device o f certain mandalas is the square grid,
which may incorporate a lotus design (made o f squares) in its centre.
E xam ples o f this structure are the bhadram andalas em ployed in
Smarta ritual, which are analyzed in the next essay. The square grid
is obtained by drawing a certain num ber o f vertical and horizontal
base lines to form squares on a surface. The squares, called pada or
kostha, are assembled into different shapes and parts by filling them
with coloured powders or grains. The constituent parts o f the sarva31 S e e the mandala o f the trident and lotuses {triSulabjanm ndala) and the mandala
o f the three tridents and ( s e v e n ) lo tus es ( tritri&UMyumandala) in Sanderson 1986 and
P a do u x , pp, 2 2 5 f f . (w ith Illustrations 1 - 2 ) and the trident mandala reconstructed in
T orzsok , A p p e n d ix 3 (with C o lo u r Plate 19). T h e trident o f the Trika has been inter
preted as representing the universe (Torzsok, p. 195).
32 In d escrip tio ns o f m a nd a la -lik c structures, w o rd s d en o tin g parts o f a lo tus arc
s o m e t i m e s treated as in tercha n g ea b le with w o r d s d e n o tin g parts o f a w h e e l; s e c
Torzsok, p. 181.
33 For a c o m b in a tio n o f a lotus and a w h e e l in m andalas, se e, for e x a m p le , the
cakrabjam andala ( bbadrakam andala) describ ed in Rastelli, p, 124 and the mandala o f
Svacchandabhairava describ ed in Torzsok, pp. 2 0 1 - 2 0 3 .
M AN D AL A, YANTRA AND CAKRA
25
tobhadra include (see Table, p. 87): a w ell (vapi), an offset design

( bhadra),34 a c re e p e r ( valll), a c h a in (srhkhala) and a crescent
m o o n ( khandendu), In the centre is usually a lotus w ith a pericarp
(karnilca), and on the outside o f the m andala a square w ith three
nested lines, coloured white, red and black. The three lines are
identified with sattva, rajas and ta m a s 35 and coloured white, red and
black respectively from the inside to the o u tsid e .36 In addition to
these parts, the Hngatobhadras contain one or more phallic symbols
( linga) o f Siva, which are them selves called R udra or Siva. Some
hngatobhadras contain additional parts, such as a corridor or passage
( vithi) for circum am bulation w hich surrounds a throne (pitha) and
miniature creepers (laghuvallT) and miniature chains (laghusrnkhala).
The characteristic elem ent o f the ra m a to b h a d ra s is the s e a l o f
Rama, which usually consists o f the inscribed words raja rama. The
ganesa- and suryabhadras feature im ages o f G anesa and the sun
respectively.
1.3.3 Other Designs
The sarvatobhadra reconstructed by Brunner in this b o o k belongs to
a different m andala tradition from the aforementioned sarvatobhadra,
and consists o f different constituents. Structurally, the m andala
represents a co m b in atio n o f the square grid seen in the bhadramandalas and a rounded lotus shape on a throne in the centre. The
lotus consists o f the pericarp (karnika), filam ents (kesara), petals
(patra, dala) and the tips o f the petals (dalagra). The throne (pitha)
has four feet (pada) and four lim bs or b o d ies(gatra), that is, side
parts in the form o f the bodies o f men and anim als (Brunner, pp.
1 6 7 -1 6 8 ). T h e re is a c o rrid o r or p a s sa g e ( v ith i) for c irc u m
am bulation (pradaksina) and an outer enclosure consisting o f entry
and exit passages (dvara). In addition, we find parts called sobha and
upasobha 37 and c o r n e r s (k o n a ). O th e r m andalas, such as the
M In architectural te r m in o lo g y , the term bhadra d e s ig n a t e s an o f fs e t p ro jectio n
c o m m o n to N orth Indian tem p le plans.
35 T h is interpretation is a lso g iv e n to m a nd a la s o f the Paiicaratra tradition ( s e e
Rastelli, p. 139).
vFor the outer square as part o f yantras, se e se ctio n 2 .3 ,8 .
37
T h e s e term s are a ls o written sobha/upasobha in the Pauskara-S am hita (cf.
R a stelli, p. 139). For an ex p la n a tio n o f th e se term s, s e e A p p e n d i x 2 to T o r z s o k s
contribution. Brunner, p. 169 interprets sobha as a d o o r or entrance p a v ilio n o f the
first en clo su r e o f a p a la ce or t e m p l e and upasobha as p o s s i b l y s ig n if y in g a pavilion
26
GUDRUN BUHMKMANN
tirimandala and the navanabliamandala reconslructed in Torzsok (sec

her Appendices 1-2 and Colour Plates 16 and 18), feature additional
constituent parts. These include door segments termed kantha (the
upper part o f a door) and npakantha ((lie lower part o f a door). The
upakantha is also called kapola in some texts.
Different interpretations are given to the constituent parts o f these
mandalas. The Srimandala reconstructed by Torzsok (see her A p p en
dix 1 and C olour Plates 16-17) is surrounded by a square with five
nested lines coloured transparent, yellow, black, red and white from
the inside to the outside. These five lines are identified with the five
kala,s which constitute the Saiva universe. In the Pancaratra tradition
the m andala is som etim es identified with the d e ity s body. Inter
preting one o f these m andalas, the Satvata-Sam hita equates the
sobhiis with the deitys organs o f action (karana), the upaSobhas with
the subtle elements ( tanmatra), and the corners and gates with the
deitys sense organs (cf. Rastelli, p. 139). In the Pancaratra tradition
the mandala also becomes a representation o f the universe, when its
constituent parts are equated with cosm ic principles and divine
powers. T hus the V isnu-Sam hita, for example, identifies the five
colours used in the mandala with the live elements (see Rastelli, p.
141).
1.4 The Question o f the Origin and Date o f Mandalas
Several scholars have suggested that Tantric mandalas arc rooted in
Vedic traditions. The layout o f Vedic altars is taken as indicative o f
an early interest in geometric designs endowed with cosm ological
symbolism (Gaefkc 1987: 153). The method of determining the lines
o f the c o m p ass for the co n stru ctio n o f sacrificial altars, the
consecration o f bricks on the surface o f a cayana altar by means o f
m antras and the locating o f deities on those bricks arc essential
features o f V edic rituals (Apte 1926: 2 - 3 ), and aspects o f these
rituals recur in the practice o f constructing mandalas and invoking
deities into their parts. The sacred space o f mandalas and yantras can
be seen as a continuation o f the Vedic sacrificial site (Schneider
on top o f a s e c o n d a ry d o o r . T h e translation offset d e s i g n for Aoblul and recess

d e s i g n fo r u p a to b h fi is used by P.P. A p te in the in tro d u ctio n to his ed itio n and
translation o f the P a u sk a ra -S a m h ita (Part I), p. xii, For d ra w in g s o f these parts, sec
T o r z s o k s Illustration 1 and also Nikita 1991: 319.
M AN D AL A, Y A N TR A A ND C AKRA
27
1988: 100), and the square enclosure o f Tantric m andalas in parti

cular as an analogue o f the sacred fire altar (Gupta 1988: 3941). But
the sim ilarities b etw e e n the tw o traditions appear to end here.
Authors like Mitra 1958: 1 1238 are going too far w hen they assum e
that patterns displayed by yantras and mandalas can be traced back to
the S ulba-Sutras o f the V ed an g as (w hich p rescribe th e w ay to
construct sacrificial altar diagrams), since the patterns displayed by
Tantric m andalas are distinctly different. So are the m antras and the
deities invoked into m andalas and the details o f the rites. The
problem o f the sim ilarities and differences b e tw e e n V ed ic and
Tantric traditions is com plex and needs to be explored in greater
detail in a separate study. Such an investigation w ould have to trace
the influences o f other traditions on mandalas as well.
The oldest H indu m andalas m ay date back to before the sixth
century A.D. A m ong the oldest mandalas that can be dated are two
types o f vastupurusamandalas described in V a ra h a m ih ira s BrhatSamhita. This text is co m m only placed in the m iddle o f the sixth
century. The tw o vastupurusamandalas are described in chapter 53,
b u t w ere o b v io u s ly n o t c re a te d by V a r a h a m ih ira b u t rath er
incorporated from older unidentified sources. Apte 1987: 141 notes
that the first type o f vastupurusamandala is d e s c rib e d in the
Pauskara-Sam hita o f the Pancaratra, w hich he dates to ca. 400 A.D.
(Apte 1986: 3, 1999: 18) or at least 450 A.D. (Apte 1987), while
M atsubara 1994: 34 assigns the Pauskara to 500 A .D . H ow ever,
these early dates are highly speculative, th e u p p e r lim it for the
com position o f the Pauskara-Sam hita being only the tenth century.
M oreover, dating a Sam hita as a whole is problem atic, since these
texts were constantly revised and reworked by redactors, Sanderson
2001: 38, note 50 states that he found evidence that the PauskaraSamhita (along with the Jayakhya-Sarnhita and the Satvata-Samhita)
were influenced by Tantric Saiva systems. A t this time the com plex
descriptions o f m andalas found in the Pauskara-S am hita cannot be
dated with certainty. W e do not have clear evidence for establishing
dates for the d e v e lo p m e n t o f y antras either. B ro o k s 1992: 34
considers the p o ssib ility that the m o s t fam o u s o f y an tras, the
sricakra/M yantra, developed before the sixth century.
38
C h a tto p a d h y a y a 1 9 7 8 : 8 0 , too, s u g g e s t s a c o n n e c t i o n b e t w e e n yantras and
sh apes o f sacrificial d iagram s used in the V e d ic tradition.
GUDRUN
28
b DMNI-MANN
2 Ynntrn
2.1 General R em arks
The word yantra designates an instrument, machine,' mechanical
device or appliance (especially one used in warfare), and also a
magic diagram. It is derived from the verbal root yum , to control.
w For this m e a n in g o f y a n tra , se c, for e x a m p le , Bhag a v a dg ita 18,61. M e c h a

nical a p p lia n c e s and m a c h in e s ca lled yantras arc d escrib ed in chapter 31 o f the
Sam a ra ng a n a -S utra d h a ra , a w ork on architecture a scribed to B hoja, w h ic h w a s
studied b y R a ghavan 1956: 2 1 - 3 1 . S ee also the o v e r v ie w in Slink In 1967: 3 0 - 5 2 .
For different a stro n o m ica l in struments called yantras for use in o b serv a to ries , se c
V o lw a h s c n 2 0 0 1 : 40ff.
M a y r h o f c r 1 9 8 6 - 2 0 0 1 , v o l u m e 2: 3 9 8 e x p la in s the w ord yantra as an
instrument for fastening. Kramrisch 1946, v o lu m e I : I 1 1 2 , w h o apparently d o e s
not differentiate b etw e en the terms mandala and yantra, d efin e s yantra as fo llo w s:
A Y antra is a g e o m e tr ic a l co n tr iv a n c e by w h ic h any a sp ec t o f the S u p r e m e
Principle m ay b e bound (yantr, to bind; from the root y a m ) to any spot for the
purpose o f w orship . It is an artifice in w h ic h the ground (b h um i) is co n v er ted into
the ex te n t o f the m a n ifes ted u n iv e rse , For two r e lig io u s e t y m o l o g i e s o f the w ord
yantra, s e e the fo llo w in g verses from tw o different chapters o f the Kulfirnava-Tantra.
T h e first v er se d erives the word yantra from the verbal root y a m with the prefix in,
m ea n in g to restrain, prevent, t a m e .
kam akrodhadidnsatthasarvndiihkliaiiiyantranfiS /
ya n tra m ily ahur ctn sm in dcvn h prinfiti pujitah // 6.86
B e c a u s e it restrains all s u fferin g arisin g from the d efec ts (in the form) o f desire,
anger and so forth th ey call it yantra. T h e g o d w h o is w orsh ip ped in it g r a c e s (the
practitioner).
ya m a b h u ta d isa rv e b h y o b h a y e b h y o 'p i kulcdvnri /

trayatc satatam caiva ta s m a d y a n tm m iiiritam II 17.61
O m istress o f the kula, b cca u se it protects a lw a y s from a b so lu te ly all dangers, such
as Y a m a and (ev il) spirits (blnll;i), therefore il is called yantra."
Purnanandas &rilattvacintamani 17.2 explains the word yantra in a sim ilar way:
y a m a y a ty n kh iln m papain trayatc m ahato bhayat /

sadh a ka m p u ja n a d dhyanal ta sm a d yantrah prakirtyato / /
It su bd ues all e v il, it protects the practitioner from great danger w h e n w orsh ip p e d
(and) v isu alized ( dhyiina ); therefore it is called yantra,
(The m a sc u lin e g en d er o f yantra here is rather unusual, but s e e a lso the citation,
from the Kalivilasa-Tantra b elo w .)
R a g h a v a b h a t t a s c o m m e n ta r y , pp. 5 1 9 , 6 - 7 on Saradatilaka 24.1 q u o t e s the
f o llo w in g e t y m o lo g y o f the word yantra from an unsp ecified Samhita:
m a norathaksarany aim n iya n lrya n tc tapodhanah /

In this the letters o f (= c o n v e y i n g ) desires are affixed. O a s c e tic s .
H e co n tin u es with a line reminiscent o f Kularnava-Tantra 6 .8 6 cited above:
kainakrndh;1 didosan va (correct to dosotthadi) dirghadiihklm uiyantranat II
ya n tra m ily aliuh / Hi

T h e y call it yantra b cca u se it restrains p rolon ged suffering arisin g from d efec ts (in
the form) o f desire, anger and so forth.
M AN D AL A, Y AN TRA A N D C A KR A
29
A general characteristic o f yantras is that they are small in size. In

contrast, mandalas vary in size and can be large enough to allow for
priests or initiands to enter them through doors and to w alk around in
them, for example, during an initiation (diksa). With the exception of
yantras placed below temple statues at the time o f their consecration
and yantras installed permanently for w orship in m athas or temples,
and a few other eases,41 yantras are generally mobile, w hereas m an
dalas are not. While mandalas can employ different colour schemes,
the use o f colour is less com m on if not indeed irrelevant in the case
o f most yantras. Texts may prescribe that the lines o f a yantra be
traced with a specific colour, for example, with turm eric or blood,
but the space inside a yantra is never filled with colours as it is in the
case o f mandalas. And while pictorial representations o f deities can
appear in mandalas, such images are generally not found in y antras.42
Like mandalas, yantras are believed to be effective only w hen w or
shipped. However, some texts claim that the act o f merely view ing a
mandala43 or draw ing or recollecting a yantra44 brings about b en e
ficial results. How ever, according to K alivilasa-T antra 7 .9 c d -1 0 a b
and 2 7 . 2 lab, w orship o f a deity in a yantra is n o t recom m ended in
the present ka li era.
Based on an analysis o f texts o f the Trika School o f Kashm ir,
Brunner, p. 162 briefly defines a yantra as a linear representation on
a specific surface, such as birch-bark. She adds that yantras alm ost
inevitably have letters, seed (bija) syllables or mantras inscribed in
them. Since m antras frequently em ploy verbs in the im perative to
express an order, B ru n n er suggests the translation co erciv e dia
g ra m s for yantras. Similarly, Rastelli, p. 142 concludes from her
study o f the Pancaratra texts that yantras have inscribed mantras.
A n o th er ety m o lo g y o f y a n t r a is found in K a liv ilasa-T an tra 33.1:
b ijm iim k o n a vijn iiim m ya tn a ta s trayate ya ta h /
ten a yantra iti khyiita isananmkhanihsrtah //

B e c a u se it with effort protects the k n o w le d g e o f the angles/co rn ers ( o f the d ra w in g )
(reserv ed ) fo r the seed (syllables), w h ic h (k n o w le d g e ) c a m e forth f r o m ISanas (that
is, S iv a s) m o u th , therefore it is called y an tra,
(T h e form srta h th e text re a d s e r ro n e o u s ly sm rtah can b e e x p la in e d as an
e x a m p le o f ease attraction; srta m is the cx p ected form.)
41 Sec Rastelli, p. 143 for y a n tra s d escrib ed in the A n ir u d d h a - S a m h ita , w h ic h are
d raw n and then w o rsh ip p ed on a p latfo rm and are therefore n o t m obiie.
42 T h e sa u d a rfa n a ya n tra ( sec Rastelli, pp. 1 4 8 -1 5 0 ), w h ic h is a c o m b in a tio n o f a
y antra an d a figure o f V isnu, is an exceptional case.
43 S ee the S u p ra b h e d a quoted in B ru n n er, p. 175, n o te 53.
44 S ec the description in Rastelli, p. 146.
30
GUDRUN B O ll NBM ANN
Authors such as Kscmariija,<5 consider il characteristic o f certain

yantras that mantras arc inscribed in them. But at least in later texts
and in modern practice mantras or syllables are not necessarily part
o f yantras (sec sect ions 2.2,1 and 2.2.2),
Brunner concludes that yantras, like some mandalas, arc used for
w orship in desire-oriented (k f i m y n ) rites, but their main purpose is
magical. She asserts that mandalas are employed in rituals to obtain
the d eity s favour, but without the presence o f a magical element.
This may be true o f the limited group o f texts Brunner analyzes, but
it is not the com plete picture, for Torzsok, pp. 201 -209 discusses
mandalas described in the Bhairava-Tantras which are used in rituals
leading to the attainment o f supernatural powers (s i d d h i ),
Yantras can be two- or three-dimensional. Two-dim ensional yan
tras are designs on paper, textiles and other materials. T hree-dim en
sional yantras are raised structures usually made o f metal. The wellknown s r i y n n t r n or s r i c t i k m can be represented either two- or threedimensionally. Three-dimensional S r i a i k r n s are classified differently,
depending on the authors use o f terminology.'Ul
,|5 See Kscmiiraja's statement quoted by Sniulcrsnn in Padoux 1986: 33.
l6 K hanna 1986: 101 lists the following kinds o f Jliree-dim cnsional srfenknts:
1) bhuprastara - e ngraved or em b o sse d on a Hal metal surface;
2) m cruprastfira - p y r a m id - s h a p e d , with e ith e r tw o, three o r nine e le v a tio n s
rising one above the other; and
3) kailaiapraslara (for kailasa0), having a slightly raised surface,
F o r in fo rm a tio n on nicruprastara-, bhuprastara- and ka ila sa p ra sta ra -^ric a k n a , see
Rao 1990: 116-1 17 and S h a n k a ra n a ray a n a n 1970: 9 2 -9 4 , A ccording to K hanna, the
bhuprastara type o f Sricakra features slightly raised shapes on a flat surface. O th e r
authors assert that the bhuprastara (also called b hup rsth a) form o f the Sricakra has a
c o m p le te ly flat surfacc and is tw o-dim ensional.
T w o k in d s o f bhuprastara (called bhuprstha) yantras are referred to in C h a w d h ri
1990: 2 and C h aw d h ri 1992: 8. T h is author lists the following kinds o f yantras:
1) 1b h o o p rish th (= bhuprstha) y a n tra s - yantras from m aterials that indicate the
use o f earth; they are fu rth er s u b d iv id e d into raised y a n tra s and c a rv e d
yantras;
2) m cru p rish th (= m cruprstha) y a n tr a s - raised yantras w hich are s h a p e d like
m ountains;
3) patal (for piitala) y a n tra s - carved yantras shaped like inverted m ountains;
4) m cru prastar(= nicruprastara) y a n tr a s - cut yantras, and
5) rurajii p rish th (m isprint for k u rm n p rsth n ? ) y a n t r a s - yantras that h ave the
shape o f a rcctangle at their base and that o f the back o f a tortoise ab o v e their
base.
Finer d istinctions am o n g these categories are not provided, C h a w d h r is classification
h as b e e n a d o p te d by B e c k m a n 1996: 5 0 - 5 2 . A so m e w h a t sim ila r c la s s ific a tio n
appears in Johari 1986: 58:
1) B h u -P rash th a Yantra: engraved or draw n on a fiat surface.
M AN D AL A, Y AN TRA A N D C A KR A
31
2,2 Types o f Yantras

A ttem pts to establish a distinction betw een m andalas and yantras
have already been discussed in section 1.1. This section will discuss
attempts to classify yantras. Renou/Filliozat 1947-1953, volum e 1:
568 attempt to distinguish betw een two types o f yantras. The first
type o f yantra is supposed to be purely linear-geometric, and to be
employed in temples for purposes o f perm anent worship. T he second
type reportedly features a linear design with iconic representations
and is used in tem porary and individual worship rites. N o exam ples
or further details are given, and in the absence o f explanations and
textual evidence this unusual distinction remains unclear.
In a book on yantras written for a general readership, S.K.R. Rao
1988: 14-15 distinguishes the following three types o f yantras:
1.
2,
yantras for worship or for actualizing a deity (pujanayantra),

yantras for magical protection (raksayantra) and
2) Meru-Prashtha Yantra: three-d im en sional form c o m p o s e d o f m etal or sto n e or

g c m - s t o n o s and sh aped lik e a pyram id, h a v in g a broad b a s e and n arrow in g
gradually toward the top lik e a m ountain ( m cru ).
3) Fatal Yantra: d e e p ly engraved e x a c t ly the reverse o f the pyram id yantra.
4 ) Mcru-Prastar Yantra: c o m p o s e d o f p i c c e s that are g lu e d or w e ld e d to each
other instead o f b ein g o n e solid p ic c c .
Metal iric a k u s su pported by a tortoise pedestal are c o m m e r c i a l l y so ld in India as
kurmaprstha-Mcakras. B r o o k s 1990: 107 lists o n ly the t w o - d im e n s io n a l bhuprastara
( sp re a d in g o v e r the e a r t h ) iricakra as a c a t e g o r y a lo n g w it h th e f u l l y threed im e n s io n a l m cru or sum cru form and the m o re fla ttened, e l o n g a t e d kailasa form .
Rao 1990: 118 refers to a cla ssific a tio n o f sricakns into bhuprstha , kacchapaprstha
and meruprstha W h e n a cakra is drawn on a flat s u rfa ce , it is c a l le d bhuprstha.
W h en supported by a d w a r f pedestal, raised like the back o f a tortoise it is referred
to as kacchapaprstha. W h en the cakra has the form o f a m o u n ta in w ith d iffer en t
e l e v a t io n s it is k n o w n as meruprstha. U m an and an ath a in his N i t y o ts a v a (written in
1745 A .D .), p. 6 5 , 4 - 9 , lists o n e type o f bhuprastara-sricakra in w h i c h the lin es o f
the cakra are e l ev a ted , and three ty p e s o f m eruprastara-sricakras in w h i c h s p e c i f i c
parts o f the structure arc elev a ted . Pranavananda < 1 9 7 7 > : 3 5 - 3 9 c la s s i f i e s sricakras
into the f o l lo w i n g three categories:
1) bhuprastara or bhukrama;
2) kunnaprastara or kurmaprstha and
3) mcrukrama, mcruprastara or meruprstha.
A c c o r d i n g to this author, the first k ind is a M cakra w h o s e c o m p l e t e d e s i g n is
en graved or e m b o s s e d . T h e sricakras o f the s e c o n d type all feature n ine triangles in a
slig h tly e lev a ted p o sitio n. Such a cakra m a y a lso be p la cc d on the b a ck o f a tortoise
(and h e n c e the n a m e kurmaprastara or kurmaprstha). T h e third ty p e o f sricakra has
all its triangles raised like a mountain.
32
G UD RU N BUHNli MANN
3 . yantras which are also callcd mandalas; they arc defined as sur
faces on which ritual objects arc placcd.
In the same book (Rao 1988: 19) the author introduces yet another
threefold classification o f yantras:
1,
2,
3,
yantras for magical purposes, generally called protective yantras

(raksayantra),
yantras for actualizing divinities (dcvatayantra) and
yantras that facilitate meditation (dhyanayantra).
The items in the first two categories in both lists arc identical, even if
their sequence differs. The third type o f yantra in the first list will be
discussed below. The third category in the second list appears to
refer to certain Buddhist m andalas (Rao 1988: 27). Even though
descriptive details are missing and the categories arc presented in a
som ew h at unsystem atic way, R a o s classification is helpful, but
clearly not sufficient. Rao has pointed the reader in the right
direction by taking the ritual function o f yantras into consideration
when attempting to classify them.
Building on B runners and R a o s work as well as on the basis o f
my study o f Tantric texts o f the later period, 1 would like to suggest
the following tentative classification o f yantras as a guideline. This
classification, according to the distinctive features and ritual use o f
yantras, is not intended to be exhaustive and may not be applicable to
all South Asian Tantric traditions.
1.
2.
3.
Yanlras which function as supports for ritual implements during

a worship ritual, being referred to as yantras for (establishing) a
foundation (stlm panayanlm ),
yantras employed in a practitioners regular Tantric worship o f a
deity, often referred to as yanlras (which arc supports) for w o r
ship (pujayantra, pujadharayanira) and named for tlicir presiding
deity, for exam p le, y an tra for the w o rsh ip o f G a n a p a ti
(,ganapa iipujayan tra), and
yantras em ployed in optional dcsirc-oricntcd rites, which arc
perform ed on special occasions. Yantras used in a special ritual
for a certain deity arc included here as well as yantras which are
prepared for specific magical rites, and which are often named
for these rites, for exam ple, yantra for attra ctio n (akarsana-
M ANDAL A, Y A N T R A A ND C A KR A
33
yantra) (Illustration 1). A fter the ritual is com plete, the instruc
tions may recommend that these yantras consecrated for magical
purposes be made into amulets and worn on the body ( dharanayantra) in order to obtain the desired results, such as protection
or the acquisition o f pow er and wealth. A m ong these yantras, the
yantras for protection (raksayantra) figure prom inently in texts.
In the category o f yantras for desire-oriented rites I also include
magic (num ber) sq u a re s .47 T hese are d iag ram s w ith num bers
inscribed, the sum o f which remains the same, regardless o f the
direction in which one adds them up.
T hese three categories are detailed below.
2.2.1 Type 1: Yantras for Establishing a Foundation
These yantras feature simple geometric shapes, such as a triangle or
a circle. They function as supports for ritual im plem ents, such as
lamps or vessels, in special desire-oriented (ka m ya ) or m agical rites.
Such supports also figure in the regular Tantric puja, in w hich they
ap p ear to be referred to as m andalas (see Illustration 2 ).48 T heir
function can be compared to that o f the seat-m andalas in B ru n n ers
m a n d a la classificatio n (sectio n 1.2); h o w e v e r, B r u n n e r s seatm andalas, w hich are m ade o f cow -dung and similar materials, are
without any clearly recognizable structure.
2.2.2 Type 2: Yantras Employed in Regular Worship
Yantras o f this type usually feature com m on geom etric shapes, but
generally do not have mantras inscribed, at least according to the
later Tantric texts that I have studied. H ow ever, the deity and her/his
em anations are invoked into the yantra with mantras. A few yantras
have the names o f these em anations or surrounding deities inscribed
fo llo w in g the expression salutation t o ( nam ah). R e g a rd le s s o f
w h e th e r the m antras are only used to invoke the deity or w hether
they are also inscribed in the yantra, they are o f utm ost im portance.
It is for this reason that the K ularnava-T antra states that a yantra
47 F o r m agic (n u m b e r) squares, sec, for ex a m p le , the illustrations in A b b o tt 1932:

5 1 5 -5 2 1 and the discussion in C a m m a n n 1969.
4,1
T h e m a n d a la in Illustration 2 s e rv e s as a su p p o rt fo r the v a r d h in l'/ v a r d h a n i
vessel in a T a n tric pujn.
34
G UD RU N BUHNKMANN
consists o f the deitys m a n t r a . Y a n l r a s employed in regular worship

are often m ade o f durable materials, such as copper.
In this category o f yantras 1 would also include the .sricukra, also
called sriyantra. In addition to being worshipped in ritual, this cakra
is also visualized and experienced in the practitioners body as a
manifestation o f the cosmic proccss o f creation and resorption with
spatial and mantric aspects, as Padoux explains in his contribution to
this book.
2.2.3 Type 3: Yantras Employed in Optional Desire-Oriented Rites
The third category o f yantras is required for the perform ance o f
optional rites, such as specific magical rites, and they arc often made
o f perishable materials, such as birch-bark or paper. T hese yantras
are draw n, acco rd in g to the instructions, with special w ritin g
materials and substances, such as animal or human blood or ashes
from a cremation ground. Johari 1986: 63 reports that such yantras
may be made from wheat flower, rice paste, beans or (grains of)
rice.5" They may also be incised on more permanent materials, such
as metal plates. D iscussions o f the various styluses used and the
m aterials on which protective yantras can be written are com m only
found in texts. The materials are considered extremely important for
the success o f the ritual, and correspond to the nature o f the rite
perform ed. Thus cruel rites require repulsive m aterials, and the
yantra used in the rite o f liquidation (m firnnn) as described in
M a h ld h a ra s 16th-century M antram ahodadhi 25,56ab and 25.59ab
should be written on human bone with certain poisonous substances.
General instructions for draw ing yanlras for different purposes,
including reducing fever, keeping snakes away and countering the
effects o f poison, can be found in various texts, such as chapter 24 o f
L aksm anadesikas Saradatilaka (10th-11th century), which is based
4,1
Cf. K u larn av a-T an tra 6.85ab and 6.87:
y a n tra m m a n tra m a ya m p ro kta tn devata m antrarupini I 6.85ab
sariram iv a jiv a s y a d ipasya sn ch a va t p riy c /
sarvcslim api d c va n a m tatha ya n tra m pra tisth ita m // 6,87
Sec also the sim ilar quotation from the K au lav aliy a-T an tra in W o o d ro ffe 1956: 93,
note 2:
ya n tra m m a n tra m a ya m p ro k ta m m antratnnl d evalaiva h i I
d ch a tm a n o r ya ttu l b hedo ya n tra d cva ta yn s tatha II
s'T w o yantras m ade from beans, rice and coloured sto nes are repro d u ced in Plate
2 o f his book.
M AN DAL A, Y ANTRA AND C A KR A
35
on chapter 34 o f the Prapancasara (ca. 10th century), and in chapter

2 0 o f the M antram ahodadhi. Y antras for m agical p u rp o ses (for
exam ple, Illustration 5) are described in detail in D a m o d a ra s 17thcen tu ry Y antracintam ani, also know n as the K alpacintam ani. The
ap p licatio n s include the six rites o f m agic (a b h ic a ra ), n am ely ,
a p p e a s e m e n t (ganti), s u b ju g a tio n ( vaslkarana), im m o b iliz a tio n
(stanibhana), e n m ity ( vidvcsana), e r a d ic a tio n (uccatana) a n d
liquidation (marana). Depending on their purpose, these yantras are
n a m e d y antras for s u b ju g a tio n ( vagyakarayantra), y a n tra s for
attraction* (akarsanayantra) (Illustration 1), and so on. The use o f
y antras in rites o f magic, which has been docum ented by previous
sch o la rs,51 continues up to the present day and can be observed even
in m odern Indian cities. Yantras featuring H anum at are sold in India
for the safety o f o n e s vehicle (vahanasuraksayantra). O ther yantras
are used for curing diseases at the recom m endation o f astrologers.52
Yantras used in magical rites may be ritually destroyed after their
use, inserted into a statue o f a deity that will then undergo burial, or
be crushed and eaten, tied to a tree or concealed in the intended
p e rs o n s home, depending on the instructions. They may be enclosed
in an am ulet container, such as a tube or a locket,53 sealed and then
worn around the neck, on the head, in o n e s h ead g ear,54 in a tuft o f
51 See, fo r e x a m p le , the list o f y a n lra s p o p u la r in S o u th In d ia p u b l i s h e d in
T h u rs to n 1912: 1 8 5 -1 8 7 and references in A bbott 1932, s.v. yantra.
52 See, for e x am p le, (he n u m e r o u s yantras in C h a w d h ri 1990, S h u b h a k a ra n 1992,
B c c k m a n 1996 and K h u rra n a 2000, and the section on y a n tra s ( ja n tr a ) in D ietrich
1998:172-175.
3 F o r y antras preserved in container am ulets, see U n tra c h t 1997: 132.
51 A m u le ts h id d e n in hats, tu rb a n s and o th e r h e a d g e a r arc d o c u m e n t e d in U n
tra c h t 1997: 89. In his p o p u la r b o o k s on y an tras, C h a w d h r i (1 9 9 0 : 6, 1992: 10)
re fe rs to a c a te g o ry o f yantra w h ic h is kept u nder o n e s cap or tu rb an or in o n e s
pocket, He calls th e m 'ch h a ta r' (v e rn a c u lar form for chalra [u m b re lla ]) yantras. T h is
c a t e g o r y has b e e n b o r r o w e d by B c c k m a n 1996: 52. C h a w d h r i (1 9 9 0 : 4 - 6 , 1992:
9 - 1 1 [cf. B e c k m a n 1996: 5 1 - 5 3 ] ) in clu d es c h h a ta r y a n tr a s as c a te g o r y 6 in his
fo llo w in g classification o f yantras, w h ic h is also su m m a riz e d in B u n c e 2001: xv:
1) sh a rir' (= sarira) y a n tra s - the y a n tra d e sig n s in th e c a k ra s o f th e h u m a n
bod y ;
2) (lim nin' ( = dharana) yantras - yan tras w o rn on the body;
3) a a sa n ' (= asana) y a n tr a s - y a n tr a s k ep t u n d e r o n e s se a t (a sa n a ) d u r in g
w o rsh ip or u nder the foundation o f h ouses, te m p le s or a statue o f a deity;
4) m a n d a l (= m a n d a la ) y a n tra s - y a n tra s fo rm e d by n in e in d iv id u a ls , o n e o f
th e m seated in the centre and the others in the eig h t d ire c tio n s; th e p e rs o n in
th e c e n tre p e rfo rm s the w o r s h ip o f th e 'ish a t' (= ista ) y a n tra ( th a t is, any
p articu lar yantra), w hile the oth ers recite certain m antras;
5) poo/7} (= pujfi) yanlras - yan tras installed in h o u s e s or te m p le s for w o rsh ip ;
36
G UD R UN B OI I NI ' MAN N
hair, on the arm, under the armpit, on the wrist or a linger and so
forth.55 A yantra which is to be inserted into a locket is first drawn on
a piece o f paper or similar material and consecrated in a w orship
ritual by a specialist. These lockets can be attached to the necks o f
anim als, such as cows, for their protection. Yantras may also be
attached to protective dolls hung near the entrance to a hom e or be
placed above a door.
Yantras employed in desire-oriented rites may be similar in design
to the yantras for establishing a foundation (type 1), but they often
have mantras inscribed. The mantras can be seed syllables (b ija )
combined with verbs in the second person singular imperative, such
as su b ju g a te ( v a slk u ru ), w hich ask the deity to carry out the
magical effects o f a rite on its recipient. The centre o f the yantra is
frequently inscribed with the nam e o f the person to be influenced,
termed the recipient or intended person (sftdhya). The place in which
the p e rso n s nam e is to be written is often indicated by the nam e
D evadatta. The re c ip ie n ts nam e is either surrounded by, or its
syllables are intertwined with, the syllables o f the mantra,
Yantras may also contain longer mantras* or even w ell-know n
hym ns (stotra, stud). The com position and ritual use o f hym ns or
devotional poem s in praise o f deities has a long history in South
Asia. Such hymns arc found in the Purana literature and the Tantras,
and in independent collections attributed to sages or seers as well. To
reinforce the efficacy o f hym n-recitation in bringing about the
prom ised material benefits, the practice arose o f reciting hym ns a
given num ber o f times. This practice is modelled on that o f repeating
powerful mantras. In time, hymns came to be regarded as powerful
magical formulas. W hereas the shorter m antras may be repeated
millions o f times to achieve a particular result, hymns are recited at
most hundreds or thousands o f times. Hymns em ployed for such
purposes include hymns for protection. These hymns often include in
their titles such terms as arm o u r (kavaca), protection (raksa), or
c a g e (panjara). In these hymns the deity is asked to protect each
6) 'ch h a la r' (= chntra) yantras - sec above; and
7) cfarshan' (= d a ria n a ) y a n tra s - y a n tra s w h ic h the d e v o te e b e h o ld s in the
m o rn in g for the sake o f auspiciousness.
MRastelli, p. 146 also refers to yantras hidden betw een the breasts o f w om en.
C o m p a r e the p ractice o f inscribing d lm ra im in m a n d a la -lik e stru c tu re s in a
B u d d h ist contcxt, w h ich is d o c u m e n te d , for ex a m p le , in D rcge 1 9 9 9 -2 0 0 0 , F igures
1-9.
M A N D A L A , Y AN T RA A N D C A K R A
37
part o f the practitioners body. The different parts, from head to feet,
are s y stem atically enu m erated . F or each p art o f the body, the
practitioner addresses the deity using a different descriptive epithet,
which is often connected with the respective body part. The d e ity s
nam es are assigned to and d e p o s ite d on the bo d y parts o f the
practitioner, and are believed to protect him like divine armour. As
well as being recited, these hym ns can be arranged in the form o f
yantras. For those w ho cannot themselves recite the hymn, a yantra
w ith the hym n inscribed in it is thought to bring about the same
beneficial effects as recitation. An e x a m p le o f a y an tra in this
category is the ramaraksayantra, which represents in a graphic m ode
the Ram araksastotra ascribed to B udhakausika. In my study o f the
R am araksastotra 1 reproduce two yantras in w hich the R a m a ra k s a
stotra is inscribed (B uhnem ann 1983: 93 and 107). Another, yet u n
published ramaraksayantra is included here as Illustration 3. T he
yantra consists o f a hexagram with a drawing o f R am a and different
seed ( blja) syllables in the centre. The hexagram is surrounded by
concentric circles and by squares, the first o f w hich has elaborate
gate structures w hich open in the four card in al directions. The
innerm ost square contains the text o f a version o f the R a m a ra k s a
stotra,57
In addition to yantras containing the text o f entire hym ns, there
are also yantras w hich are associated w ith individual stanzas o f
hym ns o f praise. W ell-known' exam ples are the yantras associated
with the Saundaryalahari and the Bhaktam arastotra. T he S aundarya
lahari is a hymn to the Tantric goddess Tripurasundari in 100 (som e
times 103) stanzas, It is traditionally ascribed to Sam karacarya, iden
tified with the Advaitin Samkara. The B haktam arastotra by the Jain
poet M anaturiga is a hym n to the first Jina R sab h a in 44 stanzas
according to the Svetam bara version, or 48 stanzas in the D igam bara
version. Each verse o f the Saundaryalahari becam e associated with a
specific seed (blja) syllable, w hich is inscribed in a yantra shape,
such as a square, a hexagram , a triangle, a lotus, and so forth (for
exam ple, Illustration 4), O nly one o f these yantras has the nam e o f
the intended person (sadbya) o f the rite inscribed on it, and only one
57 D iffe re n t v ersions o f the h y m n are p re se n te d and d icu ssed in B u h n e m a n n 1983.
T h e text in sc rib e d in the ra m a ra ksa ya n tra re p ro d u c e d in Illu stratio n 3 c o n ta in s an
in tro d u c to ry s c c tio n w ith m is c e lla n e o u s v e rs e s and v e rs e s 2 - 1 5 (cf. B u h n e m a n n
1983: 2 6 - 2 7 ) o f the stotra, w h ic h latter r e q u e s t R a m a to p r o te c t the p r a c t i t i o n e r s
b o d y parts.
38
GUDRUN BIJIINHMANN
yantra contains a verb in I he second person singular im perative.

T hese yantras arc worshipped, and the seed syllables inscribed in
them are recited a large num ber of limes, for the attainm ent o f
desired, usually m undane, benefits, Kneh individual stanza o f the
Bhaktamarastotra is associated with a mantra addressing not the Jina
Rsabha but goddesses, Yaksas and gods, and each mantra is prefixed
by seed syllables. The mantras often contain second person singular
im perative verbs. In a similar fashion individual yantras are also
associated with the 47 stanzas o f the Jain Kalyanamandiraslotra. The
yantras associated with the stanzas o f these three hym ns w ere
obviously created later, their connection with the stanzas not being
evident from the text itself.511
The general instructions require that yantras be infused with life in
the rite o f pranapratistha, which is also perform ed on statues o f
deities. According to R aghavabhattas 15th-century com m entary on
the S arad atila k a and texts such as the M a n tra m a h o d a d h i, the
pranapratistha rite entails that certain mantras are inscribed in the
yantras. T hese m antras can be seen in some yantras w hich are
prepared on perm anent m aterials, such as metal plates. T he two
ramaraksayantras reproduced in Buhnemann 1983: 93 and 107 also
contain them. The mantras include the syllable hsauh, which repre
sents the soul (jiva) o f the yantra; the syllables hamsah so 'ham ( I
am that goose), which represent its life breath (pnina); the vowels
i/i, w hich represent the y a n tra s eyes, and the syllables u/tJ, which
represent its ears; and the seed syllables lam ram m ain ksa m vain
y a m sam ham lirim am o f the ten directional guardians, beginning
with Indra in the east, which represent the heart ( h relaya) o f the
yantra. In addition to the p rrin a p ra tisth a m a n tru , the fo llo w in g
yantragayatri, an im itation o f the w ell-k n o w n gayafri (sfiv i(ri)
mantra, found in Rg-Vcda 3,62.10, is inscribed in circular form:
yantra raja y a vidm ahe varapradaya dhim ahi /
tan no yantrah pracodayat / / 5'
5K R e g a r d i n g the c o n n c c t i o n b e t w e e n the y a n lr a s and the s t a n z a s o f th e
S a u n d a ry a la h a ri, A n a n ta k rsn a astrl c o m m e n ts: T h e r e se e m s lo be s o m e m ystical
c o n n c c tio n b etw een cach sloka and its BIjakshara. But it is not intelligible; no r has
any o f the P rayoga Karlas explained the s a m e (introduction lo his translation o f the
Saundaryalahari, 1957: 13).
SJ Cf. R a g h a v a b h a t t a s c o m m e n ta r y , p, 519, 3 3 - 3 4 on S ar.uiatilaka 24.2 . T h e
y a n tra g a y a tri a p p e a rs w ith m inor variants in m any texts. B ra h m fm a n d a g iris 6 a k la nandatarangini, p. 257, 13-14 gives the following version:
om
yantranljaya vid m a h e m ahayantraya d h im a h i I
M A N D A L A , Y A N T R A A N D C AKRA
39
We know the king o f the yantra; we think o f the bestow er o f boons.

Therefore may the yantra inspire us.
T e x ts such as B r a h m a n a n d a g ir is 16th -c en tu ry S a k ta n a n d a tarahgini, pp. 264, 6ff. also enjoin that certain purificatory rites
( samskara) be perform ed on yantras, ju st as they are perform ed to
purify mantras.
In this section I have suggested a classification o f yantras into
three types: 1 yantras for establishing a foundation, functioning as
supports for ritual implements; 2 yantras em ployed in regular Tantric
worship; 3 yantras em ployed in optional desire-oriented rites. The
three types o f yantras can be distinguished according to their ritual
functions. The first and second types have similar geometric designs
but usually no mantras inscribed. Yantras o f type 2 usually consist o f
m ore com plex geom etric designs than type 1 yantras. B oth types
differ in ritual function. The first type is used as a support for objects
in rituals, w hile the second type is the m ain object o f w orship.
Yantras o f type 1 are similar in function to the aforem entioned seatm an d alas (see B ru n n ers category 1 described in section 1.2) and
are occasionally also referred to as m andalas (see 2.2.1). B ut they
differ from B r u n n e r s s e a t-m a n d a la s o f c o w -d u n g and sim ilar
materials w hich have no clearly recognizable structure. Yantras o f
type 3 are u sed in d e s ire -o rie n te d m ag ical rites, u s u a lly have
inscribed mantras and may have unusual designs.
2.3 S o m e C onstituent Parts o f Yantras
Y antra and m andala designs com m only feature a triangle and/or a
hexagram, inscribed in one or several lotuses (padm a) o f four, eight,
10, 12, 16, 100, 1000 or more petals ( dala). The lotus petals are often
surrounded by one circle or three concentric circles (vrtta) and a
square (caturasra) w ith som etim es three nested lines. In yantras o f
the Saiva and akt:a traditions the lines o f triangles or a square may
be formed by tridents w hose prongs project beyond the lines o f these
shapes (see Colour Plate 1).
The m ain deity is w o rsh ip p ed in the centre o f the y an tra at a
p oint ( bindu) which may be visible or remain invisible/unmanifest,
while his/her retinue is w orshipped in various parts o f the structure
(see Illustration 6). T hese parts include the angles (asra) or co m ers
tan n o yantrah p ra co d a ya t //
40
GUDRUN BUHNi-MANN
( ko n a ) of a triangle or hexagram, the points of intersection (saindhi)

o f two t r i a n g l e s , t h e lotus petals (dala) and (he tips o f lotus petals
( dalagra). The most important surrounding deities or em anations are
invoked into the parts o f the yantra closest lo the centre. One obvious
advantage o f a yantra compared to an icon is that a yantra allows for
the deities who surround the main deily in enclosures (iivarann, iivrti,
literally covering or veil) to be worshipped in it as well.
The structural elements o f yantras vary, as do the interpretations
given to these elem ents. Som e im portant constituent parts are
described in the following, together with examples o f interpretations
from texts. Most descriptions and interpretations o f the constituent
parts o f yantras found in the literature concern the titiymUni or
ncakra, the most important and influential o f yantras.'1
Preliminary studies o f the constituent parts o f yantras are found in
the works by A. Danielou. The authors Hindu Polytheism , 1964:
35 1 -3 5 4 contains an enlarged and revised version o f the scction on
yantras printed in D anielous older French edition o f the book, Le
polytheisme hindou, 1960: 525-539. Danielou docs not indicate the
sources o f his interpretations o f the yantra designs clearly, but it can
be inferred that he draws on articles in H indi.1'2 In his later work on
the Hindu temple (Danielou 1977: 26-28 [2001: 37-38]), the author
takes up the discussion o f the constituent parts o f yantras once again.
D an ie lo u s interpretations o f yantra constituents continue to be
U) Special term in o lo g y is used in c o n n e c tio n with (he Sricnkru. T h e three circles

outside o f the ov erlap p in g triangles are referred to as throe girdles or belts ( valaya):
a point o f intersection b e tw e e n tw o lines is called sum dhi\ a point o f intersection
b etw een three lines is kn o w n as m a rm n n ( vital p o i n t ), and a point o f intersection
b etw een a sa m d h i and m a rm n n is term ed g ranthi ( k n o t ); sec B h a s k a r a r a y a s S ctubandha, p. 31, 9 and U m a n a n d a n a th a s Nityotsava, p. 64, 4 -5.
fil D e ta ile d n o te s on d iffe re n t i n te r p r e ta tio n s that have b e e n g iv e n to the
c o n s titu e n ts o f the Grictikra and a critical ev a lu a tio n o f these in te r p r e ta tio n s arc
provided b y Pran av an an d a < I9 7 7 > .
D c v a ra j V i d y a v a c a s p a t i s article en titled T a n tr a m e m yantra au r m a n t r a ,
p rin ted in the m a g a z in e Kalyfin, p u b lish e d by the Gita Press ( G o ra k h p u r), Sakti
ai'ika, 1934: 3 8 7 - 3 9 7 , is cited in D anielou 1964: 353. T h e original article does not
indicate the so u rc e s on which D cvaraj V id y a v a c a s p a tis interpretations are based.
D an ie lo u 1977: 26 (2001: 37) refers to an articic by Da bra 1, entitled isriyantra ka
s v a r u p , printed in Sakti ankn, 1934, p. 5 9 2 - 6 0 9 . Danielou e rroneously refers to the
author as K alik a-p rasad a Dabriil and specifies the page n u m b e r as 591. H o w ev er,
the a u th o r s n a m e is Lalitiiprasad Dabral and the article is fo und on pp. 5 9 2 - 6 0 9 . A s
the title indicates, Dabriils in terpretation is c o n cern ed with the c o n stitu e n ts o f the
sriyantra,
M AN DAL A, Y AN TRA A N D C A KR A
41
influential and are partially adopted and summ arized in Johari 1986:
52 and Bunce 2001: 27-29.
2.3.1 The Point (bindu)
The point (b in d u ) is located in the centre o f the yantra and may be
visible or rem ain invisible. It is often interpreted as the principle
fro m w h ic h all form and c r e a tio n r a d ia te s (fo r e x a m p le ,
Shankaranarayanan 1970: 29). Verses 11-12 o f the first chapter o f
the Y oginlhrdaya describe the point as throbbing consciousness
whose supreme nature is light and which is united with the flashing
flow [ o f divine p o w er], the seat (baindavasana) w h ic h is the
[birthjplace o f the flow made up o f the three m a trka s (Padoux, p.
241). A ccording to D anielou 1964: 351, the point represents the
elem ent ether. T he m ost likely seventh-century G anesapurvatapaniya-U panisad, section 3 equates the central point with the void o f
space.
2.3.2 The Triangle (trikona, tryasra)
The triangle is a com m on constituent o f yantras. It can be either
downward-oriented or upward-oriented, and less frequently oriented
toward the right or left sides.63 The dow nw ard-pointing or inverted
triangle is know n as a symbol o f the female pubic triangle and the
female sex-organ or w om b (yoni, bhaga)M The letter e is identified
with it because o f its triangular shape (in certain Indian scripts).65
This triangle is known as a symbol o f the feminine in other cultures
as well. In Buddhist Tantric texts the dow nw ard-pointing triangle is
referred to as the dharm odayaldharm odaya, the origin o f existents
(dhanna).*6 This triangle is visualized in sadhanm as the place in
which everything originates. The dow nw ard-pointing triangle also
symbolizes w ater/ 7 This symbolic significance is know n from other
B u n c e 2001: 28 co n sid ers triangles w h o s e apexes p o in t to the left or right sides
as constituent parts o f yantras.
M See, for ex a m p le , J a y a r a t h a s c o m m e n ta r y on T a n tra lo k a 3.94. Cf. also B u d
dhist texts quo ted in W a y m a n 1973: 172.
1,5
B. B h a tta c h a ry y a h a s d iscu sscd this issue in m o re detail in his fo r e w o r d to the
s eco n d edition o f the J a y a k h y a - S a m h i t a (1967: 30 ) B e c a u s e o f its s h a p e th e e is
called the w o m b o f the w o r l d (ja g a d y o n i) and is referred to as a trian g le (tryasra)cf. also ibid., Figure 1, p. 34+.
"l F o r so m e re m a rk s on the d h a rm odaya/dharm odayS, sec B a h u lk a r 1979.

" See, for ex a m p le , the V astu su tra-U p an isad , cited in B a u m e r 1986: 56.
42
G U DR UN BUIIMEMANN
cultures as well, for which the downward-pointing apex suggests the

direction o f falling rain. D a n id o u 1977: 26 (2001: 37) fu rth e r
explains the downward-pointing triangle as a symbol o f Visnu.
Tantric texts com m only describe the reverse triangle, that is, a
triangle sitting on its base with its apex upwards, as the sym bolic
shape o f the element fire,'lKThe apex o f the upward-pointing triangle
indicates the direction o f the flame. In N epal, u p w a rd -p o in tin g
equilateral or isosceles triangles cut into stone or metal are fre
quen tly seen in shrines and tem ples. The tria n g u la r h o le 69 is
considered a symbol o f the Nevar god o f music, dance and dram a,
N asahdyah, who is sometim es identified with NarteSvara or N rtyanatha.
In connection with the irlcakra, authors such as B haskararaya70
refer to the dow nw ard-pointing (a d h o m u kh a ) triangles as S akti
triangles and the upw ard-pointing (urdhvum ukha) triangles as fire
( vahni) or Siva triangles. The inverted triangle is also ta k e n as
representing prakrti; the upright triangle, purusa (D anielou 1964:
352). Both types o f triangles are intertwined in the hexagram (see
2.3.3). In yantras o f Kali, five triangles appear in the centre. In other
traditions, triangles are represented with a protruding gate on each
side,71 These gates are identical in shape with the T-shaped gates o f
the outer square o f yantras (see section 2.3.8).
The triangle is naturally connected with the sym bolism o f the
num ber three. Its three lines are usually interpreted as tripartite units
(most com m only, metaphysical concepts). Thus Shankaranarayanan
interprets the lines o f the central or primary (m u/a) triangle (w hen
understood as the k a m a k a la n) in the gricakra as representing the
pow ers (gakti) o f will (iccha), cognition (jnana) and activity (kriya)
(1970: 37), following an interpretation already attested, for instance,
in Jay arath as com m entary on Tantraloka 3,94. In another context
Shankaranarayanan 1970: 38 interprets the lines o f the triangle as
C f . s for e x a m p le , a ra d a tila k a 1.23cd, w h e re a triangle with sv a stika s r e p r e
sents fire.
m F o r m o r e in fo rm a tio n an d illu s tra tio n s o f the t r i a n g u la r n iisa h h o les, se e
W e g n e r 1992: 126, F igure 1 an d K olver 1992a: 214, Illustration I.
711 Cf. B h a s k a r a r a y a s c o m m e n t a r y S e tu b a n d h a , p. 31, 2 - 3 (c o m p o s e d in 1741
A .D ,) o n N ity a so d a îk a rn a v a 1.31 (the text is c onsidered to be a part o f the Varna*
kegvara-Tantra) and U m a n a n d a n a th a s Nityotsava (1745 A.D.), p, 64, 6.
71 Cf. the yantra o f C h in n a m a s ta described in K u b jik a-U p an isad 17.5.
72 F o r a re c e n t d iscu ssio n o f this te rm and possible translations, see W h ite 1998:
176ff.
M A N D A L A , Y AN T RA A N D C A K R A
43
representing the three g u n a s or the three states, w aking (jagrat),

dream (svapna) and deep sleep ( s u su p ti) (1970: 38). The GaneSapurvatapaniya-U panisad, section 3 interprets the lines o f the central
downward-pointing triangle in a yantra o f Ganesa as the three worlds
and the three Vedas.
D. C h atto p ad h y ay a 1973: 3 0 0 -3 0 1 asserts that not o nly the
triangle inside o f yantras but the yantra in general represents the
female reproductive organ w hen he writes: A Tantrika, w h e n he
really confides in you, will frankly confess that these diagram s are
but representations o f the female organ. In support o f this statement,
he cites B handarkar 1965: 140, w ho m akes the follow ing rem ark
about the ritual w orship (puja) o f the Sricakra: [The C akrapuja]
consists in the worship o f a picture o f the female organ draw n in the
centre o f another consisting o f a representation o f nine such organs,
the w hole o f w hich forms the Sricakra. In a som ew hat generalized
statem ent D. C h atto p ad h y ay a 1973: 301 adds th at th ere are in
Tantrism various yantras... bearing different names... b u t the essen
tial feature in all o f them is the same. It consists in the representation
o f the female organ either by the picture o f a lotus (padm a) or b y the
diagram o f a triangle, usually by both. This author is correct w hen
he observes that both the triangle and the lotus are sym bols o f the
female reproductive organ, and that both are im portant constituent
elements o f yantras. But he goes too far w hen he takes every yantra
as a representation o f the female organ. This claim has rightly been
challenged by S. Chattopadhyaya 1978: 81, w h o em phasizes the fact
that not all yantras contain triangles.
2.3.3 The Hexagram (satkona, sadara, tara73)
T he hexagram consists o f two equilateral triangles w ith the same
centre b u t pointing in opposite directions, u s u ally u p w a rd s and
dow nw ards. The apexes o f the tw o triangles o f the h ex ag ram can
also be oriented to the right and left sides.74 The triangles are show n
either lying one on the other or intertw ined w ith one another. The
dow nw ard-pointing and upw ard-pointing triangles (see also 2 .3 .2 )
sym bolize the sexual union o f the fem ale and m ale principles, o f
13
T h e w o rd s ta r (tara) ap p e a rs as a s y n o n y m for sa tk o n a in A h i r b u d h n y a - S a m hita 26,5 and P a ram eS v ara-S am h ita 23.29, as B e g le y 1973: 85 notes.
74
F o r tw o illu stra tio n s, see, fo r e x a m p le , E n c y c l o p a e d i a J u d a i c a ( N e w Y o rk :
M a c m illa n C o m p a n y , 1971), v o lu m e 11: 69 0 , F igure 2; an d 69 3 , F ig u re 10.
44
GUDRUN BUHNIMANN
Sakti and Siva, o f water and (Ire,75 In Buddhist Tantrism, the w o rd

evam is thought to be represented by two intertwined trian g les,
symbolizing the union o f insight (pnijnii) and m eans (upaya). T h e
triangular shapes o f e and vn in certain Indian scripts lend them selves
to such an interpretation.7' In descriptions o f the sym bolic shapes
(mandala) o f the elem ents (bhuta), the hexagram represents th e
element w in d .77
In the hexagram the deities are often worshipped at the points o f
intersection o f the two triangles, while in the eight-petalled lotus they
are worshipped in the petals, which ideally face in the cardinal and
intermediate directions. Occasionally a six-pointed star7* or a sixpetalled lotus, such as the va/ra-lotus, can replace the hexagram in
rituals (Heilijgers-Seelen 1994: 131). Like these objects, the hex a
gram is equated with sextuple concepts and groups, T hus GaneSapurvatapaniya-Upanisad, section 3 interprets the hexagram in a y a n
tra o f Ganesa as representing the six worlds and seasons.'7q
T he hexagram has been used for decorative purposes o r as a
magical sign in m any civilizations around the world. It is also
75
In section 2,3.2 the sym bolism o f the tw o triangles is explained: the d o w n w a rd p o in tin g tria n g le s y m b o liz e s the fem ale p rin cip le and w ater, w h ile the u p w a r d p o in tin g tr ia n g le s y m b o liz e s the m ale p rin c ip le and fire. In a l c h e m y , the t w o
t r ia n g le s o f th e h e x a g r a m also r e p re s e n t the union o f fire and w a te r. F o r th e
sym bol ism o f the w a te r and fire triangles in the Tantric ngnihoirn ritual o f N ep al, see
Witzcl 1992: 788,
* Cf. S am p u ta-T an tra, c h ap ter 4 (= Elder 1978: 109 [text], 189 [translation]); cf,
also W a y m a n 1973: 1 7 2 -1 7 3 , w h o discusses three m e a n in g s o f eva/ii, a n d K o lv e r
1992b. K o lv e r d is c u s s e s the s h a p e s o f the letters c a n d the va, w h ic h w e r e
rem in isc e n t o f d o w n w a r d - p o in tin g and u p w a rd -p o in tin g triangles aro u n d the sixth
c entury A.D., and w ere v isualized as intertw ined to form a h ex ag ram . T h e nasal o f
eva m c o rre s p o n d s to the central p oint (b in d u ) inside the h e x a g ra m . W h e n V a jra y ogini is d e sc rib e d as situated in c r a m this m e a n s that she is visu alized in sid e a
h e x a g ra m . In addition, the syllabic vn is the seed syllable o f the w o rd vajra, w h ic h
can signify the penis (E nglish 2002: 150).
77 T h e s y m b o lic s h a p e s o f the o th e r e le m e n ts are a c c o r d in g to a r a d l t i l a k a
1.23 -2 4 : a square with th u n d erb o lts (vajra) - the earth elem ent; a lotus w ith a h a lf
m o o n - w a te r; a trian g le with svastikas - fire; a circle w ith six dots, th a t is, a
h ex ag ram - w ind; and a circle - ether,
n F o r N e p a le s e c o in s o f the M alla period s h o w in g the s ix -p o in te d star, see
R h o d e s /G a b ris c h /[ della R o cch ctta 1989, Plate 31, no, 724 rev., Plate 33, no. 8 2 6
rev.
w F o r interpretations o f the sa tk o n a current in N epal, sec Joshi 1981 an d D e e p
1993: 9 8 - 1 0 0 . Joshi s u m m a r iz e s various interpretations w h ic h id entify th e h e x a
g ra m w ith w e l l - k n o w n g r o u p s o f six, such as the six s y s t e m s o f p h i l o s o p h y
(dartiana).
M AN D AL A , Y AN TRA A N D C A K R A
45
known, for example, as M agen David, the Shield o f D a v id or as the

Seal o f Solom on.,so It appears on the inside o f Hindu yantras and is
also seen in Buddhist mandalas o f V ajrav arah l/V ajray o g in i.81 The
hexagram is a decorative m o tif in Islamic m onum ents o f N orth India,
Its centre features a point ( bindu), a lotus or a dancing peacock (Nath
1975-1976: 7 4 -7 5 ).112
In Nepal, the hexagram is frequently represented and considered
an a u sp ic io u s sym bol o f the goddess by b o th B u d d h is ts and
H indus.85 It is sometimes found superimposed on the latticework o f
window s o f temples or shrine room s.84 The point (bindu) in its centre
often bears an image o f the deity w orshipped in the tem ple or shrine.
The hexagram also appears in mediaeval coins o f India85 and, along
w ith other geom etrical designs, on N ep alese coins o f the M alla
period .86 Auer/G utschow 1974: 106 report that the hexagram is also
called grlmandala in Nepal. It is considered a sym bol o f education,
science and o f the goddess Sarasvatl, and therefore becam e the logo
o f Tribhuvan University, colleges and other educational institutions.
811 F o r in fo rm atio n on the h e x a g ra m in m e d iaev al E u ro p e an d the N e a r E ast, see
E n c y c l o p a e d i a J u d a ic a ( N e w Y o rk : M a c m i l l a n C o m p a n y , 1 9 7 1 ), v o l u m e 11:
6 8 7 -6 9 7 .
F o r a 13th-century T ibetan m a n d a la dep ictin g the g o d d e s s V ajra v a rah l standing
in side a h e x a g ra m , see B 6guin 1990: 173; o th e r e x a m p le s f r o m T ib e t are f o u n d in
R h i e / T h u r m a n 1999: 118 and 44 0 , K o s s a k /S in g e r 1998: 97 an d L e id y / T h u r m a n
1997: 105. F o r a p h o to g ra p h o f a h e x a g r a m re p re s e n tin g V a jra y o g in i in T ibet, see
Sto d d ard 1999, Figure 30, T h e question n e e d s to b e e x a m in e d w h e th e r the h e x a g ra m
in m a n d a la s o f V a jra v arah l/V ajray o g in i indicates the m a n d a t e 's S aiv a origins. W h ile
th e h e x a g ra m is c o m m o n ly f o u n d in H in d u m a n d a la s and y antras, w ith intersecting
u p w a rd -p o in tin g and d o w n w a rd -p o in tin g triangles alread y ap p earin g in th e gncakra,
in B u d d h is t tra d itio n s th ey a p p e a r e s p e c ia lly in m a n d a l a s o f V a jr a v a r a h i/V a jr a yogini, T h e n a m e V ajrav arah l itse lf indicates a B u d d h is t v ersion o f the B ra h m a n ic a l
g o d d ess Varahi.
n A c c o r d in g to N a th 1 9 7 5 - 1 9 7 6 : 78, th e h e x a g r a m is a ls o f o u n d in In d ia n
tem p les, e s p ecially in R a ja sth a n , w h e re it is b e lie v e d to h a v e b e e n a s s o c ia te d w ith
the w o rsh ip o f Siva and Sakti.
B a n g d e l 1999: 4 6 4 w rites that : [ s p e c i f i c a l l y , th e d o u b le -tria n g le d ya n tra in
th e T a n t r i c tr a d itio n is a u n iv e r s a l s y m b o l fo r th e g o d d e s s s g e n e r a t i v e a n d
de stru c tiv e p o w e r s ... and 1999: 540, note 118 that: [t]he y a n tr a s y m b o liz e s the
se a t/p r e s e n c e o f the g o d d e s s .... G u ts c h o w 1982: 97, P la te 105 a s s u m e s th a t the
h e x a g ra m represents T ripurasundari.
84 F o r p h o to g ra p h s, see, fo r e x a m p le , B e rn ie r 1978: 259, P late 11 and G u ts c h o w /
K o lv e r/S h resth acary a 1987: 203, no. 416.
S5 See, for ex a m p le , S ircar 1968, P la te xxiii, no. 7 obv ., C o o m a r a s w a m y (1 9 2 7 )
1985: 45 and Smith 1972, v o lu m e 1, Plate xxx, no. 14 rev.
S ee R h o d e s / G a b r i s c h / f d e lla R o c c h e tta 1989, P la te 2 0 , nos. 3 8 2 - 3 8 4 rev.,
Plate 21, nos. 4 1 2 - 4 1 7 rev., etc.
46
GUDRUN BOIIMKMANN
It is found on the k in g s headgear and on N ep alese o rd e rs ,

decorations and medals. Bernier 1978: 252 assumes that this design
has been borrowed from (he Islamic tradition o f North India, b u t
Begley 1973: 84 considers it most likely that the hex ag ram w as
already an important Tantric symbol before the Islamic h e x a g ra m
becam e w idely circulated in India, More research would, b e
necessary to trace the history o f this important motif. The hexagram
is an im portant m o tif in the later Tantric iconography o f the
Sudarânacakra, V isnu's sudiirsnnn wheel, which often has a y an tralike structure inscribed on it."' This .structure consists o f a hexagram
surrounded by one or more rings o f lotus petals. The personification
o f V is n u s wheel, the Sudarânacakrapurusa, is show n inside (or
standing against) the structure/"* The reverse side of these icons often
depicts Narasimha. If combined with a representation o f the Sudarsanacakra, this deity may be represented inside an upw ard-pointing
triangle,^
2.3,4 The Pentagram (punenkomt)
The five-pointed star, the pentagram, pentacle or Star o f Solom on, is
less com m only found in yantras. It is known as a sym bol also in
o ther civilizations. The pentagram is a constituent part o f som e
yantras o f Guhyakuli (see Colour Plate I), since the num ber five has
special significance for the goddess Kali. It is again found on N epa
lese coins o f the Malla period as well as being the logo o f som e
educational institutions in modern Nepal. Daniclou 1977: 28 (2001:
38) equates the pentagram with 6 iva as the destroyer o f love and
lust (smarnhara). This interpretation is not compatible with another
statem ent by the same author (Daniclou 1964: 353) according to
w hich the pentagram signifies love and lust and the p o w e r o f
disintegration.
117 D etails on the m u lti-a rm e d S u d a r.iin a icons can lie found in B e g le y 1973:
8 4 -9 2 .
S ec B e g le y 1973: 90 (with Figure 70) for a ca. 17th-century b ronze statue o f
the 16-arm cd S u d a rS a n a e a k ra p u ru s a in the S ri-K a ln m c k n p c ru m iil T e m p l e , T iru m o h u r , M a d u ra i D istrict. T h e d e ity is s ta n d in g ag ain st a h e x a g r a m w h i c h is
surro u n d ed by rings o f eight, 16, 32 and 64 lotus petals, T hese lotuses arc su p p o rte d
by an eight-petalled lotus (see also Illustration 1 in Rastelli \s contribution),
w S ec B e g le y 1973: 8 8 - 8 9 and Figures 67 and 69,
MA ND AL A , Y AN TRA A ND C AKRA
47
2.3.5 The Octagon (astakona, astara)

T he octagon appears less frequently as a constituent part o f yantras
and can be formed in several ways. A com m on m ethod to obtain an
octagon is to draw two crossed or intersecting squares, T h e tw o
overlapping squares appear as a symbol in various c iv ilizatio n s .91
The symbolism of the octagon, like that o f the eight-petalled lotus, is
connected with the eight directions. The octagon appears on Indian92
coins and on Mai la coins o f N e p a l.93 It also decorates a w indow in a
religious building o f the T rip u resv ara tem ple c o m p lex in K a th
m andu .94
2.3.6 The Lotus
T he s y m b o lism o f the lotus is discu ssed in se c tio n
connection with mandalas.
1.3.1
in
2.3.7 The Circle

O ne circle or three concentric circles frequently surround the inner
structure o f yantras. According to Danielou 1964: 352, the y a n tra s
o uter circle, given its revolving te n d en cy , ch ara c te riz e s m a n i
festation, A m ong the sym bolic shapes (m a ndala) o f the elem ents
( bhuta), the circle represents ether.95
2.3.8 The Outer Square
T he circle or circles in a yantra are usually surrounded b y an outer
square which often consists o f three nested lines. The square, w hich
also appears on the outer part o f m andalas, is called earth h o u s e
( bhugrha), earth city or earth citadel ( bhupura),% since the square
>n See R h o d es/G ab risch /'l' d ella R o c c h c tta 1989, P la te 21, nos. 4 0 6 - 4 0 9 rev.,
Plate 28, nos. 5 6 6 - 5 6 7 obv.
,J1 See, fo r e x a m p le , E n c y c lo p a e d ia J u d a ic a ( N e w Y ork: M a c m i l l a n C o m p a n y ,
1971), v o l u m e 11: 6 8 9 - 6 9 0 , F ig u r e 3, for this d e s ig n as u s e d in 1 3 th - c e n tu r y
G erm an y .
92 See S irc a r 1968, Plate XIX, no. 11.
Sec R h o d e s / G a b r i s c h / t d ella R o c c h c tta 1989, Plate 15, nos. 2 7 4 - 2 7 5 ; P late 20,
nos. 3 9 1 - 3 9 2 , 396, Plate 21, no. 400, etc.
<MFor an illustration, s e e Gail 1 9 8 4 - 1 9 8 8 , v o lu m e 2, Plate X V I, no. 2 and p. 97,
an d G u ts c h o w /K o lv e r /S h r e s th a c a r y a 1987: 203, no. 41 7 . T h e t e m p l e is c u rre n tly
bein g renovated.
1,5 See, for ex a m p le , the description in Saradatilaka 1 .2 3 -2 4 .
*
F or a d escrip tio n o f the d iffe re n t stag es o f c o n s tru c tin g a bhupura, see B u n c e
2001: 2 0 - 2 2 ; for variations in the shapes o f bhupuras, see B u n c e 2 001: 2 3 - 2 5 .
48
GUDRUN UUHNI-MANN
is a symbol o f the earth. Among the symbolic shapes o f the elem ents,
a (yellow) square represents the e a r t h , 'T h e symbolism o f the square
is connected with that o f the num ber four and the four card in al
directions. The square has a T-shaped gate {dviira) in the cardinal
directions. Like the sides o f a square, the gates are equaled w ith
groups o f four, as in M irras interpretation o f an unidentified Sakta
yantra (1959: 482- 483) in which they arc taken to represent the four
Vcdas. Pahcaratra Samhilas interpret the three nested lines o f the
outermost square as representative o f the three constituents (guna) o f
primary matter (prakrti) in the Samkhya system, namely saliva, rajas
and lamas. This interpretation is also attested for the three nested
lines o f the outer square o f the bhadram andalas o f the Sm arta
tradition, which are white, red and dark and symbolize respectively
saliva, rajas and lamas.VHThe square also appears on Nepalese coins
o f the Malla period.4''
Influenced by C.G, Jung, Dehcjia 1986: 42 w ould have us
recognize the mediaeval alchem ists m otif o f the squaring o f the
c irc le ( quadratura circuit) in South Asian m andalas and yantras
which feature a circle surrounded by a square. Such an interpretation
docs not find support in the South Asian traditions. M oreover, in
m ediaeval E uropean draw ings o f this m o tif the square alw ays
touches the circle. In contrast, South Asian mandalas and yantras
show a significant gap betw een the circle and the square that
surrounds it.
Gupta 1988: 39-41 offers the hypothesis that the square enclosure
o f mandalas represents a Tantric analogue to the sacred fire altar and
thus has Vedic origins (see section 1.4).
2.3.9 Other Structural Elements
It is not possible to survey all constituent elements o f yantras, which
include shapes such as the heptagon.1One unusual design is found
in a yantra o f G uhyakall from Nepal (see C olour Plate 1). This
'n Sec, for exam p le, Saradatilaka 1.24ab, where a square with th u n d e rb o lts ( vtijrn)
represents the earth element,
*
T h e s y m b o lis m o f th e se three c o lo u rs is frec|iiently re fe rre d to; for m o re
in form ation, see G o u d ria a n 1978: 166-175 and the S uprabheda, q u o te d in B runner,
p. 173, note 49.
Sec R h o d c s /G a b ris c h /f della Rocchctta 1989, Plate 20, nos. 382-3K 3 o b v 390
rev., Plate 22, no. 4 4 6 obv.
Ii(l S o m e additional shapes arc described in Bunce 2 0 0 ] : 2 7 -2 9 .
M A N D AL A , Y AN T RA A N D C A K R A
49
com plex yantra features a pen tag ram (see 2.3.4) in the centre,
surrounded by a shape with nine corners (navakona), T he n in e
angled shape is situated within two overlapping squares (see 2.3.5),
which in turn are surrounded by a lotus design with rings o f eight, 12
and 16 lotus petals. Outside o f this is a square whose four sides are
m ade up o f tridents (trisula) surrounded by skulls. F our enclosures
surround this structure, w hich feature (from the inside to the
outside): ( 1) w ater (here representing the O cean o f B lood [somfoda]), (2) the eight cremation grounds (smasana), (3) skulls (m unda)
and (4) flames ( vahnijvala/).U)I A circle o f flames is also kn o w n to
surround Tibetan Buddhist mandalas, some o f which include a circle
representing crem ation grounds. A ccording to M acd o n ald /V erg ati
Stahl 1979: 91, crem ation grounds (sm asana)'02 have been part o f
Buddhist as well as Hindu mandalas and yantras in N epal since the
15th century.103
3 Cakra
T h e term cakra, c ir c le or w h e e l, has several p rim a ry and
secondary m eanings.IM
101 T h e S a n s k rit te rm s arc ta k e n fro m a d e s c rip tio n o f G u h y ak alT s y a n tr a in

P uraScaryarnava, pp. 1149, 22
1150, 8. T h e P uraScaryarnava quotes the M a h ak alaSam hita, a basic w ork ad v o c a tin g the w o rsh ip o f G u h y a k a ll, w h ic h w a s m o s t likely
written in M ithila (M ichacls 1996; 319).
T h e o u te r e n c lo s u re s o f G u h y a k a l i 's yantra also a p p e a r in o th e r N e v a r H in d u
yantras. T h e p a 6 c im a k a rin a y a n tra , in clu d e d in the B o o k o f P ic tu re s C o n ta in in g
Images and Y a n tr a s , preserved in the B h arat Kala B h a v a n , V aran asi, no. 10054, and
dated 1764/65 A.D. (see neg ativ e 312 o f the A m e ric a n Institute o f In d ian Studies,
G u rg a o n ), featu res e n c lo s u r e s w ith (1) a floral d esig n (?), (2) a w a t e r d esig n , (3)
skulls and (4) flames.
m F or a study o f cre m a tio n g ro u n d s, see M eisezah l 1974 and P o tt ( 1 9 4 6 ) 1966:
7 6 -1 0 1 ( T h e S a c re d C e m e te r ie s o f N e p a l ) an d the d is c u s s io n in E n g lis h 2002:
1 3 6 -1 4 3 . F or the cult o f m aanas, see T su d a 1990. T su d a 1990: 98 notes th at the
eight c rem atio n g ro u n d s a p p e a r on the o u te r structure o f m a n d a la s o f th e S a m v a ra
c lass o f T a n tr a s , w h ic h d a te s b a c k to the e n d o f th e e ig h th c e n t u r y A .D . D e
M a llm a n n 1986: 42 refers to c re m a tio n g ro u n d s in the m a n d a la s o f H e ru k a /H e v a jra ,
S am v ara and Vajravarfihi.
103 Pal 1 9 7 4 - 1 9 7 8 , part 2: 93 is o f the o p in io n th at g e n e r a lly no s c e n e s o f
cem etcries are included in Hindu m a n d a la s.
I(MF o r a d isc u ssio n o f the te rm c a k ra , see also B ru n n e r , p. 163; T o r z s o k , pp.
180 -1 8 3 contrasts the term s cakra and m andala.
50
1.
2.
3.
4.
5.
( i U D R U N BUI I N T M A N N
C a k ra can refer to a wheel as (lie centra! part o f a m andala

structure, to whose hub and spokes deities are assigned. In this
function, the wheel can either substitute for a lotus or appear in
combination with it (cf. section 1.3.1).
The term cakra can refer to a group or circle o f deities invoked
into a mandala or yantra structure.
Som e later authors appear to use ca k ra synonym ously with
yantra and m andala.105
The word cakra also refers to a diagram/tabular device em ployed
in ritual.
A w ell-know n use o f the word cakra is with reference to the
w h e e ls or lotuses believed to be located in the human body.
In the following, only the meanings 2, 4 and 5 are discussed.

3.1 Cakra as a Circle o f Deities
The term cakra refers to a group o f deities invoked into a mandala or
yantra.106 At the same time, the term also denotes the support for
these deities in the form o f a specific surface."1' This is most likely
the reason why the word cakra appears synonymously with yantra
and m andala in later texts a use o f the term that requires further
investigation. It also explains why the parts o f the .tiicakra, which
consist o f a variety o f shapes such as triangles and lotus petals, are
referred to as the nine cakras.
1115 See B ru n n e r, p. 163 and P adoux 1987a: 4, T o rz sn k , p. 181 e x p re s s e s so m e

re s e rv a tio n a b o u t the in tc rc h a n g e ab ilily o f the te rm s cakra and m a n d a la in early
a iv a texts. T h e S r i v i d y a s s y n o n y m o u s use o f the term s cakra and y a n tra in the
iricakral iriya n tra in d icates a looser use o f these term s (cf. S a n d e r s o n 's re m a rk in
P a d o u x 1986; 33). T h e baliharanacakra describ ed in the next article is also called
baliharanam andala in sonic texts.
' Cf. A. S a n d e r s o n s rem ark rcp ro d u c c d in Padoux 1986: 33: "... the m a n d a la is
the adharah (lo c u s ) and the c a k r a ( o f d e itie s /m a n tra s ) the a d h c y a m (lo c a te d )...."
S n ellg ro v e 1959, part 1: 135 defines c a k r a sim ilarly as a circle o f divine forms o f
w hich the m andala consists.
m H c ilijg c r s - S c e le n 1994: 37 gives several in stan ces from the K u h jik a m a ta T antra, w here c a k ra refers to the scat o f a deity.
MANDAL A, YANTRA A N D C A KR A
51
3.2 Cakra as a Ritual Diagram/Tabular D evice

The word cakra also refers lo diagrams containing specific arrange
ments o f letters o f the alphabet or o f numbers. Diagrams such as the
akathahacakra, the akadamacakra, the naksatracakra and the rasicakra, which are used to determ ine whether a mantra suits a candi
date, fall under this categ o ry .M
IK Unlike the distributive d iag ram s,
namely, mandalas o f type 3 in B runners classification (see section
1.2), w hich are ritual diagram s divided into squares into w hich
divine or demonic beings are invoked to receive food offerings, these
diagrams are simply tabular devices, into which no deity is invoked.
3.3 Cakra as a W h e e l or L o tu s in the B ody
The word cakra is used with reference to the currently rather popular
Tantric concept o f the energy centres or pow er centres w hich are
believed lo be located in the body. They are referred to either as
w h e e ls (cakra) or as lotuses (padm a).m In som e system s these
cakras have lotus shapes and are populated by deities. The lotuses
may also have inscribed geometric figures (triangle, hexagram). The
num ber o f cakras and their location in the body varies according to
traditions. Some systems assume sets o f four, five, six, nine, 12, 16,
24, 27, 32 or more cakras.
The system o f six w heels or lotuses is particularly well known
and listed below acco rd in g to S arad atila k a 5.131 cd 136 and
20,66cd-67.
N um ber o f Petals
Lotus
4
muladhara
svadhisthana
6
10
manipura
12
anahata
16
vi&uddhi
2
ajna
Syllables
v a sti
ha-la
da-pha
ka-tha
16 vowels
ha and ksa
Deity
Brahma
Visnu
Rudra
Isvara
Sadasiva
Siva
Goddess
Dakini
RakinI
Lakini
Sakini
Kakinl
Hakim
I(* For an ex ten siv e treatm ent o f these diagram s, see B u h n e m a n n 1992.
"" As in the case o f a wheel c o m b in e d with or in p lace o f a lotus in the ccntrc o f a
m an d ala (cf. section 1.3.1), the te rm in o lo g ies d en o tin g parts o f a l o tu s and parts o f
a w h e e l in the bo d y are often used interchangeably.
52
GUDRUN b OHNI'MANN
The nmladhara or adhani is located at the bottom o f the spine

between the anus and the penis. The svadhisthana is at the root o f the
penis. The manipura is located at the base o f the navel, and the
anahata in the heart. The visu d d h i is situated in the throat, and the
ajfia in between the eyebrows. The thousand-petalled lotus (sahasradalapadma), also called the thousand-spoked wheel (sahasraracakra),
being located at the top o f the head, is added to the six eakras, but is
usually not counted as one o f them.
Since these cakras are associated with the process o f creation, they
are connected with the five elements (hhuta) and the mind (m anas),
and with the syllables o f the Sanskrit alphabet. Specific syllables are
inscribed on each lotus, one syllable per petal. Iiaeh lotus is presided
over by a specific deity and associated with a goddess.
This system o f the Saradatilaka differs from earlier stages o f
developm ent, which arc preserved in sections o f the 10th-century
Kubjikamata-Tantra."" In one section o f the Tantra the cakras are
assumed to be five circles o f goddesses, located below the navel, in
the belly, the heart, the throat and at the top o f the head or the
brahmarandhra."' These circles arc called the dcvicakra, dullcakm ,
matrcakra, yoginlcakra and khccarwakra, and are believed to be
populated by a larger num ber o f goddesses. In another section o f the
text, the cakras, with the exception o f the anahata, are not viewed as
lotuses with varying number o f petals and geometric figures ins
cribed in t he petals. Instead of the term cakra, the word pada is used
(Heilijgers-Seelen 1994: 38).
" " S e e H eilijgcrs-S cclcn 1990: 59 for details on cakra sy stem s taught in sections
o f the K u b jik a m a ta -T a n tra . T h e au th o r d iscu sses one system o f liv e c ak ras as r e
f le c te d in the K u b j i k a m a t a and tw o s y s te m s (n a m e ly , the u tin n i and d a k s iiu n
varieties) o f six cakras.
111 Sec Heilijgcrs-Scclcn 1994: 18 for more details.
MAN DAL A, Y AN T RA AND C A KR A
I. A yantra for attraction (a k a rsa n a y a n tra )
2. A supp o rtin g m a n d a la for the v a rd h a m vessel
54
G UD RU N BUHNEMANN
3. A yantra with the RFunaraksa,stotra inscribed
M ANDAL A, Y AN T RA A N D C A KR A
+'
4. A yantra assigned to stanza 55 o f th e S a u n d a ry a la h a ri
5. A yantra for s u b ju g atin g o n e s m a ste r
56
GUDRUN B0HNHMANN
~ i __________________ I
6. A pujayuntru o f M nhagiinapati
M A N D A L A S A N D Y A N T R A S IN S M A R T A R I T U A L
Gudrun Buhnemann
P A R T I. S E L E C T E D M A N D A L A - L I K E S T R U C T U R E S , M A N D A L A S
AND YANTRAS
1 Introduction
T his essay describes some m andalas, m andala-like structures and
y antras that are currently used in the ritual practices o f Smarta
Brahm ins in Maharastra. It draws on my observation o f rituals in that
part o f India as well as on the study o f ritual manuals, This paper is
divided into two parts, Part one is designed to introduce the reader to
com m on structures o f mandalas and yantras in the Smarta tradition.
A s an exam ple o f an early m andala-like structure, the baliharanacakra o f the vaisvadeva rite is described. This opens the discussion
on the directional orientation and basic designs o f mandalas. D escri
b in g specific m an d alas and yantras cu rren tly used by S m arta
Brahmins, I analyze the navagrahamandala, the saptasatlniahayantra
(which originally belonged to the Sakta tradition) and the rudrapithamahayantra. Part two o f this paper describes a category o f mandalas
called bhadramandalas.
Initially a b rie f explanation o f the w ord Smarta may be in order,
Smarta is a rather loosely used term which refers to a B rahm in who
is an adherent o f the Srnrti and o f the tradition w hich is based on
the S m rti. The Smarta tradition considers itself to be based on the
V edic heritage and the ancient orthodox texts in the V edic tradition,
such as the D harm asastras and the Smrtis. It claims to be neither
exclusively Siva-oriented (Saiva) nor exclusively V isn u -o rien ted
(Vaisnava), and often combines the worship o f five deities,1 Despite
the rejectio n o f T a n tric elem ents in rituals by so m e S m arta
1
2003.
F or a discussion o f the term Sm arta and o f the Sm arta tradition, see B u h n e m a n n
58
GUDRUN BUI INHMANN
authorities,2 the Smarta tradition has incorporated such elem ents,

including yantras.
2 The M andala-likc Arrangem ent in the linlilm vnncakra
The baliharanacakra* or balihamnamandala appears to be am ong the
earliest mandala-likc arrangements o f the Smartas. In current ritual
practice o f M aharaslrian Rg-Vcdins, it is a circular arrangem ent
formed by food offerings (bah) (see Colour Plate 2) placed on the
ground in the bhutayajha, pitryajha and m anusyayajha rites. These
three rites arc variously considered parts of, or appendices to, the
vaisvadeva * rite. The vaisvndcva, which is a ritual offering o f
cooked food before eating, is prescribed twice daily, in the morning
and in the evening. There is little difference between these two ritual
p ro ced u res, In current M aharaslrian practice, the co o k ed food,
usually rice, is divided into three portions. The offerings from the
first portion o f food are made to the sacred domestic fire to deities as
part o f the sacrificial worship o f deities (dcvayajha) o f the vaisva
deva rite. T he offerings o f the second and third portions o f food are
balis, that is, offerings made outside the sacred fire. They are placed,
strewn or thrown on the ground. Bali offerings arc made to deities
and other beings as part o f the sacrificial worship o f beings ( bhutayajha), and to the ancestors as part o f the sacrificial w orship o f
ancestors (pitryajna). The final offering forms part o f the honouring
o f guests (m anusyayajna), The ten deities worshipped as part o f the
devayajna are identical with the first ten o f the group o f 32 deities,5
2 T h e nyasa rite, for e x am p le, w h ich form s part o f the current S m arta puja, is
rejected by so m e as Tantric; see Biihnemann 1988a: 12!.
3 In the p rev io u s essay the term s m andala, yantra and cakra are discussed. Section
3 deals csp ccially with the m e a n in g s o f the w ord cakra. T h e term cakra in the c o m
p o u n d baliharanacakra appears to refer both to the circular layout o f the o fferings o f
rice w hich arc placcd on the g round and to the group o f deities and other b ein g s w ho
are in v o k e d w h ile setting d o w n the offerings. In so m e texts the term baliharanam andala is used sy n o n y m o u s ly with baliharanacakra but is m o re a p p ro p ria te in the
context o f the Y aju r-V c d a tradition in Mnhiirastra, w here its layout is square. In that
tradition few er deities arc invoked and som e o f their nam es vary.
4 For details on the vaiSvadcva ritual as it is dcscribcd in ancicnt texts, sec K ane
19 6 8 -1 9 7 7 , volum e 2: 7 4 1 - 7 4 8 , and G o n d a 1980: 4 1 7 - 4 1 8 .
5 I f o llo w h ere th e text p r in te d in R g v c d i y a b r a h m a k a r m a s a m u c c a y a , fols.
2 9 a . l 0 - 3 0 b . 3 . T h e n u m b e r and nam es o f b ein g s differ slig htly from those listed in
the a n c ie n t tex ts, such as M a n u - S m r ti 3 . 8 4 - 8 6 . S o m e o f th e se d if f e r e n c e s arc
MAN D AL A S A ND Y AN TRA S IN S MA RT A R IT U A L I
59
b e in g s and ancestors who receive offerings m ade as part o f the

bhutayajna. The 32 deities and beings who receive offerings m ade in
the baliharanacakra o f Ihe R g-V edins (see Illustration 1) are the
following:
(as part o f the bhutayajna, in a circle, beginning from the east:)
(1) S u r y a / (2) Prajapati, (3) Som a and V anaspati, (4) A g n i and
Som a, (5) Indra and Agni, (6) heaven and earth, (7) Dhanvantari, ( 8)
Indra, (9) the VisveDevas, (10) B rahm a,7 (11) the waters, (12) herbs
and plants, (13) house, (14) deities o f the house, (15) deities o f the
site ( vastu); (outside the circle, in the cardinal directions, in groups
o f tw o) (16) Indra, (17) In d ra s men, (18) Yama, (19) Y a m a s men,
(20) V aruna, (21) V a r u n a s men, (22) Som a, (23) S o m a s men;
(inside the circle, in the cardinal directions, beg in n in g from the
north) (24) Brahma, (25) B ra h m a s men, (26) the V isveD evas, (27)
all beings that move by d a y ;8 (outside the circle, in the intermediate
directions, beginning from the north-east:) (28) Raksasas;
(as part o f the pitryajna:)
(29) Svadhapitrs, (30) 6 yama, (31) Sabala;
(as part o f the m anusyayajna:)
(32) Sanaka and other humans.
T he practitioner presses together a small num ber o f rice grains b et
w een the fingers and the th u m b o f his right hand. In v o k in g the
deities and other beings, he first places the offerings so that they
form o f a circle. He then places some offerings inside and outside the
circle, always proceeding in a clockwise direction. This clockw ise
m o v e m e n t rep resen ts the usual order o f m o v e m e n t in rituals,
counter-clockwise m ovem ent being used, for example, in rites for the
dead.
In the mandala-like structure o f the baliharanacakra, offerings are
p laced in a circle, attention being paid to directional orientation.
H ow ever, there is no single deity in the centre. This arrangem ent
differs from later m andala structures w hich follow the concentric
dis c u s s e d in K ane 1 9 6 8 - 1 9 7 7 , v o lu m e 2: 741 ff. T h e d ia g ra m in B o u rq u in 1884: 86

includes 36 deities and other beings.
(l In the e v en in g ritual, Agni is invoked instead o f Surya.
7 B ra h m a appears again in 24.
8 In the e v en in g ritual, the beings th at m o v e b y night are invoked instead.
60
G U D R U N I3UI I NI i M A N N
pattern in which one ccnlral deity is surrounded by other entities on

the outside,
3 The M andala-like Arrangem ent in Pancayatana W orship
A m andala-like pattern with one ccnlral deity surrounded by four
other deities in the interm ediate directions or corners appears in the
Smarta pancayatana worship. The pancayatana worship focuses on
the icons or aniconic representations o f five deities, and becam e
popular in the mediaeval period. In modern times, the five deities are
the Vedic sun god Surya; the goddess called Devi or Durga; Visnu;
the elephant-headed Ganesa (Ganapati); and Siva (MaheSvara). In
earlier sources we find evidence for the worship o f Brahma instead
o f Durga or Ganesa. W orshipping these five deities is referred to as
the five-fold worship (pahcopnsana), or the puja o f the five shrines
(pahcayatanapuja). T he concept o f the five-fold w orship is also
reflected in temple architecture. Pancayatana temples in North India9
place the main deity in the centra! shrine and (bur subordinate deities
in smaller shrines at the corners o f the square. In domestic worship,
the positions o f the four other deities vary with the central d eity 10
(see Illustration 2). A contem porary tivapanciiyalana in a M aharastrian home, for example, may feature a sivalihga in the centre and the
icons o f Balakrsna (representing Visnu), the sun, G anapati (a red
stone) and the goddess (a metallic stone) respectively in the north
east, south-east, south-w est and north-west (see C olour Plate 3).
W hen GaneSa occupies the ccntrc, Visnu, Siva, Surya and the
goddess will be placcd respectively in the north-east;, south-east,
south-west and north-west.
It is remarkable that in this pattern one deity is positioned in the
centre and the four others in the intermediate rather than the cardinal
directions, as is com m only seen in mandala-like structures. Perhaps
the practitioner wanted to be able to face all the deities at the time o f
w orship. One should keep in mind that the four deities w ho are
w orshipped along with the central deity arc not considered to be
9 F o r a study o f pancayatana te m ples in North India, see Reitz 1998.

10 T h e a rran g em en t o f the deities in dom estic w orship follow s patterns p rescrib ed
in a stanza w h ic h has been attributed to B o p a d c v a s YamalaprakiiSa and is cited in
m a n y texts, such as M itram iS ras Pujaprakasa, K a m alak arab h attfi's N irn a y a s in d h u
and c o n tem p o rary ritual m anuals. For details, see Biilmemann 1988a: 5 0 -5 1 .
MAN DAL AS AND YANTRAS IN S MART A RI TU AL I
61
manifestations o f the latter, even though their position at the time of

w orship is clearly subordinate.
4 Lotus D esigns and the Navagrahamandala
A different directional orientation is followed in lotus designs, which
often form the core o f m andala and yantra structures. T h e lotus
pattern and the symbolism o f the lotus have already been described
in section 1.3.1 o f the previous essay. In this structure, the main deity
is positioned in the centre and the surrounding (avarana) deities are
placed on the petals, and sometim es also on the filaments o f the
lotus. The surrounding deities are positioned in the four cardinal and
interm ediate directions, and occasionally they are also placed at the
z en ith and nadir o f the m andala. The zenith and nadir are then
indicated by special markings placed near two o f the interm ediate
directions in the mandalas. The surrounding deities may form one or
several circles around the central deity, follow ing the co m m o n
concentric pattern. Frequently the deities are not invoked into icons
but into areca nuts (puglphala), which are readily available and can
be used repeatedly.
A com m only em ployed m andala with an interior lotus design in
the contem porary M aharastrian Sm arta tradition is the navagraha
mandala, which features the nine heavenly bodies. The na va g ra h a
m andala is constructed and w orshipped to propitiate the evil co n
stellations o f heavenly bodies and to rem ove obstacles. It is e m
ployed in the gralm m akha or grahayajna, the sacrificial w orship o f
the heavenly bodies, a preparatory rite preceding major rituals, such
as the vastuSanti or the life-cyle rituals (samskara) . 11
In this mandala, each heavenly body (graha) is typically rep re
sented by a specific symbolic shape and is prepared from grains or
coloured powders. The grains used for each shape may be a different
colour. The grahas (listed in the order o f the w eekdays) w ith their
symbolic shapes, colours and associated directions are usually:
" F or the n a v a g n ih a sa n ti and the w o rs h ip o f the nine h e a v e n ly b o d ie s, see K a n e

1 9 6 8 - 1 9 7 7 , v o lu m e 5: 7 4 9 - 7 5 5 . F o r descrip tio n s o f the g ra h a ya jn a c o m p ile d from
d iffe re n t Sanskrit texts, see D harm akoSa: S a m s k a r a k a n d a (e d ite d b y L a x m a n s h a s tri
Joshi, v o lu m e 3, pari 3, Wat: Prajna Press, 1981): 1 5 5 4 -1 6 3 9 .
GUDRUN BUHNEMANN
62
Heavenly Body
sun
moon
Mars
Mercury
Jupiter
Venus
Saturn
Rahu
Ketu
Shape
circle
square
triangle
arrow
rectangle
pentagram
bow
winnowing fan
banner
Colour
red
while
red
yellow
yellow
while
dark
dark
smoke-coloured
Direction
centre
south-east
south
north-east
north
east
west
south-west
north-west
The sun is positioned in the centre o f the mandala, which contains

another lotus. In Illustration 3 a circlc drawn outside the central lotus
contains instructions for the sequence and placement o f the heavenly
bodies. Here each heavenly body is accompanicd by one deity on
either side. The Vedic sun god Savitr appears in the centrc with A gni
and Rudra on either side. The moon is accompanicd by the waters
and Uma; M ars by ihe earth and Skanda; Mercury by V isnu and
Narayana; Jupiter by Indra and Brahma; Venus by Indra and Indrani;
Saturn by Prajapati and Yama; Rahu by snakes and Kala; and Ketu
by Brahma and Citragupta.12
The mandala is surrounded by a square called the earth c ity or
earth h o u se ( bhupura, bhugrhn), already described in section 2.3.8
o f the previous essay. The square has four protruding gates in the
cardinal directions. Inside the gates and in the four corners o f the
square, the eight directional guardians (dikpfiln) arc invoked:
Directional Guardian
Indra
Agni
Yama
Nirrti
Varuna
Vayu
Soma (= Kubera)
Isana
Direction
east
south-east
south
south-west
west
north-west
north
north-east
12
T h e d esc rip tio n o f the m a n d a l a s d eities is found in R g v e d i y a b r a h m a k a n n a sa m u c c a y a , fols. 1 3 2 a . 2 - l 3 7 a . 9 (this description is said to be based on an u nspccificd P ari& sta),
M ANDAL AS AND Y AN TRA S IN S MART A RI TU AL I
63
T h e navagrahamandala is a good exam ple o f the lotus pattern

co m m o n ly seen in m andalas and yantras w hich use a concentric
a rra n g e m e n t o f deities. The eight-petalled lotus w ith its petals
po in tin g in the directions is especially well suited for placing the
deities in their respective directions. The group o f eight (and so m e
tim es ten) directional guardians is also invoked into parts o f the outer
structures o f other yantras, such as the two yantras described below.
5 Yantra Structures
I will n ow address the structure o f two y an tras, the saptasatlmahayantra and the rudraplthamahayantra, in some detail. Both yan
tras belong to the category o f yantras which are employed in optional
desire-oriented rituals (see type 3 described in section 2.2.3 o f the
previous essay).
5.1 The Saptasatimahayantra
T h e durgasaptasathnahayantra, abbreviated saptasatim ahayantra, is a
yantra w hich originally w as exclusive to the Sakta tradition. It is
com paratively well known in M aharastra and is printed in the Rgv e d iy a b ra h m a k a rm a s a m u c c a y a , the m an u al co m m o n ly u sed by
priests in the Rg-Vedic tradition (Illustration 4 ).13 The yantra becam e
increasingly important as the popularity o f the D urgasaptasatl (also
called D evim ahatm ya or Candi) grew. This text eulogizes the deeds
o f D urga in approximately 700 stanzas. It dates back to ca. 5 0 0 -6 0 0
A.D . and is attributed to the M a rk an d ey a-P u ran a. T he y an tra is
em ployed in the ritual worship o f the goddess known as candlyaga,
especially for Durga worship during the N avaratri festival.
A lthough used by Sm artas, the yantra show s ty p ical T antric
elem en ts both in its structure and in the use o f the seed ( blja)
syllables inscribed in it. As noted in section 2.2.3 o f the preceding
essay, yantras e m p lo y ed in optional d e s ire -o rie n te d rites often
contain seed syllables. These are one-syllabled m antras frequently
derived from the first letter o f a deitys name to w hich a nasal sound,
the anusvara or anunasika, is added. A total o f 85 deities are invoked
13
F o r a sim p le r v ersion o f the y a n tra , cf. C o b u r n 1991: 115 ( F ig u r e 5 3) and
1 3 8 -1 3 9 .
64
GUDRUN BUI INIiMANN
into the y an tra .1'1 In the centre, a dow nw ard-pointing triangle (trikona) is found containing salutations to the three principal e m a n a
tions o f the goddess, namely, M ahakali (I), M ahalaksm i (2) and
M ahasarasvatl (3), along with the goddesses respective seed s y lla
bles, aim, hrim and klim. This triangle is located inside a hexagram
{satkona) in whose six corners the names o f two groups o f deities are
inscribed. Group 1: Sarasvati and Brahma (4), Gam i and Rudra (5),
L a k sm i and 11 rsikesa (6 ), Asladatfabhuja (7), D a s a n a n a (8 ),
A s ta b h u ja (9); g ro u p 2: N an d aja (10), R a k ta d a n tik a (1 1 ),
Sakambhari (12), Durga (13), Bhima (14) and Bhramari (15). G roup
1 consists o f Brahma, Siva and Visnu with their consorts and the
three great m anifestations o f the goddess in the D e v im ah atm y a,
namely, Astadasabhuja (the Eighteen-Armed O n e, that is, M a h a
laksm i), Da&anana ( the T en-Faced O n e , that is, M ahakali) and
Astabhuja (the Bight-Armed O n e, that is, Mahasarasvatl). Group 2
consists o f special m anifestations o f the goddess d e s c rib e d in
D ev im ah atm y a 11.42-55. Jaya (16), V ijaya (17), Jayanti (18),
Aparajita (19) and the two vehicles of the goddess, the lion (sim ha)
(20) and buffalo (mahisa) (21), are invoked into the points in w hich
the sides o f the two triangles forming the hexagram intersect. T he
hexagram is inside a lotus into whose eight petals again two groups
o f deities are invoked. The first group consists o f the m o th e r
goddesses (matrka): Brahmi (22), M ahesvari (23), K aum ari (24),
V a is n a v i (25), V arahi (26), N a ra sim h i (27), A in d rl (28) and
C am unda (2 9 ).15 The second group consists o f the eight Bhairavas:
A sitarigabhairava (30), R urubhairava (31), C and a b h airav a (32),
K rodhabhairava (33), Unm attabhairava (34), K apalabhairava (35),
Bhisanabhairava (36) and Samharabhairava (37), Outside the eightpetalled lotus, a lotus with 24 petals is found. In it the follow ing
saktis are invoked: V isnumaya (38), Cetana (39), Buddhi (40), Nidra
(41), Ksudha (42), Chaya (43), Sakti (44), Trsna (45), Ksiinti (46),
Jati (47), Lajja (48), Santi (49), Sraddha (50), Kanti (51), L aksm i
(52), Dhrti (53), Vrtti (54), Smrti (55), Daya (56), Tusti (57), Pusti
(58), M a tr (59), Bhranti (60) and Citi (61), Outside this lotus is a
square (bhupura, bhugrha) w ith four protruding gates. Into the
14 F or a text listing alm ost the sam e nam es as those inscribed in the yantra skctch,
see R g v e d ly a b ra h m a k arm a sa m u c c a y a , fols. 349a. 12350b.8.
15 T h e n a m e s differ slig h tly fro m those in the rudrupitlmnalvlyantra. In that
yantra, C a n d ik a replaces N arasim h i and the se q u e n c e o f the Iasi three n a m e s in the
list differs.
M AN DAL AS A ND Y AN TR AS IN S MART A R IT U AL I
65
corners o f the square and the four gates two groups o f deities are
invoked. The first group consists o f the ten directional guardians
( dikpala), who are invoked into their respective directions beginning
from the east:!( Indra (62), Agni (63), Yama (64), Nirrti (65), V anina
(6 6 ), V ayu (67), Som a (= Kubera) (68 ) and Rudra (Tsana) (69).
B ra h m a (70) is assigned to the zenith, which is located betw een the
north-east and east in the yantra. Sesa (71) is assigned to the nadir,
w hich is located between the south-west and west. The second group
consists o f the attributes (ayudha) o f the directional guardians, which
are assigned in the same sequence (cf. also the rudrayantra described
below); the thunderbolt ( vajra) (72; east), the spear (sakti) (73; south
east), the staff (danda) (74; south), the sword ( khadga) (75; south
w est), the noose (paa) (76; west), the goad (ankusa) (77; n o rth
w est), the mace (gada) (78; north), the trident ( trisula) (79; no rth
east), the lotus (padm a) (80; assigned to the zenith) and the wheel
(cakra) (81; assigned to the nadir). Outside the gates Ganapati (82),
Ksetrapala (83), Batuka (84) and the Y oginis (85) are found.
Even though the num ber o f deities invoked into this y an tra is
rath er large, its design is com m on: a dow nw ard -p o in tin g fem ale
trian g le inside a hexagram w hich is in the pericarp o f an eightpetalled lotus. The lotus is surrounded by another lotus and a square
with four gates. In addition to specific deities connected with Durga,
groups o f deities appear which are often found in other yantras. They
are the group o f eight m others, the Bhairavas, the ten directional
guardians and behind the latter their attributes.
5.2 The Rudraplthamahayantra
T he rudraplthamahayantra ( great yantra o f R u d ras seat), or simply
rudrapitha! rudrayantra (see Illustration 5), is em ployed in connection
w ith the r e c ita tio n o f the R u d ra d h y a y a o f T a ittiriy a -S a rn h ita
4 .5 .1 -1 1 o f the Yajur-Veda. The R gvediyabrahm akarm asam uccaya
ascribes this yantra to the Skanda-Purana17 and prescribes it for all
rituals w orshipping Rudra. A fter the deities are invoked into the
yantra, a vessel (kalaSa) is placed on it to serve as the seat for Siva,
the principal deity o f the ritual (see Colour Plate 4).
1(1T h e na m e s o f the directional g u ard ian s and their directions are listed in scction
4 in connection witli the navagraham andala.
17 See R g v e d iy a b ra h m a k a rm a s a m u c c a y a , fols. 3 15a. 1 - 3 18a.8 for th e description.
66
GUDRUN BUMNIMANN
The yantra consists o f a lotus pattern with live rings o f petals that
contain five m ajor groups (divided into subgroups) o f surrounding
deities. Outside the petals is a square (bhupura) with three nested
lines and four protruding gates. The lines o f the square are coloured,
from inside to out, white, red and black. They arc called sadva, rajas
and tamas and identified with the three constituents (g u n a ) o f
primary matter (prakrti) in the Samkhya system."* Outside the gates
are depicted the eight great snakes, In this yantra, a total o f 141
deities are invoked (the diagram numbers them only up to 121),
generally beginning in the west. The five-faced Rudra (1) is invoked
into the pericarp o f the lotus and w orshipped with the m an tra s
corresponding to his five aspects, Sadyojala, V am adeva, A ghora,
T a tp u ru sa and ISana. The first group o f surrounding deities is
assigned to the innermost ring o f eight petals: Nandin (2), M ahakala
(3), GaneSvara (4) Vrsabha (5), Bhrhgiriti1'' (6), Skanda (7), U m a (8)
and C andi^vara (9). In the adjacent ring o f 16 lotus petals the
following aspects o f Siva are invoked: Ananta (10), Suksm a (11),
Siva (12), Ekapat (13), Ekarudra (14), Trimurti (15), &rikantha (16),
V am adeva (17), Jyestha (18), Srestha (19), Rudra (20), Kala (21),
Kalavikarana (22), Balavikarana (23), Bala (24) and Balapramalhana
(25). The third group o f deities is positioned in the next ring o f 24
lotus petals. This group consists o f three subgroups: the supernatural
pow ers (siddhi), the m other goddesses (m atrka) and the Bhairavas.
The eig h t supernatural pow ers2*1 are: atom ization (a n im a n ) (26),
m agnification (m ahim an) (27), levitation (laghim an) (28) heaviness
(gariman) (29), extension (jprapti) (30), efficacy/non-obstruction o f
desire (prakam ya) (31), sovereignty (Ltita) (32) and mastery (vasita)
(33). The eight m other goddesses are Brahmi (34), M ahesvari (35),
Kaumari (36), Vaisnavl (37), Varahi (38), Aindri (39), C am unda (40)
and C andika ( 4 1).21 The eight Bhairavas (4 2 -4 9 ) arc identical with
the ones listed above for the sap{aatimahayantra, namely Asitahgabhairava and the others. To the fourth ring o f 32 lotus petals are
assigned several subgroups o f deities. The first subgroup consists o f
lsSec also m y rem arks in section 2.3.8 o f the p revious essay for these three lines.
I.e., the tw o attendants o f Siva.
aT h e s e supernatural p o w ers arc explained in V y a s a s c o m m e n ta ry on Y o g a su tra
3.45.
21
T h e n a m e s d i f f e r s o m e w h a t fro m th o s e in th e e a r l i e r list for th e
saptasatim ahayantra. In that yantra, N a ra sim h i appears in place o f C a n d ik a and the
sequence o f the last three n a m e s in the list differs.
m a n d a l a s a n d y a n t r a s in s m a r t a
RI TU A L I
67
the eig h t manifestations (m urli) o f Siva: B hava (50), Sarva (51),

I&ana (52), Pasupati (53), Rudra (54), U gra (55), Bhim a (56) and
M a h a t (57), The second subgroup consists o f the eight great serpents:
S esa (58), A nanta (59), V asuki (60), T a k sak a (61), K u lira (62),
K arkotaka (63), Sahkhapala (64) and Kambalasvatara (64). The third
subgroup consists o f the k in g s:22 V ainya (66), Prthu (67), Haihaya
(58), A rjuna (69), Sakuntaleya (70), Bharata (71), N ala (72) and
R a m a (73), The fourth subgroup consists o f the eig h t principal
m o u n tain ranges (kulacala): H im avat (74), N isadha (75), V indhya
(76), M alyavat (77), Pariyatraka (78), M alaya (79), H em akuta (80)
and G andham adana (81). The fifth ring has 40 lotus petals with the
fo llo w in g five subgroups o f divine beings: the eig h t directional
guardians, their consorts, the gu ard ian s attributes, the g u ard ian s
v e h ic le s and the d ire c tio n a l e lep h an ts. T h e eig h t d ire c tio n a l
guardians are as in the navagrahamandala and the saptasatimahayantra Indra (82), Agni (83), Yama (84), Nirrti (85), V aruna, ( 86),
V ayu (87), Kubera (88) and Isana (89). Their eight consorts are: Saci
(90), Svaha (91), Varahi (92), K hadgini (93), V aruni (94), V ayavi
(95), Kauberl (96) and ISani (97). The eight guardians attributes are
as in the saplasatlm ahayantra: the thunderbolt (vajra) (98), the spear
( sa kti) (99), the s ta ff (danda) (100), the sword (.khadga) (101), the
noose (pasa) (102), the goad (ahkusa) (103), the m ace (gada) (104)
and the trident ( trisula) (105). The guardians vehicles ( vahana) are
I n d ra s elephant Airavata (106), the ram (m esa ) (107), the buffalo
(m ahisa) (108), the corpse (preta) (109), the sea-m onster (makara)
( 110),- the deer (harina) ( 111), the man (nara) ( 112) and the bull
( vrsabha) (113). The directional elephants (diggaja) are specified as
22
T h e s e n a m e s a p p e a r to h a v e b e e n tak en fro m the f o llo w in g v e r s e w h i c h is
tr a d itio n a lly re cited by M a h a ra s tria n B r a h m in s up o n risin g in the m o rn in g . T h is
v erse, w h ich is believed to s ecu re w ealth and victory, is prin ted in texts such as the
B h a k t i m a r g a d l p (edited by G.N, D a n d e k a r , M u m b a i: M a je s tic B o o k Stall, 1978
[third edition]): 2:
vainyam prthum haihayam arjuimin ca

sakuntalcyam bharatam nalaip ca /
ramaip ca y o vni smarati prabhatc
tasyarthaklbho vijn ya i ca h a ste//
A c tu a lly , the verse lists only the na m e s o f five kings: Prthu, th e son o f K in g V e n a
(also spelled V ena); A rju n a (K artav iry a) o f the H a ih a y a fam ily; B harata, the so n o f
S ak u n ta la ; N a la and R am a. In the yantra, the n a m e s are clearly in te rp re te d as eig h t
s e p a ra te nam es. T h e individual w h o first assig n ed the n a m e s to the y a n tra did n o t
realize that three o f the epithets (vainya . haihaya and sakuntaleya) are a d jectiv es that
m o d ify three o f the n a m e s (Prthu, A rjuna and Bharata).
68
GUDRUN BUHNHMANN
Airavata (114), Pundarika (115), Vamana (116), K um uda (1 1 7 ),

Anjana (1 18), Puspadanla (I 19), Saxvabhauma (120) and S u p ratlk a
( 121). In the gales o f the square outside the lotuses, the e ig h t
directional guardians (1 2 2 -1 2 9 ) appear lor the second time in th e
yantra. Four additional deities, who represent forms o f Siva, are
invoked into the intermediate directions: Virupaksa (130) ( s o u th
east), Visvarupa (131) (south-west), Pasupati (132) (north-west) and
Ordhvalihga (133) (north-east). The eight great serpents outside the
square arc: Sesa (134), Taksaka (135), Ananta (136), Vasuki (137),
Sarikhapala (138), Mahapadma (139), Kambala (140) and K arkotaka
(141). They appear here for the second time with variants for som e
o f their names.
This yantra features a lotus design with five rings o f petals,
surrounded by a square with three nested lines. A m ong the deities
invoked are special groups associated with Rudra/Siva. In addition,
com m on groups o f deities who are also assigned to other y an tras
appear. T hese are the m other goddesses, the B hairavas, the eig h t
great serpents, the directional guardians, their consorts and th e ir
attributes. The groups o f directional guardians and great serpents
appear twice in the yantra, each time in a different part o f the yantra,
which is not unusual.
The second part o f this paper introduces a structure quite different
from the lotus designs described here. It deals w ith types o f
bhadraniandalas, all o f which share the square grid.
16
17
^ r^ r .
29
\
32
YANTRAS
fg ^ttrr ^ r jf :
9. I gRrr
26
^m s^Tnf m m
19
18
SMARTA
Ot
.Uf
4]/
***r f^ rr^ riT ^ !

27
31
>P*
30
RITUAL
v^,
O'
IN
-3SL
W fW ^rg : ffmPT
22
23
*sror
24
AND
&>
25
MANDALAS
f^TTsar:
28
ijL f t* * *
Os
V.O
fôrnr 20
srw p ^ T t 21
I, A diagram showing the deities of the baliharanacakra
GUDRUM BU UN EM ANN
70
NE
EAST
NORTH
NW
SE
SOUTH
WEST
SOr y a p a Sc J
SW
SlVAPA8cXYATANA
V j^ n u p a S cX y a t a n a
Vignu (2)
Sflrya (3)
Siva (1)
D ov T (5)
Gancfta (4)
Siva (2)
Ganesa (3)
Vi^nu (1)
Devi (5)
Sflrya (4)
yatana
Siva (2)
Ganeaa (3)
Sflrya (1)
Devi (5)
V i^ u (4)
e v Tp a R o A y a t a n a
Vigpu (2)
Siva (3)
DevT (1)
SOrya (5)
Gancda (4)
a ije Sa p a S c X y a t a n a
Vigiju (2)
Siva (3)
Ganeaa (1)
Devi (5)
SQrya (4)
2. The arrangem ent o f the five deities in (domestic) p a n c a y a tu n a shrines
il
II
3. A mandala of the heavenly bodies
M ANDALAS A ND Y AN T RA S IN S M A R T A R I T U A L I
4, The (durga)saptasatimahayantra
71
72
GUDRUN B UHNEMANN
5. T h e ru d ra p ith a m a h u ya n tra
PART II. B H A D R A M A N D A L A S 1
1 Introduction
Bhadram andalas are square-shaped mandalas that are divided into a
grid o f squares.2 Specific shapes are traced within the fram ew ork o f
these squares. In the course o f the ritual, deities are invoked into
different parts o f these shapes, The m andalas are used m ainly as
supports (pitha) for vessels ( kalasa). The vessels function as seats for
icons o f deities. The bhadram andalas are em ployed m ainly in the
concluding ceremonies o f religious observances (vratodyapana). It is
believed that a ritual perform ed without the support o f a bhadra is
fruitless, w hereas a ritual em ploying such a support is tho u g h t to
y ie ld excellent results (B hadram artanda, fol. 2 a .5 -6 ). The c o n
stru ctio n o f the b h a d ra m a n d a la s is still alive in M ah arastra and
d escrib ed in ritual handbooks o f priests. In the follow ing I will
explore the different types and structures o f these m andalas. M y
m ain textual source here is the 19th-century Bhadram artanda written
by Harikrsna, a text which has not been analyzed previously. M o s t3
o f the m andalas found in the printed edition o f the Bhadram artanda
have been reproduced in this book.
1 In m y pap er entitled B h a d ra m a n d a la s in the Ritual P r a c tic e , p u b lish e d in the

W i e n e r Z c its c h rift fiir die K u n d c Siidasicns in 1987, 1 d isc u ss sarva to b h a d ra s and
s e v e r a l t y p e s o f lingatobhadras, the d e itie s in v o k e d into th e m a n d th e ritu a l
e m p l o y m e n t o f the bhadram andalas. T h e paper includes d ia g ra m s o f m a n d a l a s and
p h o t o g r a p h s o f sta g e s in th e ir c o n s tr u c tio n , to g e t h e r w ith tr a n s l a t i o n s o f text
p a s s a g e s and n u m e r o u s rc fc rc n c c s in th e n o tes. T h is in f o r m a tio n h a s n o t b een
d u p lic a te d here. W h ile the earlier pap er refers only briefly to so m e m o d ific a tio n s o f
th e se m andalas, such as the ram atobhadras, these m o d ific a tio n s are d e sc rib e d h ere in
g re a te r detail. T h u s this essay c o m p le m e n ts m y earlier paper.
2 T h e sq u are grid is described in section 1.3.2 o f the p rev io u s essay.
3 T h e p rin te d ed itio n d o c s not illu stra te all m a n d a l a s d e s c r ib e d in th e text.
I llu s tra tio n s o f m a n d a la s 36, 38, 5 3 - 6 5 are m issin g . M o r e o v e r , s o m e m a n d a l a
d r a w in g s arc in c o m p le te ( 2 4 - 2 5 , 3 9 - 4 0 and 5 0 - 5 1 ) an d h a v e th e re fo re n o t been
rc p ro d u c c d in this book.
74
G U D R U N BUI I N K M A N N
2 The Bhadramartanda and its A uthor

The Bhadramartanda (BM) is, to my knowledge, the most c o m p r e
hensive com pendium on hhadram andalas. The text describes as
many as 76 bhndras, which are to be employed by the followers o f
the V ed a ( vaidika).
The BM forms chapter (adhyaya) 17 o f the sixth skandha o f the
Brhajjyotisarnava, written by the astrologer (jyo tirvid ) H arikrsna in
1871 A.D. Harikrsnas father was Vehkatarama, an m id Icy a B rahm in
o f Gujarat o f the audicya-sahasra branch who resided in A urangabad.
Venkatarama belonged to the vajasancya-sakha o f the W hite YajurVeda and the gautania-gotra.
To date, only parts o f the comprehensive Brhajjyotisarnava have
been printed by the VehkateSvar Press (also known as Khemraj Srik rsn ad as). The printed parts o f the text include the B a tu k a bhairavopasana (Bom bay 1909), the 124th adhyaya in the dharm askandha o f the work; the CakravalLsamgraha (Bom bay 1900), in clu
ded in the sixth skandha; the KridakauSalya (Bombay 1901), the 20th
adhyaya in the sixth skandha, the Hanumadupasana (B om bay 1899),
the 1 14th adhyaya of the upasanastabaka in the eighth skandha; and
the D urgopasanakalpadrum a (Bom bay 1907), the 128th a d h ya ya in
the eighth skandha. H arikrsna is also the author o f the K arttikamasamahatmyatika and the VaiSakhamasamahatmyatika.
A lthough H arikrsnas compilation is recent, descriptions o f m ost
o f the m andalas included in the BM can be traced back to o ld er
sources, such as the vratakhanda o f H em adris Caturvargacintam ani
(ca. 1 2 6 0-1270 A.D.), the manoharakiinda o f the A nandaram ayana
(AR) and to several other texts on vra/as and dharma.4 A m ong these
texts, the A R 5 is the most important source for the BM. Bulcke 1962:
73 assigns this text to ca. 1500 A.D. A ccording to R aghavan ( t )
1998: 121, the AR was most likely produced during the Maratha rule
4 H arik rsn a nam es the follow ing texts as his sources: the < A n a n d a - > R a m a y a n a ;
the V rataraja; the Santisara; the T altv asag ara-S am h ita; 'th e P a n c a ra ira; H c m a d r i < s
vratakhanda in his C aturvargacinlam ani> ; the P u ra n a s , cspccially the S k a n d a and
the L a iiig a ; the M an d a la sa m g ra h a; the T a n t r a s , cspccially the R u d ra y a m a la ; the
R u d r a p a d d h a ti, a text that m a y be identical with the M a h a r u d r a p a d d h a ti, o r th e
R u d ra n u s th a n a p a d d h a ti (A u frcch t !: 530), w hich w as au th o red by N a ra y a n a b h a tta ,
son o f R a m c v a r a b h a tta ; an d the K a u m u d i , w h ic h m a y be id en tical w ith th e
V ra to d y a p a n a k au m u d l.
5 F o r a s u m m a r y o f the narrative co n ten ts o f the A R, see R a g h a v a n ( f ) 1998:
7 2 -1 2 4 .
MAN D AL A S A ND Y AN TRA S IN S MARTA R I T U A L II
75
in South India, between the end o f the 17th and beginning o f the 18th
centuries. This later date is more likely. Several o f the m andalas
H arikrsna describes are detailed in contemporary ritual handbooks o f
priests in Maharastra, and are currently employed in rituals.
3 The Structure o f the Bhadramartanda and its M andalas
T h e B M has been out o f print for a long time and it is difficult to
obtain copies. The work, comprising 36 printed folios, is written in
verse with occasional b rie f prose explanations. M ost verses consist
o f quotations from other texts. The w ork begins w ith a general
section that provides definitions and explanations (paribhasa), and
then proceeds with technical instructions for drawing the mandalas.
T h e 76 bhadras described in the text are divided into the following
m ain categories:
1)
2)
3)
4)
5)
6)
7)
8)
sarvatobhadras
lihgatobhadras for Siva, including bhadras for GaurT
bhadras for Rama
the pancabhadra
bhadras for Surya
the bhadra for Ganapati/Ganesa
the cakraravindamandala
the svastikamandala.
A com plete list o f m andalas described in the B M is found in the

appendix at the end o f this paper.
Sarvatobhadra means auspicious from all sid es6 and m ost likely
refers to the sym m etry o f the m andala d esig n .7 The nam e has been
D a k s h in a r a n ja n Shastri 1940: 171 tra n sla te s the n a m e as b e a u tifu l in e v e ry

a s p e c t , B r u n n e r , p. 167 a s th o r o u g h l y a u s p i c i o u s a n d R a s t e l l i , p. 124 as
a u s p ic io u s in every w a y .
7 S arv a to b h a d ra is also the n a m e o f a bandha c o m p o s itio n in literature, in w h ic h
v e rs e q u a rte rs can be read in all four directions, fo rw a rd s, b a c k w a rd s , h o riz o n ta lly
an d vertically. A s L icnhard d em o n strates, na m e s o f bandhas can be tra c e d b a c k to
c e rta in a rm y fo rm a tio n s (v y u h a ), A s a r v a to b h a d ra v y u h a can attack an e n e m y from
all sid e s (L ie n h a rd 1997: 346, 350, 351). F o r a d ra w in g o f a sarvatobhadra stanza,
s e c L ic n h a rd 1997: 353. T h e te rm sa rvatobhadra is used in arch ite c tu re to d e n o te a
h o u s e w ith a v e ra n d a all a round (B rhat-Sarnhita 53.31) an d a ty p e o f te m p le (B rhatS a rn h ita 56.18, 56. 27). In to w n - p la n n in g th e term re fe rs to an o b lo n g or s q u a r e
s h a p e d town. F or a sarvatobhadra to w n plan, sec L ien h ard 1997: 358.
76
GUDRUN BOlINHMANN
used by different texts to designate distinct types o f mandalas. For

exam ple, the variations o f this m andala in the BM d iffer fro m
descriptions in some other texts, for example, the s a rv a to b h a d ra
according to the third chapter o f the Saradatilaka, w hich B ru n n e r
reconstructs in her contribution. The sarvatobhadra, o f w hich four
types are explained in the BM (Illustrations 1-4), is the most v e rsa
tile o f all the bhadras in application. It can be em ployed in all the
religious observances (vra/a), regardless o f w hether the principal
deity is Visnu, Siva or another deity. But it is believed to be
particularly well suited for Vaisnava rites. One type o f sarvatobhadra
described in the BM (Illustration 3) is an eight-petalled (a stadaia)
lotus, which functions as a substitute for a bhadramandala.
The terms lihgatobhadra, ramatobhadra and so forth are derived
analogically from the term sarvatobhadra. The term Iihgato<bhadra>,
<auspicious> because o f a l ih g a /lih g a s is used synonymously with
lihgasamudbhava, <auspicious(ness)> arisen from a lihga/lihgas.'
The lihgatobhadras described in the main section o f the BM fall
into many subcategories. The subcategories are named according to
the num ber o f phallic symbols o f Siva ( lihga) they depict, ranging
from one to 1,008. In addition to their number, the arrangem ent o f
the lingas within the mandalas is significant, as is the num ber o f lines
drawn while constructing the mandalas. As one w ould expect, the
lihgatobhadras are em ployed in vratas connected with deities o f the
Saiva tradition. T hree lihgatobhadras that are nam ed gauritilaka 9
(Illustrations 5 -7 ) are prescribed for vratas connected w ith S iv a 's
*
F o r o th e r types o f sa rvatobhadra, sec the references in B iih n em an n 1987: 4 3 ,
note 2 and, in addition, the c o lo u r d ia g ra m in B anerji 1978: 17(H- a n d the line
d ra w in g in V.V, D w ivcd/J, P an d ey /S .S , B aluilkar (B haratiya Tantra&istra, S arnath,
Varanasi: Central Institute o f Higher T ibetan Studies, 1995): 682. A sarvatobhadra is
included in the sk etc h b o o k o f the painter V ish n u B a h a d u r C hilrak n r o f B h a k ta p u r,
N epal, w hich is rcp ro d u ccd in Vcrgati 1982: 56. G acffk c 1987: 154 states that the
sa rv a to b h a d ra is identical with the n a v a p a d m a m a n d a la ; these m a n d a la s, h o w e v e r,
usually differ con sid erab ly in structure. For a sketch o f a n a va p a d m a m a n d a la r e c o n
structed acco rd in g to the description in the J a y a k h y a -S a m h ila o f the Pan caratra, se c
A pte 1973: 514+, illustrated Figure 1. His skctelt is based on a colour print p re fix e d
lo the fo r e w o r d o f o n ly the first edition o f the J a y a k h y a - S a m h ita (1931); sec also
C o lo u r Plate 15. T h e n a va p a d m a m a n d a la is also described in the L a k s m l-T a n tra . A
c o lo u r diagram is a p p e n d e d to G u p t a s translation o f the L ak sm l-T an tra (= D ia g r a m
III), w h ic h is r c p ro d u c c d in black and w hite in K hanna 1979: 92. T h e T a n t r a s a d b h av a (see T o rz s o k , p. 205, note 95) app aren tly refers to a type o f srim a n d a la as
sarvatobhadra(ka),
9 F o r d ra w in g s o f tw o d ifferent g a u ritila k a m a n d a la s each w ith four lin g a s an d
five lotuses, see SP3, Plate XV and also p. 696; see also S ab arath in am 1995: 62.
M A N D AL A S A N D Y AN TR A S IN S MAR TA RI TU A L II
77
consort Gauri, also referred to as Sakti. Combinations o f the sarvatobhadra and the lingatobhadra are frequently term ed harihara(atmakajlihgatobhadras, lingatobhadras (consisting) o f Hari and H a r a
(for example, Illustrations 19-21). The sarvatobhadra, which usually
appears in the centre of these mandalas, represents Hari/Visnu, while
the lingatobhadra on the outside represents Hara/Siva. T he hariharaUngatobhadras can be employed in both Vaisnava and in Saiva rites.
L in g a to b h a d ra s are also com bined with the au sp ic io u s svastika
sym bol. Five additional lingatobhadras, w hich are used on special
occasions, such as festivals, and in rites performed for the attainment
o f specific goals, are described separately in the final section o f the
B M . The total num ber of lingatobhadras described in the text is 41.
T h e 25 bhadras for R am a, called ram atobhadras, auspicious
because o f <the seal o f > R am a, are em ployed in rituals for R am a
and Visnu. They are mandalas inscribed with one or several seals
(m udra). The seal is defined as the words raja rama ( King R a m a )
(see Table, Figure 9), or, in reverse order, rama raja. The bhadras for
R a m a may also be inscribed with the nam e (naman) rama rama. In
such cases, they are classified as ramanamatobhadras (cf. AR, m a n o harakanda 4 .4 9 -4 5 ). Alternatively, they can be inscribed w ith the
n a m e s rama ram a, in w hich rama refers to R a m a s wife. Som e
bhadras are inscribed with the seal raja rama together w ith lihgas.
T h e s e bhadras are term ed ramalingatobhadras (see C o lo u r Plates
6 - 7 ) .10 Various bhadras for Rama, such as ramatobhadras and ramaJihgatobhadras, m any o f w hich contain a sarvatobhadra in their
centre, are elaborately described in sargas 4 and 5 o f the m anoharakanda o f the AR. C om bining a ramamudra and a sivalihga in the
sam e m andala poses no doctrinal problems. R am a is said to have
w orshipped Siva and established lihgas (cf. AR, sarakanda 10.124),
w hich are often called ram alihgas in M ah arastra and K arnataka.
Siva, on the o ther hand, is said to u tter a ram am antra called
tarakamantra for the benefit o f those w ho pass away in V aranasi.
AR, rajyakanda 18.19 refers to a stone inscribed w ith the ramamudra
in Ram anathapura, a site close to Ramesvara. A ccording to legends,
the stones used to build the bridge to Lanka w ere inscribed with the
O f these tw o paintings, w h ic h c o m e from p riv ate collections, the p a tte rn o f the

first bhadra c a n n o t be id entified in the B M . T h e s eco n d hhadra b e a rs th e inscription
c a tu r m u d n lr a m a lin g a to b h a d r a and c o rre s p o n d s to Illustration 4 2 fro m th e B M . It
features four nlm aniudriis, eight lingas and a sarvatobhadra in the centre.
78
GUD RU N BUMNHMANN
ramamantra or with R a m a s name. In the AR the p o w e r o f th e

ramamudra is eulogized repeatedly. When a mandala that c o m b in es
ramamudras and sivalihgas is employed for the worship o f Rama, the
ramamudra becomes the object o f worship and the lingii/Siva acts as
the w o rsh ip p er." But when the same m andala is used in a ritual
w orshipping Siva, the Hnga becomes the object o f worship, and the
ramamudra, the worshipper. A ramatobhadra is employed in rites for
Visnu, while a ramalingatoblmdra is suitable for either Visnu or Siva.
A bhadra inscribed with the s e a l rama rama is used for th e
goddess.'2
Two types o f suryabhadras (Illustrations 6 7-68) are described for
use in vratas connected with the sun. Bach one contains 12 images o f
the sun, 12 being the num ber that sym bolizes the sun, since it
corresponds to the 12 months o f the year and the 12 zodiac signs.
The two bhadras differ in the number o f lines drawn and in the w ay
the icons are placed. The bhadra (Colour Plate 10) from Rajasthan
can be identified as the suryabhadra with 20 lines corresponding to
Illustration 67 from the BM.
The BM describes only one bhadra for G anapati/G anesa (see
Illustration 69). It is a bhadra with a total o f five icons o f G anapati,
four in the cardinal directions and one in the centre (see C olour Plate
8 ).13 B ut w e know from other texts that several other types o f
ganapatibhadras exist.1'1 One o f these features 21 icons o f Ganapati,
21 being an important num ber in the worship o f the deity (see C olour
Plate 9).
11 Cf. AR, m anoharakanda, 4.6 and 5 .9 0 c d -9 1 a b ,

12 Cf. AR, m anoharakanda, 5.375; B M , fol. 3a.8 -9 ,
13 T h e m a n d ala is rc p ro d u c c d from GancSkoS (edited by A. G adgil, Pune: r tr a m
B o o k A g c n c y , 1981): 477. T h e ganesabhadra r c p ro d u c c d in the BM fe a tu re s f o u r
icons o f GaneSa in the cardinal directio n s but no icon (only a lotus) in the centre.
T h is ganesabhadra is rcproduccd in colour in S.D. Dc&ikar (G anapati, T irv a v a d u tu ra i
Aditiam, 1984 [second edition] [in Tamil)): 141+,
14 F or ganapatibhadras, see also Biihnemann 1987: 4 8 - 4 9 , 59, 63.
MAN DAL AS AND Y AN T RA S IN S MARTA R I T U A L II
79
O th er m andalas in the BM include the cakraravinda ,15 a lotus

encircled by a wheel, and the svastikam andala ,16 w hich features the
auspicious svastika symbol inside a square. The svastika also appears
in other mandalas, such as some lingatobhadras in com bination with
lingas, lotuses or other symbols. It is an ancient auspicious symbol
w h ich A uer/G utschow 1974: 22, 38 consider an abstract form o f a
m a n d a la whose a r m s establish the unity o f the m a n d a la s four
parts.
4 C onstituent Parts o f the Bhadras
B hadras are constructed by drawing a certain num ber o f vertical and
horizontal base lines ( reklia) that form square grids. The squares are
called pada or kostha (see Table, Figure 1). The m andala is m ade up
o f padas that are assem bled to form different shapes. T h e sarvato
bhadra usually has the following parts (see Table, Figures 2-9):
enclosure (paridhi) (Figure 2)
w e ll ( vapi) (Figure 3)
15 A m o re c o m p le x m a n d a la o f this ty p e is th e cakrabjam andala. T h e P a u sk a ra S arnhita, w hich c o n tain s a d esc rip tio n o f the m a n d a la in c h a p te r 8, stales in v erse
2 a b th at m ultiple variants o f this m a n d a la exist. F o r a d e s c rip tio n o f the c a k r a b ja
m a n d a la , see P a d m a - S a m h ita , c a ry a p a d a , c h a p te r 2. T h e m a n d a la is d is c u s s e d in
v o l u m e 1, A ppendix 10: 5 3 - 5 4 o f the edited text (w ith r e fe re n ce s to parallel texts)
a n d in v o lu m e 2: 5 6 9 - 5 7 1 ; sec also C o lo u r Plate 14 in this b o o k (re p ro d u c e d fro m a
p la te inserted in the b a c k o f v o lu m e 1) and the sketch in v o lu m e 2: 568. A n o th e r
v a ria n t o f the m a n d a la is r e p ro d u c e d in the ap p e n d ix to G u p t a s tra n sla tio n o f the
L a k s m l - T a n t r a (= D ia g r a m IV); in the m a n d a la r e p ro d u c tio n the tips o f the lotus
p e ta ls are not oriented to the points o f the co m p ass. G u p t a s d iag ram is re p rin te d in
b l a c k an d w h ite in K h a n n a 1979: 93. F o r a s k e tc h o f a s o m e w h a t d if f e r e n t
cakrabjam andala, see R a o 1 9 8 8 - 1 9 9 2 , v o lu m e 5: 27. T h e cakrabjam andala is also
k n o w n as b h a d ra ka m a n d a la (see Rastelli, p. 124, note 28 ) b e c a u s e it is s im ila r to a
t y p e o f sa r v a to b h a d r a , n a m e ly the sa rv a to b h a d ra r e c o n s tr u c te d b y B r u n n e r . O n e
m a r k o f d istinction b e tw e e n the c a k ra b ja m a n d a la and th e s a r v a to b h a d r a is that a
c o n c h shell is draw n in each o f the fo u r c o m e r s o f the c a k r a b ja m a n d a la . F o r a
b h a d ra k a m a n d a la , see the line d ra w in g printed in N a r a d iy a - S a m h ita , p. 576, T h e
s a m e line d raw in g also ap p ears in an article by R.P. C h a u d h a r y (in V .V . D w iv cd /J.
P an d c y /S .S . B ahulkar: B haratiya TantraSiistra, Sarnath, V aranasi: C entral Institute o f
H i g h e r T ib etan Studies, 1995): 682. F o r a sketch o f a sim p le r bhadrakam andala, see
R a o 1 9 8 8 -1 9 9 2 , volum e 5: 20.
'
16 F o r a different sv a stik a m a n d a la re c o n stru c te d fro m th e S id d h a n ta sa ra v a ll, see

S P 3, Plate XVI and p. 696; see further P a u sS 2 , a ppendix, p. XI and R a o 1 9 8 8 - 1 9 9 2 ,
v o lu m e 5: 25 for different types o f this m andala.
80
GUDRUN BUHNEMANN
offset (bhadra)]1 (Figure 4)

creeper (vnHi) (Figure 5)
chain (srhkhala) (Figure 6)
crescent m o o n ( khnndendu) (Figure 7).
In the centre is usually a lotus with a pericarp {karnika).
In addition to these parts, the liiigatoblmdras contain one or
several phallic symbols (lihga) o f Siva (see Figure 8) . u T h e lingas
are also referred to as Rudra or Siva in 1he BM. The number o f lingas
ranges from 1 to 1,008, O ther constituent parts found in so m e
lihgatobhadras are a passage (vith i) for circum am bulation, w hich
surrounds a throne (pitha), miniature creepers (laghuval.lt) and m in ia
ture chains ( laghusm khahi).19
As noted before, a ramatobhadra may contain lingas a n d /o r a
sarvatobhadra, but it m ainly depicts the s e a l o f R am a (T able,
Figure 9), usually the words raja rama. The gancsib and suryabhadras feature images o f Ganesa and the sun.
The parts o f the mandala arc usually one o f five c o lo u rs:20 the
enclosure is yellow; the well, white; the bhadra, red; the creeper,
dark green/blue; the chain, black; and the crescent moon, white. O n
the outside, the mandala is enclosed by a square with three nested
lines which are named after the three constituents (guna) o f primary
m atter (prakrti) in the Sam kbya system: sattva, rajas and tamas.
These lines are coloured white, red and black in accordance with the
sym bolism o f the gunas. The lotus in the centre o f the m andalas
often has a yellow pericarp and while petals, while the lingas are
black and the ramamudra is white.
T he BM recom m ends that one construct these m andalas from
powders or grains. The grains include unbroken rice grains (aksata)
(white), red lentils (red), chick-peas (yellow), black beans (dark) and
m oong gram (mudga) (Phaseolus mungo) (green) (BM, fol. 3a. 1-2).
The use o f both coloured powders and grains in the construction o f
17
In architectural term in o lo g y , bhadra designates an offset p rojection c o m m o n to
North Indian tem ple plans.
IS F o r additional c o n stitu e n t parts o f th e lin g a to b h a c in s , sec B u h n e m a n n 1987:
55.
|,; T h e se parts are described and illustrated in B iihnem ann 1987: 5 4 -5 5 .
20
T h e s ig n ific a n c e o f the five c olours in an cicn t Indian c u ltu re is e l a b o r a te ly
discussed in G oudriaan 1978: 190ff.
MA ND AL A S A ND Y AN TRA S IN S MART A RI TU AL II
81
the bhadras can be observed in contemporary ritual practice in India

(see Colour Plate 11).
5 The R itual E m p lo ym en t o f the M andalas
In M aharastra, the bhadras are mainly em ployed in concluding rites
( udyapana) o f religious observances ( vrata), which frequently consist
o f a piija followed by the giving o f a gift. This tradition has con
tinued until the present. The size o f the mandalas can vary. In M ah a
rastra, the bhadras are frequently prepared on a low square table used
in worship ( caurahga) and are comparatively small in size. After the
m a n d a la is constructed, the deities are invoked into areca nuts
(puglphala) that have been placed in parts o f the m andala (see Colour
Plate 12) and worshipped with such offerings as flowers. According
to BM fol. 3b. 1, the deities invoked into the ram atobhadras and the
lingatobhadras are said to be the same as those in the sarvatobhadra,
b u t other texts appear to differ.21 The group o f deities invoked into
the sarvatobhadra is a specific group led by B rahm a (brahm adim andaladevata)22 After the offerings, a vessel (kalasa) filled with
w a te r and auspicious objects is placed in the centre o f the mandala;
additional vessels may be placed in the four comers. A flat dish filled
w ith u n b ro k en rice grains is placed on top o f the vessel, w hich
b eco m es the throne o f the principal deity o f the vrata (see C olour
Plate 13).
In addition to being used in concluding rites o f vratas, a bhadra
can be e m p lo y ed as a seat for a deity in a p u j a ritual. The
sarvatobhadra can be used in the rite o f infusing life (pranapratistha)
into a statue, in rites for the heavenly bodies (such as the graham akha) and in rites o f pacification (santi).23 Vergati 1982: 57 reports
that in Nepal the sarvatobhadra is covered with a piece o f cloth on
which a vessel (kalasa) is placed. The m andala is used in the sevenday ritual recitation o f the B hagavata-Purana (bhagavata-saptaha)
and in the nine-day ritual recitation (navaha) o f the R am ayana and
other texts for Rama. As noted above, the sarvatobhadra is em ployed
for rites centring on Visnu, but it can also be used for any other
21
F o r the d e itie s in v o k e d in to sev eral lingatobhadras, sec B i ih n e m a n n 1987'
5 8 -5 9 ,6 5 -7 0 .
'
T h e ir nam es arc listed in B iihnem ann 1987: 6 3 -6 5 .
a F or detailed rcfcrcnccs, see B iihnem ann 1987: 4 9 - 5 0 and 61.
82
GUDRUN BUIINl'MANN
deity. The lihgaiobhadnE arc em ployed in Saiva rituals, and a

subcategory o f them, the gaurltilnkas, in rituals for S iv a s co n so rt,
Gauri.
T he tradition o f constructing b h a d ra m a n d a la s is still alive in
Maharastra and other parts o f South Asia, including Nepal, although
it is disappearing fast. The tim e-consuming construction, from the
drawing o f the lines with the help o f threads to the filling o f the
squares with pow ders or grains o f different colours, is now often
obviated by the use o f com m ercially produced charts, w hich can
easily be called back into service.2'1 Such coloured charts o f one o r
more types o f sarvatobhadra, lingatobhadra, grahamandala, ksetrapalamandala and matrkamandaia arc reportedly sold in shops in
V aranasi for use by priests. The printing and distribution o f such
charts prom otes the standardization o f certain variations o f th ese
mandalas, just as the printing o f a certain recension o f a text does.
6 Concluding R em arks
It appears that the Smartas included the b hadram andalas in their
ritual practices under the influence o f mandala riluals perform ed by
Tantric practitioners and mandala patterns described in Tantric texts.
The Pancaratra and the Tantras are referred to among the sources for
the BM . O ne such source may have been the U anaSivagurudevapaddhati by ISanaSivagurudevamisra, an 11th- or 12th-century Saiva
manual o f temple worship in four sections (piida). This text describes
m a n d a la s n a m e d bhadraka, sarvatobhadra, piirva ifka n ta ( a s t a lihgalatapadmavithiprakara), latalihgodbhava, pancabrahm aniandalagauritilaka, svastikabjadvaya, svastikasarvatobhadra, cakrabja,25
m ayacakra and tripuram andala.11' Som e o f these nam es are also
included in the BM. One can observe structural similarities betw een
some bhadramandalas and the mandalas reconstructed by Apte in his
24 B ru n n e r m akes a sim ilar observation regarding the m a n d ala tradition in S o u th

India (B runner, pp. 1 6 5-166, note 28).
25 T h e se m a n d a la s arc described tw icc in the ISanaSivagurudcvapaddhati, in tw o
very sim ilar pa ssa g e s. The first d escription ap p ears in s;liniiny;ip;l(l;i 6 . 3 6 - 1 5 2 (=
v o lu m e 1, pp. 51, 10 - 62, 2), and the sccond one in kriyfipficla 8 .3 1 -1 2 3 (= v o lu m e
3, pp. 77, 8 - 85, 6). T h e A jita(-A gam a) and the T attv asag ara(-S am liita) are a m o n g
ISanaSivagurudcvamiSras sourccs.
T h e last tw o m a n d a la s arc o n ly d e s c r ib e d in sfu n iin ya p iid n 6 . 1 3 2 - 1 5 2 (=
v o lu m e 1, pp. 60, 8 - 62, 2).
M AN DAL AS A ND Y AN T RA S IN S MARTA R ITU AL II
83
edition o f the Pauskara-Sam hita (Part 1). A detailed com parative

study o f m andalas described in the A g am a texts has not been
undertaken, but would be a prerequisite for tracing the developm ent
o f the bhadramandalas.
7 A ppendix: List o f M andalas D escribed in the B M
B e c a u se the BM appears to be the m ost co m p reh en siv e text on
bhadras, a complete list o f mandalas described in it is provided here,
0 General explanations (paribhasa) (fols. 1a. 13b.3)
1 Sarvatobhadras (fols. 3b.3-1 Oa.9)
1.1 Type 1 (fols. 3b.3-4b,6) (Illustration 1)
1.2 Type 2 and list o f deities invoked into the m andala w ith their
respective mantras (fols. 4b.6-8b.12) (Illustration 2)
1.3 Type 3: The astadalam andala as a substitute for the sarvato
bhadra and the deities invoked into it (fols. 8b. 1210a.2) (Illustra
tion 3)
1.4 Type 4 (fol. 10a.2 - 9 ) (Illustration 4)
2 Lihgatobhadras (fols. 10a.9-24b.l 1)
2.0 Deities invoked into the lihgatobhadras and their m antras (fol.
10a. 10-1 Ob. 11)
2.1 Gauritilakamandalas employed in vratas connected to goddesses
(fols. 10 b. 11-11 b.4)
2.1.1 Ekalingatobhadra iaghugauritilaka (fols. 10b. 11-1 l a . 3) (Illus
tration 5)
2.1.2 Caturlingatobhadra brhadgauritiiaka (fol. 11 a.3 -9 ) (Illustration
6)
2 . 1.3 Gaurltilaka (fol. 1 l a . 9-1 lb .4) (Illustration 7)
2.2 Caturlirigatobhadras (fols. 1 lb .412a.7)
2.2.1 Type 1 (fol, 1 lb .4 -1 0 ) (Illustration 8)
2.2.2 Type 2 (fol. 1 lb. 10-14) (Illustration 9)
2.2.3 Type 3 (fols. 11 b. 1412a.4) (Illustration 10)
2.2.4 Type 4 (fol. 12a.5 -7 ) (Illustration 11)
2.3 Astalirigatobhadras (fols. 12a.8-13a,5)
2.3.1 Type 1 (fol. 12a.8-12b,4) (Illustration 12)
2.3.2 Type 2 (fols. 12b.4-13a,2) (Illustration 13)
2.3.3 Tw o minor variations (fol. 13a.2 -5 )
2.4 Dvadasalingatobhadras (fols. 13a.518a. 11)
84
G U D R U N BUI I NI ' - MANN
2.4.1 Type 1 with a list o f the names of deities invoked into the m a n
dala (fols. 13a.5 - 1 5a.3) (Illustration 14)
2.4.2 Type 2 (fol. 15a.3- 13) (Illustralion 15)
2.4.3 Type 3 (fol. 15a. 1 3-15b.6) (Illustration 16)
2.4.4 Type 4 (han'hanitniakadvfulaSalingalobhadnt) (fols. 15b.6 16a,3) (Illustralion 17)
2.4.5 Type 5 (fol. 16a.3 L0) (Illustration 18)
2 .4.6 T y p e 6 (hariluufUmakadvridniialingatobhadra) (fol. 1 6 a .10
16b,2) (Illustralion 19)27
2.4.7 Type 7 (Imrihnriitninkiidvildoiinlingntobimdrii) (lol. 16b.2 - 1 0 )
(Illustralion 20)
2.4.8 Type 8 (hnnharntimikn/[hnriharn]dvfKins;iIingn(ohImdrn) (fols.
16b. 10-17a.6) (Illustration 21)
2.4.9 Type 9 ( IntfilingalobhndnifK(fol. 17a.6-12) (Illustralion 22)
2.4.10 Type 10 (lihgasvastikablm dva) 1 (fol. 17a. 1 2 - 17b.5) (Illus
tration 23)
2.4.11 Type 1 I (lingasvastikabhadra) 2 (lols. 17b.5 - 1 8a,4)
2.4.12 Type 12 (navanabhapadmasvastikamandala) (fol. 18a.4 - 1 1 )
2.5 Sodasalirigalobhadras (fols. 18a. 1 l~19a.3)2"
2.5.1 Type 1 (fol. 18a. 11 18b. 10) (Illustration 26)
2.5.2 Type 2 (sodaSalingodblmvahanlmramandala) (fols. 18b. 10
19a.3) (Illustration 27)
2.6 Saptadasalirigatobhadras (fol. 19a.319b. 12)
2.6.1 Type 1 (fol. 19a.3-1 3 ) (Illustration 28)
2.6.2 Type 2 (fol. 19a. 1319b.3) (Illustration 29)
2.6.3 Type 3 (fol. 19b.3-12) (Illustration 30)
2.7 CaturvimSatilirigatobhadra (fols. 19b. 1220a.6) (Illustration 31)
2.8 Astavim 6ati 1ingatobhadra (fol. 20a.7~8) (Illustralion 32)
2.9 Pancavimâtilihgatobhadras (fol. 20a.8-20b. 13)
2.9.1 Type 1 (fol. 20a.8-20b.3) (Illustration 33)
2.9.2 Type 2 (fol, 20b.3-13) (Illustration 34)
2.10 Astottarasatalingatobhadras (fols. 2 0 b .l3 -2 2 a .6 )
2.10.1 Type 1 (fols. 20b. 1 3 - 2 lb. 10) (Illustration 35)
2.10.2 Type 2 (fols. 2 lb. 10 22a.6)
27 A variation o f this m andala is rcproduccd in M ookcrjcc 1971: 54 (plate 32).
2li For a d ra w in g o f the latalifigodbhavam andala, sec also Rao 1 9 8 8 - 1 9 9 2 , v o l u
m e 5: 23.
V crgati 1982: 58 rep ro d u ces a sodatialingatobhadra from the s k e tc h b o o k o f the
painter V ish n u B a h ad u r Chitrakar o f Bhaktapur, Nepal. T h e bhadra differs fro m the
tw o types e x p lain ed in the BM and is erroneously labelled sarvatobhadra on p. 59.
MA ND AL A S A ND YANTRAS IN S MARTA RI TU A L II
85
2.11 Ekavimsottarasatalirigatobhadras (fols. 22a.6-23a.14)

2.11.1 Type 1 (fols. 22a.6-23a.4) (Illustration 37)
2.11.2 Type 2 (fol. 23a.514)
2.12 Astottarasahasralingatobhadras (fols. 2 3 a .l4 - 2 4 b .l 1)
2.12.1 Type 1 (fols. 23a. 14-24a.8)
2.12.2 Type 2 (fol. 24a.8-24b. l 1)
3 Ramabhadras (fols. 24b. 1 l-3 2 a .9 )
3.0 List o f deities to be invoked (fols. 24b. 1 l-2 6 a .8 )
3.1 Ekamudraramatobhadra (fol. 26a.8-26b,7) (Illustration 41)
3.2 Caturm udrarainalirigatobhadra (fol. 26b.7 - 1 0 ) (Illustration 42)
(see also Colour Plate 7)
3.3 Astamudraramatobhadras (fols. 26b.lQ -27a.7)
3.3.1 Type 1 (fols. 26b. 10-27a,4) (Illustration 43)
3.3.2 Type 2 (astam udraranialingatobhadra) (fol. 2 7 a .4 -7 ) (Illustra
tion 44)
3.4 N avam udraramatobhadra (fol. 27a.7 -1 0 ) (Illustration 45)
3.5 D vadasam u d raram alirig ato b h ad ra (fol. 27a. 10 - 2 7 b .8 ) (Illu s
tration 46)
3.6 Trayodasamudraramatobhadra (fol. 27b.8-1 0 ) (Illustration 47)
3.7 L aghusodasaram am udraram alingatobhadra (fols. 2 7 b .l0 - 2 8 a .7 )
(Illustration 48)
3.8 S odasam udraram atobhadra and sodasamudraramalingatobhadra
(fol. 28a.7-10) (Illustration 49)
3.9 Caturvirnâtimudraramatobhadra (fol. 28a. 1113)
3.10 Pancavimsatimudraramalirigatobhadra (fol. 2 8 a . l4 - 2 8 b . l)
3.11 Astottarasataramalingatobhadra (fol. 28b.2 -9 ) (Illustration 52)
3.12 Sataramalingatobhadra (fol. 28b.9-1 1 ) (Illustration 53)
3.13 Satamudraramatobhadra (fol. 28b. 11 13)
3.14 Astottarasataramatobhadra (fols. 2 8 b .l3 -2 9 a ,3 )
3.15 A sto ttarasataram alirigatobhadras (differing from 3.11) (fol.
29a.3 29b. 10)
3.15.1 General description (fol. 29a.3-14)
3.15.2 Variation 1 (fol. 29a. 14-29b.5)
3.15.3 Variation 2 (fol. 29b.5-10)
3.16 Astottarasahasraramanamatobhadra (fols. 29b. 10-30a. 12)
3.17 Harihararamalingatobhadra (fol. 3 0 a .l2 -3 0 b .6 )
3.18 Astottarasahasraramatobhadra (fol. 30b.6-14)
.
3.19 Astottarasahasraramalihgatobhadra (fols. 30b. 1431 a, 11)
3.20 (Astottara)sahasraramatobhadra (fol. 31 a. 11-12)
86
G U D R U N BUI INIvN4ANN
3.21 Astottarasataramatobhadras (fols. 3 la. 12 32a.8)

3.21.1 Type 1 (fols. 3 la. 12-32a.2)
3.21.2 Type 2 (fol. 32a.2-8)
4 Pancabhadra (fol. 32a.9-1 3 ) (Illustration 66)
5 Siiryabhadras (fol. 32a. 13--32b, 12)
5.1 T ype 1 (fol. 32a. 13--32b.6) (Illustration 67) (sec also C o lo u r
Plate 10)
5.2 Type 2 (fol. 32b.6-12) (Illustration 6 X)
6 Ganapatibhadra viglmnmurdn (fols. 32b. 12-33a.3) (Illustration 69)
7 Cakraravindamandala (fols. 33a.3-34a. 1) (Illustration 70)
8 Svastikamandala (fol. 3 4 a .l-3 4 a .5 ) (Illustration 7!)
9 Lingatobhadras for special occasions and for the attainm ent o f
specific objectives (fols. 34a.5-35h.9)
9.1 TrayodaSalingasanuidbhavamandala (fol, 34a.5-1 0 ) (Illustration
72)
9.2 CaturdaSalingatobhadra (fol. 3 4 a.l0 -3 4 b .6 ) (Illustration 73)
9.3 VimSatilingatobhadra (fols. 34b.635a. 1) (Illustration 74)
9.4 CatvarimSallihgatobhadra (fol, 35a. 1-10) (Illustration 75)
9.5 Sastilingalobhadra (fol. 35a. 10-35b.9) (Illustration 76)
0 Colophon (fols. 35b.9-36a.3)
M A N DA L AS AND Y A N TR A S IN S M A R T A R I T U A L II
1 f
87
F ig u re 1. B asic sq u are unit (p a d a , k o s th a )
Figure 2. E n c lo s u r e (p a rid h i)
F ig u re 3. W e l l ( vapT)
1
H -4-1
F ig u re 4. O f f s e t (b h a d ra )
Figure 5. C r e e p e r (vallf)
cff
F ig u re 6. C h a i n (sr n k h a la )
Eb
=E
F igure 7. C re s c e n t m o o n (k h a n d e n d u )
p
F ig u re 8. P h allic s y m b o l (lin g a ) o f Siva
a,
' :
mm
* i . g
* fen fi
- -mmrmammmmu
n. m m
m n
a.
F ig u re 9. R a m a m u d r a ra ja ra m a
T able: C o n stitu e n t Parts o f the B h a d ra s
G U D R U N BUI IN E M A N N
n j k
.j w
j
*FT
0
*
Ll U L. =3
5 1L1
j 1 5 DB 5
*
L u n
Ll E ' -,
B Ll J
fm m m m w m w *
Sarvatobhadra, type I
'
q | N y i t 8 S ^
I * T F ** "*!
r
i i* . * 'I"
I-- '
I*' "I irm%
*J'
'*
!
1
T
.C-L:
Mi'UTjw
2. Sarvalobhadra, type 2
M A N D A L A S A N D Y A N T R A S I N S M A R T A R I T U A L II
'
1
*
. J
*
*
*
*
i i
**
*-
'
" ,
**
1
i
* :
**
i -
**l
*)
--
*
1
i * :
.1
" 1 " *
_i
1 *
I
' i
* *
* :
I " -
*
%
l "
**
j
- 1
I -
'*
**l
'
* j
* !
1
1
\
j
' H
' ;1
**
I " - I
GUDRUN BUIINI'MANN
# ra
5.
E k a l i g a l o b h ; i i l r : i l u i l t u i a n n t i l a k a
6. Catur 1ingatobhaclra h r h a d ^ a u n t H a k a
M A N D A L A S A N D Y A N T R A S IN S M A R T A R I T U A L II
7. G auritilaka
Wg
Ug Ho
t.
92
GUDRUN BUIINHMANN
U
*1 5 F tl 55*1. 0'
9. Caliii'lirignlobhadra. type 2
10. Caturlirigatobhaclra, type 3
11. C aturlingatobhadra, type 4
=grf%%Tfd*:U
\\
m r p l*
ru d rrzD u
n s r n
Bi _:h B. * THi . I ! ! I j B r r i ''*!B
m
& i m-i i -i
-is i
^ r - - r r H a ~ i ~ ri ^ ^ r r n ^ ^ i T -
m m *i '-iî'-ii
n~ r-R 12. Astalingatobhadra, type 1
94
GIJDRUN BUHNEMANN
srg iflso fd j
13.
A sta lir ig a to b h a d n i, ty p e 2
14. DvadaSalirigatobhadra, type 1
\%
,
15. D v a d asalin g ato b h ad ra, ty p e 2
16. Dvadasalingatobhadra, type 3
GUDRUN BUHNIiMANN
17.
Dvadasalingiitobhadni, type 4
s ra
18. Dvadasalirigalobhadra, type 5
*STtt V} WT?IT^
Sn^^RTh^To \\
19. D v ad asa lin g a to b h a d ra , type 6
H fim rh ^ to
20. D vadaS alingatobhadra, type 7
\o
GUDRUN BUHNHMANN
21. Dvadaalirigalobhadra, type 8
22. DvadaSalingalobhadra, type 9
23. D v a d asalin g ato b h ad ra, type 10
26. SodaSalirigalobhadra, type 1
G U D R U N B U H N I 'M A N N
27. Snda&ilingatohhadra, type 2
aro \jo
28. SapiadaSalingatobhadra, type 1
am
<-y k
GUDRUN BUllNl'M ANN
3 I. CaUirvimsatilirigalobhadia
w v*.
t *
32. Astaviinfialilingatohhadia
33. P an cav im satilin g ato b h ad ra, ty p e 1
sm
w firj
su
104
G U D R IJN B U IIN K M A N N
35. As|ottaniSul;ilingntobhacira, type 2
MA ND AL A S A ND Y AN T RA S IN S M A R T A R IT U A L II
105
\c
37. Ekavimsattarasatalirigatobhadra, type 1
GUDRIJN BUI 1NHMANN
4 1. Ekaimidnlnnnntobhadni
w*f Ml ^HTTO#
42.
C a lu n n iK lr a n lin a lih g a to b h a c Ir a
M A N D AL A S AND Y AN TRA S IN S MA RT A R I T U A L II
4 3 . Astamudraramatobhadra, type 1
107
108
G UDRUN BUHNEMANN
44.
A sla in u c lrn ia m a to h lK id n i, type 2
45. N a v a m u d ra ram a to b h a d ra
46. Dvadasamudraramalirigatobhadra
GUDRUN BUHNKMANN
47. TraycKlaSamtuiianimiilobhadra
bto Hi < y h m t m 1<ifi'intw'Jiwrwq vc
48. Laghusodasaranianuidraranialirigntobhadra
49. S o d a sa m u d ra ra m a to b h a d ra
111
112
G I J D R U N BUI 1N H M A N N
52. AsloUara6auitamalirig:itobli;idra
66. P a n c a b h a d ra
am \
tf|4 U |l4 U m H
67, Suryabhadra, type 1
S '*
GUDRIJN BUHN1IMANN
g fo y fliW H
68, Suryiihhadra, iypc 2
8W V ^TTOT%
69. G anapalibhndra vi^h n a m a rd a
\c
70. C ak rarav in d am aiid ala
mjtm
71. Svastikamandala
G U D R U N BUI I N H M A N N
am
vs*
72. TrayodaSaliiigasaniudbhavaniandala
73, CatuidaSaliiigatobhadra
wr
g im . w
.h
75. Catvanmsallingatobhadra
117
7 i b k k k h IIb
| Hil:KiSCicir' .
BBBKKbiB B
I S 'i i i K i i B ' a a a
m - nKBkiS a d
ItC SltK B ;_B B
> IBmmtNkMIVMMII NMklNWW:H
CO
K*aitmaitKt:;isicKKioai . , _
KBKBtiB SfcKli 1C1C!' * . SSISfefen b 5 b ^ S
''ICB'itS; BlfaKKKKiltB

BlfelfclClfiiCH.fi_BlLHfc.BBB
B KBKifcS BitKlfcfckWtte WICM 1CIB14 ICICH lCftttlbijjICfcB BjblCBK B
A
&
i
<n
An
B U H N E M A N N
S a s ti li rig a io b h a d ra
| a KB a KBKBttKBKlB BBKBKKibB BBJUBBJBiflHBH B BK B ,

*i B
~ BKlCBBBftB' BHKB BBKB' BUBHBIBBB B
i B BB I KBKKKIhBB BUKHBBBBBKUB BBBiKKlEBH B SB B
. BBKBBBBEBKKKIBBBBBiHBKn BKKUBjBBBKlCKUfiUBBBBKBBB
, B KBKBHitfBKKHBBBBBBBKiiSBBBBBfiBBBBBliKKBHIIiiBHBB B
I SKKSBBKHB BttBBiijBiB KKB uitbttfbBiB BifaKBUUUHB
I B KBKlSKftBBifi BAKKWttB B BBiUKURtS BUBkiKWKBK B
PBBKSKKKBBa!' BBiKKKKBBB BBB BBBKKUKIUB BBKKKKBKBBB
| l KJEKfeSKKBBBKlCKKKKKB B BKKKKKUBlBBBHlKKKBK B
I . BKHttiBBBBKBBBBfiBBBBBBfiBBBBBBBBBBBBBBBHKB
I lEBKBiB BftKiiUAIfckB KttiK KKK HKJi JBBWBVBUB BUKBH B
BBKBKiB BBKKBiiicB B B B BttttBBBBB BUBKBBB
EcBUBBB B&iKKKKKBBB BBB BBB BBB jlBBHKHKHUB BBBKBB
iiaBiait.KKHKKKKB B . .B B ; BKKKKKUKUBBfiUBKU
BBBBBBBBBBBBBiKK
ffi 8ES " " ,' E0ES 6SSBE EESB5
EB MKBttfJ BWK
* -----a n vbb
(
b i h i . ,
L > S 5 BK B U B Bib
* SSSSSS* " 5
bbbuu Eb B
iBBHaKKKEB-------B, B. B M . B____B . B BKKMKtfcBB
GUDRUN
76.
3
B MBiCKKIfeBBBibKKKtiKKa
IBB BBBKKKKIhB BBKKKKBJCBBB
BBBKfitEKtCKBB BfcKKKKBBB
B kSKKVifeSBft BBVKJbttB B BUBKKKlB SBBKKBKBK B c(S
BKKfclfcBKiBB BlifclfcMiSiBB BSKB
K B kuifaBiaiBiB
__________ BttKBiiifcKKB
B KBKBifci^lCfeBBBBfiBBBBBRBBBBBBBBBfilfciitfBfcfcBkiBlfiB
BVBtCBflBBfcBttKKffaBBBIfaKKKi BBICJfiBBBfiltlLlCICBttBJBBKIIBB s 7
b tea a b k k k b b b u jb b b iBBBfefl b b b k k k b b a Eh. a
5
BKKtalfeBfeB BlfaUB BKUB BBBIHtiBKB B
B KB B KSKBRittBlfllB BttBK IEKKB BlbKBllililBBBK B BK B
*
a kb m j b bl KBisBieiKBiiBiaaaaBBft'ieBHBB. a b k b bis b A
6
B BBBBUBfcKHKBKi&EKaBltB H
B
B BiB B KB KB B KSBjK$y<KttK B BB B8 BK
BBBKBBBB'KBBBBKB , BUB a&KBHB . . BKBBBBKBBBBKBBB
B KB B KB B MB B KBBiK%~*4ftCfiJBH B BK Bi K B BK B.." w
. B i. B BBaBKBaKlflHBKfcBKaBBBl B _ B
B KS a KB B UBKBKbBHiB>KUBiB>UBttKatCBB B BK Bl BK B
M A N D A L A S AN D YAN TRAS IN THE P A N C A R A T R A

TRADITION*
Marion Rastelli
T h i s paper deals with the preparation and use o f man dal as and
yantras in the Pancaratra tradition. The sources for this investigation
are the Pancaratra Samhitas as far as they have been edited and were
available to me.
Both mandalas and yantras are diagram-like, often symm etr ic
drawings that are invested with mantras. A mandala, also called pura,
y a g a , and cakra,' is usually prepared by means o f applying po wde r
u p o n an immobile place. It serves as a place o f worship. The deity is
made present in the mandala by imposing mantras that represent him
and his aspects upon it, and he can then be ritually worshipped there.
A yantra is usually drawn upon a mobile material and can thus be
carried around. Mantras are written on it, and the drawing and the
mantras form an integrated whole that represents the deity or one o f
his aspects. Yantras are often used as amulets, which protect their
wearers and help them to the fulfilment o f their wishes.
The Construction o f M andalas
Before drawing a mandala, the practitioner must determine a suitable
place and prepare it. The Pauskara-Samhita, which is to a great part
I am grateful to C ynthia P e c k -K u b a c z ek for reading the E n g lish m a n u s c rip t and

su g g e s tin g various corrections.
1
F or llic usage o f the term pura for m an d alas, cf. R astelli 2 0 0 0 b : 375, note 57.
T h e term yaga is used m ainly in the P au sk ara-S am h ita, sec, e.g., P a u s S j 2.8 c, 4 .1 3 d ,
14d, 96a, 5.5a. I f not indicated otherw ise, I use the P a u s k a r a - S a m h i t a s edition from
1934, as the new edition from 1991 contains only the first 26 adhyayas. T h e usage o f
cak ra in the sense o f m andala is rare (e x a m p le s arc P a r a m a - S a m h ita 7.68b, 8.7c, 8c,
P a r a m c iv a n i- S a m h ita 17.499a, 501a). In the context o f m a n d a la s, c ak ra m o re often
d e s i g n a t e s a c irc lc that f o r m s a part o f a m a n d a l a ( J a y a k h y a - S a m h i t a 3 0 .5 1 c ,
P a r a m a -S a m h ita 6.2 3b, P a d m a -S a m h ita , caryFipada 2 5 . 134a).
120
MARION RASTELLI
devoted to mandala construction,2 provides very elaborate instruc

tions on how this should be done. Places that are suitable for the
cons tru cti on o f a ma ndala arc a mo un taint op, a cave at the
mountaintop, a pleasant forest abounding with trees and herbs, a
lovely grove, a lotus-lake, a rivcrbank, a confluence o f two rivers, a
ford (tirtha), a temple (dcvatiiyatana), a meeting place (gostha), and a
her m itag e o f Brahmins that is free from evil beings, thieves,
diseases, and d a n g e r s .'1 In general, these places are con sidered
sacred, and similar lists arc also given for places that arc suitable for
worship o f the deitys descents (avafara).4
It is important that the chosen place is without any faults as
otherwise worship performed there would not bring the desired
results.5 Free from any faults means, according to the PauskaraSamhita, that the place should be provided with auspicious trees,
fruits, flowers, soft young grass, tender herbs, and various species o f
lotuses. It should be pleasant to the touch and be supplied with
cooling water, fragrances, and good flavours. It should not be saline,
burnt by fire, a meteor or sunrays, or vile-smelling. The place should
bestow ease, peace, and delight like the sight o f o n e s beloved, It
should be inhabited by beautiful birds, cows, deer, tigers, elephants,
and huma n beings, and should be free o f evil creatures. East o f the
place there should be a lake, west o f it a forest o f mango trees, north
o f it a large thicket and south o f it a mountain. Moreover, in the
north or north-east of it there should be a well or a water tank. The
soil should be even and incl ined to the north-cast, o f a single colour
and free o f serpents and similar crcaturcs. In the summer it should be
as cool as the moon, in the winter, hot, and in the rainy season,
without m o i s t u r e , 6
2 M a n d a la c o n s tr u c tio n h as b e e n g iv e n the s a m e a tte n tio n in the P a u s k a r a S am h ita as te m p le construction has been given in the later Padm a-S am hita.
3 P a u sS ] 2 .4 - 7 b . F or n cccssary em e n d a tio n s in this passage, see Rastelli 2000a:
120, note 75.
4 Cf. P a u s S ] 3 4 . 1 - 2 and 3 6 .2 3 8 - 2 3 9 c . F or translations o f the three passag es o f
the P a u sk ara-S am h ita m en tio n ed here and in note 3, see Rastelli 2000a: I20f.
5 Cf. P au sS | 2 .10cd: H o w e v e r, a faulty soil m ay p ro d u c e m a n y o b s ta c le s . (sa
tu d o s a v a li b h u m ir vig h n n n u tp fid n ycd bah u n II) and 2.26: T h e re fo re , a d ifferent
[soil w ith the] o pposite [qualities] alw ays b e sto w s u n d esired results. [It] sh o u ld be
refused from afar sincc it inhibits success. (nlo n yii vip u n lii casam 'stapha/ada sadfi
/ duratah pariharlavya sid d h ih a n ika ri ya ta h II)
h Pa u sS i 2 . 7 c - 2 0 b and 24ab. P ausS ] 2 .20b rasanvli: 'rasa ya. A p t e s em e n d a tio n
rasanvita is not possible on account o f the metre.
M A N D A L A S A ND YANTRAS IN T H E P A N C A R A T R A
121
As described, this place is certainly an ideal and found only rarely

i f at all. Other Samhitas, being more realistic, m ak e less heavy
dema nd s on the place for a mandala and only prescribe that it should
be pleasant, even, oriented toward the north-east, and free from
thorns and other faults.7 Even the Pauskara-Samhita concedes that if
o ne cannot find a place with all the described qualities, one can also
m a k e do with an ordinary one.x
Before the construction o f the mandala, the soil is tested and
prepared just as before the construction o f a building.9 The PauskaraSamhita describes several methods for testing the suitability o f the
soil ( bhupanksa), One o f these methods is digging a hole and filling
it again with the earth taken out o f it. The best result is i f the earth
overfills the hole. It is acceptable if the ground is even after filling
the hole, but if the hole cannot be filled with the earth again, the
place should be avoided.1" Another method is to sow a seed. The soil
is most suitable if it shoots forth within three days, a m edi um result
is i f it shoots forth within five days, the worst if it shoots forth within
seven days .11
The place is to be worshipped with mantras and oblations and, on
an auspicious day, ritually acquired. For this ritual acquisition, one
mu st first distribute b a li offerings for the cruel beings w h o are
present there in all quarters, and ask them to go a w a y .12 The n one
digs the ground to the water limit and fills it again. After that one
grows grass, rice, and shoots o f trees. Then one puts a herd o f cattle
out to pasture for three days. After the three days, the ground is
ploughed and filled with burnt bricks. Upon them one scatters gold
dust, silver dust, and jewel dust, fragrant flowers, grain, rice, fruits
7 E.g., J a y a k h y a -S a m h ita 23.69cd, V isn u -S a m h ita 9.2.

* PausS i 2.27.
v Cf. P a u s S ] 2 .7c8: At the b e g in n in g , he should test i f the soil there, w h ic h [he
h as] e x a m i n e d well, is e n d o w e d w ith f a v o u r a b le s ig n s fo r the sa k e o f a h o u s e o f
gods, fo r the sake o f a village for B ra h m in s , etc., for the sak e o f w o r s h ip p in g ya g a s
(i.e., m a n d a l a s ) , fo r the s a k e o f a h o u s e o f a h o u s e h o l d e r . ( tatra b h u m im
p a rlk sy iid a u la k sa n a d h y a m s u la k s ita m II 1 d c v a n a m a la y a rth a n i tu g ra m fir th a m
b ralvnanadisu / yajanartham tu ya g a n a m grhartham g rb a m e d b in a m // 8)
10 T h is p ro c e d u re is m e a n t to test w h e t h e r the lan d is v e ry lo o se or s a n d y ; cf.
A c h a ry a 1946: 384.
11 P a u s S ] 2.21 c23. P a d m a -S a m h ita , kriyapada 1 .3 2 - 3 6 g iv es the s a m e m e th o d s
for testing the soil before constru ctin g a temple; cf. also A c h a ry a 1946: 3 8 3 - 3 8 5 .
12 P a u s S ] 2 .3 0 - 3 4 c . Cf. P a d m a -S a m h ita , k riy a p a d a 1 .4 0 - 4 9 a fo r th is rite b e fo re
the c onstruction o f a tem ple.
122
MARION RASTELLI
and other substances. 11 The ground is made even, sprinkled with

water, rammed by means o f pestles, besmeared with clay, cow-dung,
and water and swept with the hand or a cloth.1'1
The mandala is to be constructed in a pavilion ( mandapa) that is
adorned with canopies, banners, etc.15 In the centre o f the m andapa, a
platform ( vedi) is made o f wood and clay and besmeared with the
five products o f the cow (milk, sour milk, butter, and the liquid and
solid excreta) or with cow-dung and water.1'1 On the platform, one
first draws the x-axis, then the y-axis, the diagonal lines, and the
boundary lines in order to obtain a square. This square is divided into
several other squares in which the various elements o f the mandala
are d r a w n .17 The preliminary grid is made by means o f threads that
are besmeared with a substance and thus leave lines on the floor
when laid on it. 18 The mandala is drawn by means o f strewn powder
(rajas). First the lines are drawn, and then the pattern is filled with
powders in various colours.1'7
While strewing the powder, one must be very careful lo apply it
evenly. The Visnu-Samhita describes in detail how the mandala
should be drawn and which mistakes are to be avoided; Many faults
[arise] through protuberances, [lines that are] crooked, [too] thin, and
[too] thick, cavities, elevations, irregularities, and holes. In no case
[should] the lines [be] interrupted. If there are protuberances, etc., on
13 P a u s S ) 2 .3 4 d - 3 9 . The lext is corrupt and thus not alw ays intelligible.
MP a u sS j 2 .4 2 c "4 4 . F or sim ilar rites before the construction o f a tem ple, cf,, e.g.,
P a d m a -S a m h ita , kriyapada 3 .1 -3 7 ,
15 Cf. J a y a k h y a - S a m h i t a 13.14, P a u sS ] 4 . 2 I f f ., L a k s m i- T a n tr a 3 7 .3d, 5 5 .10a,
V isn u -S am h ita 9 .3 - 5 b , A niru d d h a-S am h ita 7.3ab.
16 J a y a k h y a - S a m h i t a 13.1 5c16, P a d m a - S a m h i t a , c a ry a p iid a 7 .2 7b, V isn u S am h ita 9,5b, V iSvam itra-S am hita 15.26a. For the v e d i as support o f the m an d ala,
see also J a y a k h y a - S a m h ita 18.36a, ISvara-Sam hita 11.133c, 18.171c, A n ir u d d h a Sam hita 7.3c and 8a.
17 J a y a k h y a - S a m h i t a 1 3 . 1 7 - 2 4 , L a k s m i - T a n t r a 3 7 . 4 c - 1 3 b , T g v a ra -S a m h ita
1 1.1 3 5c140. For an e x a m p le o f a p re lim in a ry grid, cf, the d ia g r a m s in G u p t a s
translation o f the L aksm i-T antra, pp. 228 and 230.
111
Cf, P a d m a -S a m h ita , caryapiida 7.9c: candanardrani sutrani, th read s m o isten ed
w ith s a n d a l w o o d . N o t all S a m h ita s m e n tio n the s u b s ta n c e for b e s m e a r i n g the
threads explicitly. Ja y a k h y a -S a m h ita 13.16c gives only: hav in g first b e sm e a re d the
thread ( upa lip ya g ra ta s sutram ) and L a k s m i-T a n tra 37.5a m e n tio n s a very w h ite
th re a d ( s u s ita m su tra m ), w h ich p ro b a b ly m e a n s a thread b e s m e a re d w ith a w hite
substance. For the m a n n e r o f draw ing lines and circles by m eans o f threads, cf. also
B ru n n er 1986: 26 (cf. Brunner, p. 170) and Apte 1973: 503f. or A pte 1987: 130f.
|,J J a y a k h y a - S a m h i t a 1 3 . 2 5 - 4 0 , L a k s m i - T a n t r a 3 7 . 1 3c19, l i v a r a - S a m h i t a
11.141 161b. A c c o rd in g to the P aram a-S am h ita, the m a n d a la u sed in the sa m a y a diksa can also be draw n with paint (varnaka) (P a ra m a -S a m h ita 8 .7 c -8 b ),
M AN D AL A S AND Y ANTRAS IN T HE P A N C A R A T R A
123
accou nt o f carelessness when the mandala is drawn by disciples, then

t h e competent teach er21 [should be] alert. H e should have [the
disciples] position the lovely pericarp, the filaments and the other
[parts o f the man dal as lotus] by applying [the powder] only once
a n d not by moving [it] to and f ro .21 In addition, the various
co ns ti t u en t parts o f the m andala should be applied in var yin g
thicknesses: He should make the pericarp and the lines o f the leaves
aro u n d [it] elevated. The elevated lines are threefold, namely, equal
to the little finger, to the middle finger, and to the thumb. The
diminution o f the [lines] with regard to the highest, the middle, and
the lowest [elevation] should be modified in the order beginning with
the lotus.22
T h e powders used for drawing the mandala are made o f various
materials such as herbs, jewels, stones, leaves, fruits, seeds, fragrant
substances, wood, metals, and minerals.23 The Naradlya-Samhita and
the Bhargava-Tantra state in detail which materials should be used
for wh ich colours: ground rice or sanda lwoo d po w d er for white,
safflower for red, pure turmeric or turmeric mixed with a bit o f white
substance for yellow, burnt rice or sacrificial charcoal for black, and
green leaves for green (gyama).24 Apart from powder, materials such
as whole seeds (mustard seeds, sesame seeds, grain) or flowers can
also be u s e d .25 Flowers are especially used for mandalas employed
during initiation (dlksa) or a festival ( u tsa va )u
211 D r a w i n g m a n d a l a s is p r im a r ily a ta sk o f the t c a c h e r (guru, a ca rya ), w h o
b e l o n g s to the highest level o f initiation {cf. Rastelli 1999: 1 5 3 -1 5 8 ); cf. Ja y a k h y a S a m h ita 17.56ab, S atvata-Sarphita 21.49, and A h irb u d h n y a -S a m h ita 20.6c.
21 V is n u -S a m h ita 9 ,3 9 -4 1 :
gra n th iva ktra krS a sth u la n im n o n n a ta vim isra n a ih /
chic!mis ca bahavo dosii rekh a cch cd o na sarvatha II 39
p ra m a d a d yatra sisya n a m m a ndalasya hi Ic k h a n e l
g ra n th ya d i syiid atandrl tu sa m a y c k u ia lo gu ru h 1140
ka m ik a k c sa ra d ln i lalitani p ra y o ja y e t I
sakrdarpanam atrcna na p u n a h parivartanaih I I 41
22 V is n u -S a m h ita 9 .4 2 -4 3 b :
k a m ik a m ucchritam ku rya l p a tra rckh a i ca sarvatah I
ka n isth a m a d h ya m a h g u sth a n iita rckhiis trid h o cch rita h l/ 42
m u k h y c m a d h y e d h a m c b ja d ikra m a d u h ya s ca tatksayah /
T h is m e a n s that the m a n d a la s ccntre is the m o s t elevated and its edge the lowest.
11 P a ra m a -S a m h ita 7.61, V isn u -S a m h ita 9.37.
MN a ra d ly a -S a m h ita 8.59c(51 b, B h a rg a v a -T an tra 1 3 .17-18.
25 S a tv a ta -S a m h ita 17.50c51.
*
F o r flo w ers used for m a n d a la s e m p lo y e d d u rin g th e initiation, cf. S a n a tk u m a r a S a m h ita , indraratra 9 .8 7 c - 8 8 c , rsiriitra 5 .1 - 1 7 , N a r a d ly a - S a m h ita 10.6b; f o r flo w e rs
124
MARION RASTELLI
There are many different forms o f mandalas. The Paus kar aSamhita, for example, gives a great choice o f varying m a n d a la s .27
The mandala that is most often mentioned in the Pancaratra texts is
the cakrabjamandala , also callcd bhadrakam andala . 2K The centre o f
this mandala consists o f a combi nation o f a wheel (cakra) and a lotus
(abja) that is then surrounded by three square enclosures with doors
in the four quarters. The ccnlral combination o f a wheel and a lotus
is made up o f five concentric circles. The innermost circle is the
pericarp (karnika) o f the lotus. In the second circle are the lotus
filaments (kesara), its petals (dala), and the hub (nabhi) o f the wheel.
In the third and the fourth circle are the spokes (ara) o f the wheel,
and the fifth circle is the felly o f the wheel. In the two inner
enclosures around the five circles, there are twelve lotuses each, one
in each corner and on the left and right o f the doors (dvara). Conch
shells (Sanklia) are drawn in each o f the corners o f the outermost
enclosure.29
The use o f different mandalas in worship effects different results.
Thus, the kind o f mandala may be chosen according to the result one
desires to achieve. The Pauskara-Samhita, for example, distinguishes
the mandalas described in it as follows: The first mandala is known
as auspicious in every w a y as it causes prosperity. The second
[mandala] is called liberating from sins as at the sight o f it, the sin
acquired in many births comes to an end. The third [mandala] is the
good p at h and shows the way o f the dharma. The fourth [mandala]
is called dharma. On account o f [its] worship, it bestows the dharma.
T h e fifth [mandala] is the w o m b o f w e a l t h [and] effects the
increase o f the bhaktas wealth. (...),,3(l
u sed fo r m a n d a l a s e m p l o y e d d u rin g a festival, cf. U v a r a - S a m h i t a 1 1 .3 8 9 - 3 9 0 ,
V isn u -S a m h ita 2 0 .6 6 c - 7 5 b , V isvakscna-S am hila 2 6 .1 67c201.
See P ausS ] 519.
a S ee the usage o f the designation blmdraka for this m andala in P a d m a -S a m h ita ,
caryapiida 7.38c, N a ra d ly a -S a m h ita 8.53a and V iSvam ilra-Sam hita 15.34d.
*
P a d m a - S a m h i t a , caryapiida 7 . 1 2 b - 3 8 b , N a r a d l y a - S a m h i t a 8 . 2 - 5 3 b , ViSvam itra -S a m h ita 1 5 .6 c -3 4 b , B h arg a v a -T an trn 13.526b. PausS | 8 d e s c r ib e s v a rio u s
kin ds o f cakrabjam andalas. F or a d ra w in g o f the cakrabjam andala a c c o rd in g to the
y o u n g e r Sam hitas, see C o lo u r Plate 14. A n o th e r exam ple o f a m an d ala often used is
the n avapadm am andala (sec C olour Plate 15).
30 PausS i 5 . 2 - 4 :
m andalam sarvatobhadram bbadrakrt p ratham am sm rta m /
a g h a n in n o ca n a m niima d v itiy a m y a sya dar& m it II 2
a n cka ja m n o p a tta m tu ka lm a sa m ksa y a m cti c a l
sa d a d h va m s y a t trtiynm ca d h arm am argapradarianam /l 3
MAN D AL A S A ND YANTRAS IN THE P A N C A R A T R A
125
Another reason to choose a particular mandala may be the kind o f

ritua l that is to be performed, as some m andalas are con sidered
p articu larly suitable to certain rituals: When [the god] is sent to
slee p (prasvape) one should prepare the best yaga (i.e., mandala)
called svastika, o twice-born, when [he] is awakened, [the mandala]
ca l l ed distinction or one o f the design-complexes, at the festival,
t h e cakrapahkaja (i.e., cakrabja) with one or more lotuses.31
Finally, the man dal a differs accor di ng to the deity that is
wor shi pp ed on it:. This is relevant especially with regard to the ritual
o f the sadhaka who worships a particular mantra or deity in order to
ga in siddhis. The mandalas used here are specific to the mantras
w o r s h i p p e d , and sometimes their shape even cor responds to the
m a n tr a that is worshipped in it. Thus, the ma ndala o f the sankham antra, for example, has the shape o f the conch shell or the mandala
o f the gadamantra is surrounded by eight ma ces .32
The Use o f M andalas
A mandala serves as a place where the deity can be invited and then
w o r sh i p p e d .33 In addition to a mandala, there are several other places
that can be employed for this purpose. Pots ( kum bha, kalasa)y statues
(b im b a , pratim a), and fire ( vahni, agni) are most often used, but also
o t h e r places are possible such as the disk o f the sun or a rosary
( aksasutra).M Often the deity is w or shipped in several places one
after the other. In the daily ritual described in the Jayakhya-Samhita,
d h a rm a k h y a m sy n c caturtham tu dharm am y a c c h a ti p u ja n a t!
p a n c a m a m vasugarbham tu bhaktanam va su vrd d h ikrl // 4
F o r d iffe re n t results s te m m in g from the w o rsh ip o f d iffe re n t m a n d a la s , see also
J a y a k h y a - S a m h ita 1 3 .4 4 -4 7 b and Padma-Sarnhitii, caryapada 7 . 8 5 c - 8 9 .
31 Paramc&vara-Snmhita 12.575576b:
p ra sv a p e sv a stik a m nam a ku ry a d yag n va ra m d v ija l
p ra b o d h c tu v iv e k a k h y a m v y u h c s v cka ta m a m tu va U 575
u tsa vc h y cka p a d m a m va naikfibjam cakrapankajam /
F o r the s v a s tik a m a n d a la , see P ausS ] 5.151 b177c, fo r the v iv e k a m a n d a la 5 . 1 0 7 b 117c, fo r the d c s ig n - c o r n p lc x c s (vyuJia), P a u s S j 7 (the tran sla tio n o f v y u h a b y
d c s i g n - c o m p l c x follow s Aptc; cf. his translation o f P a u s S 2 7).
32 J a y a k h y a - S a m h i t a 30.6 3 and 42 ab . F o r the s h a p e s o f the v a r io u s m a n t r a s
m a n d a la s, cf. Rastelli 2000b; 325 and 3 6 0 - 3 7 1 .
11 E.g., Ja y a k h y a -S a m h ita 13.1, P a r a m a -S a m h ita 4.26.
;y F o r th e disk o f the sun and a rosary as places for w o rsh ip , see P a ra m a - S a m h ita
4 .2 6 c and J a y a k h y a -S a m h ita 13.4a.
126
MARION RASTHLLI
for example, the deity is worshipped in a mandala, in a pot, and in

the fire.35
In younger Samhitas, a hierarchy of suitable places for wor shi p
has been established, especially o f the four places ( cntuhsthana), as
mandala, pot, fire, and statue arc called.' Worship o f the deity in a
mandala is considered the best, then follow the pot, the fire, and
finally the statue. On the other hand, mandala worship is necessary
only on special occasions; in other cases one could be content with
the de it ys worship in fire and a statue or only in a statue: A t an
initiation, a festival, an ablution, the fruit festival,'17 the raising o f
the damana [flowers], the great llower festival, a spccial festival
o f the purifying thread, the ja y a n tf [festival],'w at the time o f the
krttika festival,411 at a lunar or solar eclipse, an equinox, a solstice,
particularly on the twelfth day o f a half-month, in all expiations, and
at the spring festival one should perform the worship in a mandala.
Otherwise, [the ritual] may cause faults. Worship [of the deity] in a
mandala is the best one, worship in a statue the lowest. Worship [of
the deity] in a pot, a mandala, a statue, and fire is the very best.
Through the very best worship, all faults are destroyed, all sins o f
men are removed, [and] the ki n gs country is promoted. A ritual
without worship in [all] four places may be fruitless, Therefore, with
all [ones] effort one should perform the worship in the four places.
Four places are the best, three places medium, two places the lowest,
the remaining single place mere appearance. A statue is taught as the
single [place], a statue and fire are the two places, [these two]
together with a pot are the three places, [all these and] a mandala are
the four places.41
35 Cf. Rastelli 1999: 2 7 1 -3 2 2 .
y Cf. P a d m a -S a m h ita , caryapiida 21.69.
37 Cf. A n iru d d h a -S a m h ita 22.9cd and 23.1 Oc26.
J* Cf. A n iru d d h a -S a m h ita 22.7cd and 23.1-1 Ob.
3<;Cf. A n iru d d h a -S a m h ita 22.14cd and 2 5 .5 c -2 7 .
^ Cf. A n iru d d h a -S a m h ita 2 2 .1 6cd and 2 5 .2 8 -4 9 .
41 A n iru d d h a -S a m h ita 6 .5 6 c -6 4 b :
d ik sa y a m u tsa vc caiva snapanc ca p h a lo tsa vc // 56
dam anH ropanc caiva tathii p u sp a m a h o tsa v c /
p a v itro tsa v c v iie s c ja y a n ty a ip ca tathaiva c a !
k rttik o tsa v a k a lc tu grahanc so m a su rya yo h II 57
v is u v e ciiyanc caiva dvfidasyiini tu viScsatah /
p ra y a scittcsu sa rvcsu vasantotsava ova ca II 58
m andaliiriidhanam k u ryiid anyathii d o sa krd b havct /
m andaliiradhanam srestham ja g h a n y a ip b im b a p u ja m m II 59
M AN DAL AS AND YANTRAS IN T H E P A N C A R AT RA
127
As in the other places that are suitable for the deitys worship, the
deity must be made present on the mandala. This is achieved by the
imposition (nyasa) o f the mantras of the deity and o f his various
aspects such as his retinue, his weapons, and his ornaments. The
d e i t y s main mantra is usually placed in the mandalas centre and the
m a n t r a s o f his retinue, etc., around him. An e x a m p l e is the
cakrabjamandala (cf. p. 124) described in Naradlya-Samhita 8, 6 1 c73. Visnu is in the mandalas centre, that is, the pericarp (karnika) o f
a lotus. His consorts rl and Pusti are on his right and left side, that
is, south and north o f him. V i s nu s twelve limbs ( aiiga), viz., heart
( hrdaya), head (siras), tuft o f hair ( sikha), armour (kavaca), weapon
( astra), eye (drs), belly ( udara), back (prstha), arms (bahu), thighs
( uru), knees (janu), and feet (pada), are on the petals o f the lotus.42
V i s n u s quiver ( isudbi) and sword (asi) are south o f the lotus, his
b o w ( dhanu) and shield (kheiaka) north o f it. The garland o f forest
flowers (vanamala) is east of the lotus, and srivatsa (a curl of hair on
V i s n u s breast) and kaustubha (a jewel on V is n u s breast) are north
and south o f the garland o f forest flowers. The conch shell (sankha)
is at the directional points. Garuda is at the eastern and western door,
V i s n u s discus (cakra) at the southern door, his mace (gada) at the
nor the rn door. The eight guardians o f the quarters (d ik p a la ) are
outside the man dal a.43 The 25 tattvas arising from the primary matter
kum bhnm nnclnlnbim b,rign;ui p u ja n a m cottam am b h a v c tl

utiam aradhancnaiva sa rva d o sa ksa ya m ga ta m // 60
sarvapiipaharam p u iv sa m riijanlstravivardhanam /
catuhsthanurcanahinam karm a n isp h a ku iim n a y c t // 61
tasm al sarvaprayatncna catiihstlulnarcaim m y a jc l /
c a tuhsthanam uttam am ca tristhanam m a d h y a m a m b h a v e t 11 62
d v isth a n a m a d h a m a m caiva cka m a b h a sa ka m param /
c k a m tu b im b a k a m p ro k ta m d visth a n a m b im b a p a va ka m II 63
tristhanam k u m b h a sa m y u k ta m catuhsthanam tu m a n d a la m l
Cf. also IS vara-Sam hitu 13 .8 8 c 89: W ith reg ard to the p rin c ip a l rule, o n e sh o u ld
w o rs h ip the all-p erv ad in g one in a pot, a m andala, a statue, and fire. W ith regard to
the s e c o n d a ry alternative rule, one should w o rsh ip [the deity] in a pot, etc., w ith the
e x c e p tio n o f the m a n d a la , or one sh o u ld w o rs h ip [him ] in a sta tu e and in fire.
( k u m b h c ca m a n d a le b im b c vahnau ca y a ja n a m v ib h o h H 88 m u k b y e k a lp e p ra ku rvlta b y a n u ka lp e tu m a n d a la m l vina k u m b h a d is u y a je d b im b a v a h n y o s tu va y a j c t / I
89) Cf. also, w ith the sam e content, P a ra m e s v a ra -S a m h ita 1 6 .20c-22b.
42Cf. N a ra d iy a -S a m h ita 3 .8c1 lb.
45 F o r their nam es, see, e.g., N a ra d ly a -S a m h ita 13.3 1 8 -3 1 9 .
128
MARION RASTELLI
(prakrti) are in the vithi." Siva, Brahma, and Visnu arc in the hub o f
the mandalas wheel, the twelve lords of the months (m asesa), Visnu,
M a d h u s u d a n a , Triv ikn tma , Vamana, Sridhara, II rsikesa, P ad manabha, Damodara, Ke.âva, Narayana, Madhava, and G o v i n d a '5 on
the spokes, the moon with its sixteen digits (kalii) and its light and
dark half between the spokes (?), the sun (surya) on each spoke at the
edge as well as the twelve Adityas, viz., Dhatr, Aryaman, Vidhatr,
Mitra, Varuna, Bhaga, Indra, Vivasvat, Savitr, Pusan, Tvastr, and
Visnu.4<
Since the mandala is a placc for the deitys worship, it is used in
many different rituals. According to the J a ya kh y a- S am h it a ,47 the
daily ritual (nityakarman) consists o f preliminary rites such as the
bath (snana), the purification o f the elements (bhutaSuddhi), and the
imposition o f mantras (mantrany;Isa) upon o n es body, o f the mental
(manasayaga, antaryaga) and the external worship (bahyayaga), and
o f the fire ritual (agnikarya).
Mental worship is composed o f the same parts as the ensuing
physical worship. Mental worship, however, consists only o f
visualization. During this process mantras are imposed upon the
lotus visualized in the practitioners heart in the same manner as they
are upon a mandala, and subsequently visualized and wor shi pp ed
there.4K Thus the lotus o f the heart serves as a mandala in mental
worship.
In external worship, a mandala is drawn as described above (pp.
122f.). Like the other materials used for worship, the mandala is
ritually purified before worship. According to the Jayakhya-Samhita
this ritual purification consists o f sprinkling water upon the mandala
and burning and inundating it mentally by means o f mantras.^ After
44 T h e vith i(kii) or v ith i is the passage surro u n d in g the pit.hn. T h e p ith a s u r r o u n d s
the inner lotus wheel; cf. N aradlya-S am hita 8.10cd, Apte 1973: 504 and the d ra w in g
ibid. Figure 1.
45 Cf. P a ram a-S am h ita 2 .8 3 -8 6 ,
4,1
F or other e x a m p le s o f m a n tra im p o sitio n s, see the d ra w in g s in A p te 1973:
513ff. a c c o r d i n g to J a y a k h y a - S a m h i t a 1 3 .1 0 7 125b o r in H ik ita 1990: 172
according to Satvata-S am hila 17.6 5 c72b.
47 T h e fo llo w in g description o f the use o f the m andala in the daily ritual relates to
the Jay a k h y a -S a m h ita . For a detailed description o f the daily ritual as g iv e n in this
text, see Rastelli 1999: 193-322.
411J a y a k h y a -S a m h ita 12.81 c102b; cf. also Rastelli 1999: 264f.
4'J H a v in g a n o in ted [his] hands w ith fragrancc, hav in g sp rin k led [water] that is
d ra w n up fro m the arg h ya vessel ov er the entire m a n d a la , pavilion [and] o ffe rin g
substances, he should cau se [them] to b urn by m e a n s o f the astram antrn [and] then
M AND ALA S A ND Y AN TRA S IN T H E P A N C A R A T R A
129
the purification, the mantras are imposed upon their places on the
mandala while imagining that they are arising from V is nu 50 and thus
m a d e present there. The n the mantras are mentally visualized in
anthropomorphic forms and worshipped by offering various things
such as iirghya,51 water used for washing the feet (padya), flowers,
fragrances, various kinds o f food, et c.52 This kind o f worship in
which the mantras are imposed upon different places on a mandala,
visualized in a concrete shape, and worshipped is called bhogayaga,
the offering o f objects o f enjoyment.53 Accordingly, the mandala is
called bhogasthana, the place o f the objects o f enjoyment.54
At the very end o f the daily ritual, i.e., after the fire ritual, the
mantras made present upon the mandala are sent forth again: The
practitioner offers arghya and fragrances again, takes the arghya and
flowers offered during the external worship from the mandala as his
prasada, puts them on his head, and draws a mar k ( tila ka ) on his
forehead with yellow and red powder from the mandala. Then the
practitioner visualizes that the mantras being present u p o n the
mandala enter the gross ( sthula) body o f the main mantra, i.e., that o f
V i s n u himself. The main m a n t r a s gross bo d y enters its subtle
( suksm a) body and the latter, its highest (para) body. This highest
body enters the heart-lotus o f the practitioner who then visualizes his
b o d y as shining and starts to tremble on account o f the m a n t r a s
in u n d ate [them] by m e a n s o f the m ula[m antra]. T h en the s u b sta n c e s are p u re and fit
for the o f f e r i n g , ( J a y a k h y a - S a m h i t a 1 3 .7 3 c7 5 b : g a n d h a d ig d h a u ka ra u k r tv a
arghyapatroddbrtena ca // 73 m a n d a la m m antapam p ro k sy a y a g a d ra v y a n y asesatab /
d u h a y e d astram antrcna m u lcn a p la v a y c t tatab // 74 n irm a lo d ra vya sa n g h a s ca y a g a y o g y o h h a v c t ta d a /) Cf. also J a y a k h y a - S a m h ita 13.89 fo r a re p e a te d ritual p u rifi
cation o f the m andala.
50 Cf. J a y a k h y a - S a m h ita 13.105c10 6 b : T h e n he sh o u ld v isu a liz e that L a k s m i
a n d the other [m an tras] arc c o m in g forth as sp le n d o u r fro m the v e n e r a b le V i s n u s
sh in in g b o d y like a m ass o f s p a rk s. ( tato b h agavato v isn o r bhasa b h a sva ra vig ra h a t
// la k sm y a d ir nissrta d h y a y c t sp h u lih g a n ica ya ya th a /)
51 A rg h y a is a m ix tu re o f v ary in g ingredients. A c c o r d in g to J a y a k h y a - S a m h i t a
1 3 ,6 5 c66b, it c o n s is ts o f w h ite m u s ta rd , s e s a m e seeds, d u r v a grass, w h i t e rice,
barley, water, m ilk, and fruits.
53 Ja y a k h y a -S a m h ita 13.106c178b.
53 J a y a k h y a - S a m h ita 1 2 .8 3 c - 8 4 b , 20.341 ab (= P a r a m e s v a r a - S a m h i t a 15.839ab).
F o r the use o f the term bhogayaga, see also P a u s S ] 19.59c and L a k s m i - T a n t r a
38,84a.
'
M S ee, e.g., J a y a k h y a - S a m h ita 13.106c an d L a k s m i- T a n tr a 38.2 8 c. T h e b h o g a
y a g a is differentiated fro m the layayaga, the o ffe rin g in d is s o lu tio n , in w h ic h the
m a n t r a s a re w o r s h i p p e d w ith o u t v is u a l i z i n g th e ir c o n c r c t e f o r m a n d w ith o u t
i m p o s i n g th e m u p o n a p a r tic u la r p la c e (see J a y a k h y a - S a m h i t a 12 .7 5 c 81 b an d
Rastelli 1 9 9 9 : 2 6 1 - 2 6 3 ) .
'
130
MARI ON RASTELLI
power. Next the food that was offered to the mantras (n a iv ed ya ) is

distributed among the persons who have participated in the ritual.
Visvaksena is invited into the mandala and the substances previously
offered to the main mantra (mulamantra), i.e., Visnu, are offered to
h i m .55 After sending forth the mantras also from the fire pit, where
they were w or shi ppe d during the fire ritual, and inviting and
worshipping Visvaksena also there, the devotee worships him again
in the mandala, asks his forgiveness for any mistakes that have been
made during the ritual, and sends him forth. The mandala is left after
having brought a vessel that is filled with milk, water, honey, and
clarified butter and in that the astramantra has been im pos ed .56 This
vessel probably serves as the man dal as safeguard, as protection is
one o f the functions o f the astramantra,57
In other rituals, the mandala is treated like any other place where
the deity is invited and worshipped. So, mandalas are used for the
d e i t y s w o r s h i p in different rituals such as festivals (utsava),
consecrations (pratistha) o f statues, temples, etc., or magic rites.58 In
the annual rite o f pavitraropana, which is performed in order to make
good all faults accumulated in the rituals during the past y e a r ,59 the
mandala is invested with threads as the other places where the deity
is present during worship are, such as the pot, the statue, and the fire
4 60
Pit.
The mandala is o f particular importance in the initiation (diksa).

Al though it is possible to perform the diksa ceremo ny without a
mandala if one does not have the means for it/ 1 using a m and ala
55 V i s v a k s e n a a lw a y s r e c e i v e s th e r e s id u e o f (lie o f f e r i n g s to V is n u . F o r
V isvaksena, cf. G upta 1976.
56 J a y a k h y a -S a m h ita 15.230231 and 2 3 3 c -2 4 9 .
57 Cf. its e m p l o y m e n t in the d ig b a n d h a rite in J a y a k h y a - S a m h i t a 9 . 2 1 c - 2 4 ,
11.6-7b, 16.91 c - 9 2 b .
'
58 C f., e.g ., P a ra m c v a r a - S a m h ita 17.89cd, ISvara-Sam hitS 1 0 .2 7 5 a b , P a d m a S a m h ita , caryapiida 1 1.2322 3 3 c (u tsiiva); J a y a k h y a - S a m h ita 2 0 .1 3 7 a , N a r a d ly a S a m h ita 1 5 .23cd, P a d m a - S a m h ita , k r iy a p a d a 31.41 (pratistha), P a d m a - S a m h i t a ,
caryapiida 32.21 Sab (fo r the p u rp o se o f e xpulsion [uccii/ii/ia]), and Rastelli 2 0 0 0 b :
325f.
'
w Cf. Ja y a k h y a -S a m h ita 21.1 4b, K ane 1977, v o lu m e 5: 339f. and G u p ta /H o e n s /
G o u d riaan 1979: 158.
W
) J a y a k h y a - S a m h ita 2 1 .6 3 64b, P a u s S | 3 0 .3 2 c 35b, TSvara-Samhita 1 4 .1 3 8 c 140b, 1 4 .2 1 5 c 2 i 6, N a r a d i y a - S a m h i l a 2 3 . 2 3 c - 2 6 , P a d m a - S a m h i t a , caryapiida
14.51 c - 5 3 and 19.93-94b.
'
'
61
Cf. J a y a k h y a -S a m h ita 16 .4 c6: For rich [people] lie should c a u sc to p e r f o r m
[the initiation] by m e a n s o f an offering [to the god] in a great m andala. F o r a perso n
w h o is dep riv ed o f p roperty an d w ealth [or] p o ssesses very little p ro p e rty [but] is
MAN DAL AS AND Y AN TRA S IN THE P A N C A R A T R A
131
during the initiation is often considered better62 and sometimes even

so essential that the term one who has seen a m a n d a la (m a n d a la d rsta ) serves as a designation o f one who has undergone the first
initiation and thus has become a sam ayin!'3
An examination o f the different Pancaratra-Samhitas shows that
th e role o f the mandala in the initation was not always the same. It
w a s used in diverse rites o f the diksa and so the meaning ascribed to
it also varied.
T h e P au sk ara -S am hi ta distinguishes four kinds o f ma ndalas
according to the rank o f initiation. The padmodaramandala is used
for the sam ayin, the anekakajagarbham andala for the putraka, the
cakrabjamandala for the sadhaka, and the m isracakram andala for the
a c a r y a M At the beginning o f the initiation, the teacher draws the
respective mandala, worships it, and then shows it to his disciple.65
His hand, upon which he has visualized and thus made the mandala
present, mandalas comprising all aspects o f the deity in the form o f
the mantras imposed on them, is subsequently laid on the disciple.
T h r o u g h this act, the teacher, who is identified with the deity
himself,66 liberates the disciple from transmigration: Having recog
nized that the devotee is helpless and plunged into the ocean o f
so rrow, he should draw [him] out by m ean s o f the yaga h and
a fra id o f tra n sm ig ra tio n [and] is tru ly a d e v o te e o f V is n u , the te a c h e r s h o u ld be
g r a c i o u s (i.e., p e rfo rm th e initiation fo r h im /h e r) o n ly w ith seed s, s e s a m e , an d
clarified b u tter [that are oblatcd to the god w h o is present] in the fire, [and] for [a
p erso n ] w h o is d ep riv ed o f any substanccs, [he should be gracious] only by m e a n s o f
a w o r d (i.e., a m a n t r a ) . ( m a h a m an d a la y a g c n a v itta d h y a n a m tu k a r a y e t 11 4
v itia y o g a v im u k ta s y a sva lp a vitta sya d e h in a h l sa m sa ra b h a ya b h ita sya visn u b h a k ta sy a
tattvatah H 5 a gn a u c a jy a n v ita ir bljnih sa tila ih k e v a la ih tatha / d ra v y a h m a sy a vai
k u r y a d va caivanugraham g u ru h 11 6) Cf. also L a k s m i-T a n tra 4 1 ,9 -1 Ob fo r a variant
o f this passag e. Possessing the n ecessary m aterial m e a n s is an essentia! p rerequisite
fo r o b ta in in g the initiation by m e a n s o f a m andala. Also acco rd in g to the PausS, the
first thing that the te a c h e r asks from a disciple w h o h as c o m e for th e initiation is to
a c q u ire the necessary m e a n s (P au sS ] 1.14).
62
Cf. B hiirg av a-T an tra 24,28: T h e principal in itation in the cakrabjam andala is
p ra ise d for the [person w h o is d evoted to the tw elve syllable m antra]. T h e s e c o n d a ry
alternative initiation is [perform ed] only n e a r the fire. ( ca krabjam andale diksa tasya
m u k h y a p ra k irtita l aim ka Ip a b h a v e d diksa ke v a la m vahn isa n n id h a u 11)
w See S a tv a ta -S a m h ita 2 0 .2 a (= ISvara-Sam hita 2 1 .4 6 4 c ) and A laîn g a B h a t t a s
c o m m e n ta ry on this p assag e (S a tv ata-S am h ita-B h asy a, p, 410, 9 - 1 1 ) .
MPausS j 1.8b11. F o r different p ro c e d u re s with regard to the u s e o f the m a n d a la
in the d ik sa for the different ranks, cf, also P a ra m a - S a m h ita 8 .7 c1 lb , V isn u -S arn hita 2 . 4 0 - 4 7 and 1 1.3-6.
65 Pau sS ] 1.6-8a.
wF o r the identification o f the teach er with the deity, cf. Rastelli 1999: 1 6 8 -1 7 0 .
132
MARION RASTELLI
(yagahasiena). It: is known that this teacher is similar to me (i.e.,

Visnu). 67 This procedure is called yfigadiksa, initiation into the
yaga (i.e., mandala). After it, the teacher destroys the disciples past,
present, and future bonds68 by sacrificing them into the fire, and
causes the disciple to become identical to the deity.m The diksa is
completed.
After the diksa, the devotees belonging to the various ranks o f
initiation vary in their worship o f the mandala. The sa m a y in only
looks at and worships the mandala on the twelfth tit hi o f every ha l f
(paksa) o f a lunar month. After one year he may become a putraka. A
putraka should look at, worship and visualize the mandala within his
heart for another year. Then he may become a sadhaka. Me should
worship and visualize the mandala for three months minus five days.
I f he is successful in this, he may undergo the initiation for an
acarya. After having worshipped and visualized the mandala for four
years, he may reach final emancipation.71 According to the PauskaraSamhita, emancipation can be attained only by means o f ma ndala
worship: The ocean o f wordly existence is not passable by any other
[means] than the y a g a 7'
In the diksa according to the Sanatkumara-Samhita, the mandala
plays an even more central role than in the Pauskara-Samhita. In this
text, one should undergo the initiation is expressed by the phrase
p ra vised diksam cakram andale, one should attain the initiation in
the cakramandala,72 which shows the centrality o f the mandala in
the diksa. The initiation ritual as described in the San atk um ar aSamhita is very simple. After constructing a mandala by means o f
flowers73 and distributing various offering substances on it,74 the
teacher worships it, takes the disciples right hand with his right
hand, and, carrying a handful o f flowers (jjuspahjali), leads him
f'7 P a u sS ] 1.28:
jn a tv a b h a kta m anatham ca n im a g im m sokasagarc I
u d dharcd yagahastcna sa g u m r m a ts a im s sm r ta h // 28
F or the visualization o f the m andala on the hand, see below , pp. I37f.
T h e s e arc often s y m b o liz e d by a thread; cf., e.g., J a y a k h y a - S a m h ita 1 6 .1 3 1 134b and 2 6 0 c - 2 7 4 .
'
m P a u sS ] 1.37d
41. T his passage is corrupt and nol intelligible in all details.
7,1P a u sS i 1.13-23.
71 PausS i 1,26ab: bhava rn a vo h y alahghyas tu vina s y a d y iig a to 'n ya ka ih /
72 San a tk u m a ra -S a m h ita , indranltra 4. led, 3ab, 4ab, 5ab, 6nb, ctc.
73 Cf. above, p, 123.
74 S a n a tk u m a ra -S a m h ita , rsiratra 5.1 30b.
MANDALAS AND YANTRAS IN THE PANCARATRA
133
a r o u n d the ma ndala while turning their right sides towar ds it

(pradaksina). Then the teacher has the disciple approach the mandala
an d worship it with offerings o f fragrances, flowers, etc. Following
this he teaches him the rules that must be observed and makes him
en ter the mandala. How far he may enter into the mandala depends
on his varna and his/her sex. A Brahmin m ay enter into the inner
m o s t enclosure (prathamavarana), a ksairiya into the second enclo
sure, and a vaiSya into the third enclosure. Sudras and wo m en must
stay outside the mandala and are not allowed to enter it. 75 In the
mandala, the disciple is told three mantras, a Vedic one ( vaidika), a
Ta ntr ic one ( tantrika), and one that is both Vedi c and Ta ntric
( vaidikatantrika). These mantras can be used in the various rituals
henceforward.76 With this the initiation is completed.
In other Samhitas the initiation ce re m o n ie s are m u ch more
elaborate. In the following description, only the rites concerning the
mandala will be discussed. A rite in the diksa that is described very
ofte n in the texts is the tossing o f one or m o re flowers, and
sometimes also other substances such as fragrances and arghya, onto
the mandala. Usually, the disciple is led blindfolded around the
man dal a and then given two handfuls (anjah) o f flowers and other
substances, or only one flower, which he must toss onto the mandala.
Then the blindfold is removed and the disciple may see the mandala.
The older Samhitas do not: clarify explicitly the meaning o f this
rite. The Ja yak hya-Samhita reads: Having prepared an anjali o f
flowers (puspanjaii) [for each o f the disciples] in sequence, he should
have [the disciples], whose eyes are blindfolded as before, toss [the
flowers] that are en do we d with gold, jewels, and pearls. T h e n he
sh ould unveil [their] eyes and show [them] e v e r y t h i n g . 77 Th e
Satvata-Samhita states: Taking the [disciple] by the hand, he should
go near the g o d s abode (i.e., the mandala). Having placed [him] on
his left side and again blindfolded [his] eyes, he should have [him],
w h o is deprived o f [his] eyesight [now], toss a r g h y a [from] the
aiijaii. [Then the disciple] may see the mantras highest abode, which
75
A lso the kind o f m a n d ala varies w ith regard to th e d i s c ip le s va m a o r sex; cf.
S a n a tk u m a ra -S a n ih ila , rsiratra 5 .4 0 c -4 3 .
7f>S an a tk u m a ra -S a m h ila , rsiratra 5.30c-4Qb.
77 Ja y a k h y a -S a m h ita 16,217 c - 2 18:
k r a im tp u s p a n ja lim krtva b a d d h a n ctra m s ca p u r v a v a t I I 217
k sc p a y e c ca h ira n ya d h ya m m a n im u k ta p h a la n v ita m /
u d ghatya n a ya n e pascfid a kh ilu m sam pradarSayct 1/ 218
134
MARION RASTELLI
bestows the fruit o f [the fulfilment o f every] wish.71' The reason for
tossing flowers and other offerings onto the mandala may be that the
disciple, prior to seeing the mandala, must worship the deity present
in it in order to dispose him favourably, and that only then does he
grant him his favour. This can be concluded from the SatvataSam hitas statement that the mandala that has been looked upon by
the disciple, fulfills every wish.
A passage in the ISvara-Samhita also shows that the tossing o f
flowers, etc., brings the disciple into contact with the deity present in
the mandala. This contact causes or increases his devotion ( b h a k ti)
and purifies his internal organ (antahkarana) as a result o f the deit ys
grace: Me should have him, with a bowed head, toss the a njahY s
contents] into the g o d s abode. If he sees that [the disciple], whose
eyes have [then] been unveiled, who has not been seen by any other
person [and] who has entered the auspicious path, [shows] the sign
o f devotion ( bhaktilaksana), which involves bristling o f the hair,
eagerness, joy, tears o f bliss, prostrations, exclamations o f vi ct or y
(jaya), and circum amb ulatio ns in [all] quarters (Idikpradaksina),
[and] if he has understood that [the di sciples] internal organ is
purified, he should recognize that he is suitable. When [all this has
happened], then [the disciple] has received grace (anugrhlta) through
this (i.e., the t e a c h e r s) thought, which is named A cy ut a (i.e.,
Visnu). 7<) A passage from the Tantraloka, although from the K a s h
mirian Trika tradition, may help further to understand the idea that is
behind the described ritual. According to this text, when the blin d
fold is removed the mantras present in the mandala enter the disciple
78 Satvata-Sam hitii 19.39c

41 b:
tarn adaya karad d cva d h a m a sa n n ika ta m vrajct ff 39
krtv a tm a n o vam a b h a g c bhuyah snm cchiidya lo c a n c /
p ra k sc p a y c t tatbfi sarghyam anjalim m u k ta lo c a n a m ll 40
s a m p a sy c t param am dham a m antram icchaphalapradam (v .\.) /
w I S v a r a - S a r n h i t a 2 1 . 1 3 1 - 1 34b:
p ra k sc p a y c d d c v a d h fm m i n a tam urdhnam jai in ca Iam /
tasyodghatitanctrasya tv adrstasyciarair ja n a ih // 131
ku sa ln d h v a n iv ista sy a drstvii vai b haktiluksanam /
ro m a n ca u tsu kya h a rsa d h ya m anandagrusam anvitam II 132
sa pranam ajayalapadikpradaksinasam yutam /
p u tantahkaranam bu<d> dhva y o g y o y a m iti b h a v a y c t/l 133
ya d ii ta d a cyu ta kh ya y a m ig rlu to d h iya lay;1 /
T h ro u g h the identification o f the tc a c h c r s thought with the deity, the identity o f the
tcacher with the deity is also e m p h asized here.
M A N D AL A S AND Y AN TRA S IN THE P A N C A R A T R A
135
in an instantaneous possession (avesa).80 Perhaps the author o f the

I sv ar a- S am hi ta did not consider the signs o f d e v o t i o n 5 (bhaktilaksana) as possession, but it is obvious that also according to the
Isvara-Samhita, the disciple, when seeing the mandala for the first
time, comes into contact with the deitys power.
In many Samhitas, the tossing o f a flower onto the mandala is
considered a means to determine the initiands mantra or name. As
de sc rib ed above (pp. 1 2 7 f), varying mantras are placed on the
different constituent parts o f the mandala. The mantra that is present
on the place where the flower tossed by the disciple falls is then his
p e r s o n a l mantra or gives him his name. The P a r a m a -S a m h i t a
desc ribe s the first variant: He should blindfold the disciple by
mea ns o f a cloth, take [his] hand and enter the mandala through its
door. He should lead [the disciple around the mandala,] turning their
right sides towards [it] and have [the disciple] bow down before the
guardians o f the quarters. On which flowed' this flower falls by
chance, this [flowers] mantra is for his protection and prosperity.
Hav in g effected [his] faculty o f seeing [again], he should show him
the cakramandala. K2 Several y oun ger Samhitas indicate that the
purpose o f tossing flowers onto the mandala while blindfolded is to
give the disciple a name (namakarana), A m o n g these texts is the
Padma-Samhita, which states: Directed by the teacher, [the disciple]
sh oul d toss the flowers onto the mandala. [The teacher] should
assign to the disciple the names o f the manifestations (m u rti) such as
Kesava, etc., that are the sovereigns o f that place upon which most of
th e flowers fall o f their ow n accord. The w o r d bhagavata or
bhattaraka is to be joined to the end of the name i f [the disciple] is a
m l A 1 5 .4 5 1 c ^ t5 2 b . Cf, also Sanderson 1986: 169 and T a k a s h im a 1992: 5 If.
Sl
O b v io u s ly , the m a n tr a s arc in v o k e d in flo w ers b e in g p la c e d on the v a rio u s
p a rts o f the m andala. B iih n e m a n n 1987: 47 d escrib es a sim ilar m e th o d u sin g areca
n u ts for invoking m antras. T h is p assag e p r e s u m a b ly does not m ean that the m a n d a la
is m a d e o f flo w e rs (cf. p. 123), b e c a u s e th e v e r s e s P a r a m a - S a m h i t a 8 . 7 e - 9 b
p re s c rib e that the m a n d ala is to be draw n w ith paint in the case o f a sa m a y a d ik sa and
w i t h p o w d e r d u r i n g the ta n tra d ik sa ( P a r a m a - S a m h i t a 8 .9 b tatra d ik s ifa h em .
tantradiksitah).
10 P a r a m a -S a m h ita 8.4446:
vastram cka m upadaya sisy a m ba<d>dhva m u k h a n ta r e l
h a ste g rh itva p in v isc t svena dvfirena m a n d a la m /l 44
n itv a p ra d a ksin a m caiva d ikpalan p ra n ip a ta yet /
ta t p u sp a m n ip a tc t p u s p c y a s m in n cva ya d rcch a ya 1145
sa m a n tra s tasya raksayam u d a yc ca v id h iy a te /
utpadya ca ksu si cainam darsaycc ca kra m a n d a la tn l/ 46
136
MARION RASTELLI
Brahmin. He should assign a name ending with dcva to a ksatriya,

one ending with the word pain to a vnitiya, [and a name] ending with
dasa to one who is born from [the Purusas] f e e t Then he should
remove the blindfold and show [him] the cakranmndaki.M
Although most o f the Samhitas examined connect the ritual o f
tossing flowers upon the mandala with the name-giving, this was not
its original purpose. This can be seen in the .layakhya-Samhita that
prescribes the name-giving ceremony as being at another moment o f
the initation than that o f the first showing o f the mandala, and
prescribes the ceremony being done with the help o f a platform made
o f earth (sthala) upon which a mandala is only i m agi ned .85 In the
Satvata-Samhita, although the name-giving immediately follows the
tossing o f the puspanjali upon the mandala, the purpose o f the
tossing act is not to give a name. It is just done on the same occasion.
The passage Satvata-Samhita 19.39d-41b quoted above (pp. 133f ,)
continues as follows: On this occasion, he should give [him] a
10 I.e., a sudra; cf. Rg-Vccla 1 0 .90.12d: padb h ya m iu d ro ajayata. For the choicc o f
the n a m e s for the m em b ers o f the diHerein varnas, cf. also M a n u -S m rti 2.31 -3 2 . I
am thankful to Professor Biihnemann for this reference.
1,1
P a d m a -S a m h ita , caryiipada 2 .5 7 -6 0 :
p u sp lin i vikirct tasm in m andate g uruenditah /
y a s m in p a ta ti bhubhagc p uspanam utkaras s v a y a m /l 57
ta d bhagadhipam uriinam niim iiny asya vinirdiSct /
sisya sya k c h v a d i n i padam bhagaviitcti ca / / 5 8
n a m n a m ante p ra y o k ta v y a m y a d va bhattarakcti ca /
b riih m a m s cot ksatriyasya d cva n ta m nfima nirditicl II 59
va isya sya pala& ibdantam dasanlam padajunm anah /
tato vim u c y a drgbandham darsaycc cakram andalam II 60
See also V is n u - S a m h ita 10 . 6 1-64b, SripraSna-Samhilfi 16.106 a n d 1 3 6 - 1 3 7 , and
ViSvamitra-Samhitii 9 .6 5 -6 7 .
85 Cf. J a y a k h y a -S a m h ita 16.124c127b: l i e should blind fold the d i s c i p l e s ey es
by m e a n s o f a n ew , pure, w hite cloth upon w h ich the nciram antra h as b een used.
T h en he sh o u ld have [him] cast for V isnu a pu sp a n ja li that is not in v ested w ith a
m a n tra upon the platfo rm . H e sh o u ld bran d him w ith the n a m e o f the m e n ta lly
im agined lotus upon w hich the [puspanjali] falls before [this] he sh o u ld d ivide [the
platform ] into nine parts (here obviously the n a va p a d m a m a n d a la co n sistin g o f nine
lotuses is m eant [cf, note 29]) , to gether with the w ords visnu or pati, o B r a h m in .
{vasasa b y a h a tcn a th a su d d h c n a su sitc n a ca H 1 2 4 n ctra m a n tra b h ija p tcn a n c trc
iis y a s y a b a n d h a y c tl am antram k sc p a y c c ciito visn o h p n sp a n ja lim sth a lc 1/ 125 yatra
sii p a ta ti brahm an b u d d h isa n ka lp itc m b u jc / navadha vibhajel pra g va t ta n m lm m a m
tam a iik a y c t H 126 visn u sa b d fm vitcn a iva p a tisa m jfu lyu tcn a [cm. o f the editio n ] ca /
T h e text is corrupt.) It is u n k n o w n if th e branding o f the n am e is to b e taken literally.
U s u a lly in the diksa, a cakra m ark is b ran d ed on the right s h o u ld e r and a ia iikh a
mark on the left one (cf. Satvata-S am hita 2 2 .9d, ISvnra-Samhita 2 1 .2 8 4 c - 2 9 2 b , and
R angachari 1930: 35).
M AN D AL A S A ND Y AN TRA S IN T HE P A N C A R A T R A
137
s u ita b le name o f a [deity] who has a secret primary name and a

s e c o n d a ry one accordant with circumstances. 86 Ac co rd in g to the
Isvara-Samhita, the disciples name is determined with the help o f
tossing a puspanjali upon the mandala. This is described by means of
t h e passag e from the Satvata-Samhita quoted above, w h i c h the
I s v a r a - S a m h i t a lias a d o p t e d from this text like m a n y other
p a s s a g e s ,87 However, the name-giving rite is performed one day after
t h e first vi ewi ng o f the m a n d a l a ,88 w hi ch occurs d u r in g the
prep ara ti ons for the initiation ( d lk sa d h iv a s a n a ). Thus, the rite o f
na m e- g iv in g originally was not connected with the act o f tossing
flowers onto the mandala.
At the end o f this section on the diksa, I w ould again like to
c on s id e r the placing o f the teachers hand on the disciple that was
m e n t i o n e d above (pp. 13If.) as being part o f the initiation. This
placing o f the teachers hand can be done either before revealing the
m an d al a to the disciple or after,80 The hand is mentally invested with
the de it ys power that is then transferred to the disciple. Investing
w i t h the d eit ys pow er is done either by imposing a mantra90 or
visualizing a mandala on the hand: On [his] right hand, he should
visualize all principal deities as illuminating [it] by their rays o f light
[and] being in their respective supreme abodes in the centre o f the
cakrabja[mandala]. He should [then] touch the [disciple] with this
% S a tv ata-S am h ita 1 9 .41c-42b:
ta sm in n avasarc kuryiin nam a ya sy a y a ih o c ita m II
rahasyasam jnatp m u k h y a m ca g a u n a m vasya y a th a sth ita m /
1(7 Is v a r a - S a m h ita 2 1 .3 1 9 - 3 2 5 ~ S a tv a ta - S a m h ita 1 9 .3 9 c - 4 6 b . T h e I s v a r a - S a m h ita in tro d u ces this p a s s a g e s with th e fo llo w in g words: He sh o u ld g iv e a n a m e to
th e disciplc. T h e rule for this is given here, (Isv a ra -S a m h ita 2 1 .3 1 8 c d : n a m a k u ry a c
ca sisya syn -tn d vid h a n n m ih o c y a tc //)
"HCf. TSvara-Samhita 2 1 , 1 3 1 - 1 34b q uoted on p. 134f.
m B e fo re re v e a lin g th e m an d a la : S a tv a ta - S a m h ita 19 .3 8 c 39 b , V is n u - S a m h i t a
1 0 . 4 6 c - 5 2 b ; a fte r it: P a u s S | 1.28, J a y a k h y a - S a m h i t a 16,335, P a d m a - S a m h i t a ,
caryapiida 2 .6 6 - 6 7 b , Isvara-S am hita 2 1 .1 34c136b, V i^ v a m itra -S a m h ita 9 .7 5 c - 7 6 a .
Cf. J a y a k h y a - S a m h ita 18 .8 2 c83: H a v in g thus s p o k e n , the te a c h e r h i m s e l f
s h o u ld visualize the su p re m e V isn u w h o is the su p re m e one [and] w h o c o n sists o f
p arts and is w ith o u t parts (cf. Rastelli 1999: 1 0 1 -105), in the form o f the m an tra on
[his] rig h t h a n d , w o r s h ip h im w ith fra g ra n c e s , flo w e rs , etc., a n d lay it on the
[d is c ip le s ] h e a d . ( it y u k tv il d a k s in c h a s te s v a y a m s a n c in ty a v a i g u r u h / / 82
m a n tra tin a n a m param v isn u m pa ra m sa k a la n isk a la m / sa m p u jy a g a n d h a p u sp a d y a ir
d a d ya t tasya ca m a sta k c H 83) and S a tv a ta -S a m h ita 19 .3 8 c 39b: H e sh o u ld to uch
[h im ] fro m the feet w ith the m a n tra h a n d and th en he, w h o c o n q u e r s th e seed o f
s u f f e r in g , s h o u ld lay the m an tra hand, w h ic h h as a b la z in g fo rm , on [his] h e a d ,
(ap a d a n m antrahastcna param rsyatha m u rd h a n i // m antrahastam jv a la d r u p a m da d ya d
y o d u h k h a b ljn jit I)
138
MARION RASTELLI
Acyula hand that was [previously] wetted with water.')l If a mantra

is imposed upon the hand, it is called mantra hand (mantrahasta). If a
m andala is visualized on the hand, it is called yaga hand (i.e.,
mandala hand; yagahasta)!2 In both cases it can also be called Visnu
hand (visnuhasta) (or Acyuta hand as in the passage just quoted) as
in both cases Visnu is present on the hand: The hand on which
Visnu is present is called Visnu hand.w According lo the VisnuSamhita it destroys all sins that were ever accumulated: All sins that
were formerly accumulated in thousands of other births are dissolved
without doubt by the mere touch.'M According to the PauskaraSamhita it liberates one from the world o f transmigration (cf. pp.
131 f).
All the examples o f various Samhitas given show that the mandala
plays a central role during the initiation, although its meaning and
function vary in the different texts. First o f all, the first viewing and
first worship o f the mandala is the initiation into mandala worship,
which is in the centre o f the d e v o t e e s religious practice in the
P aus kara-Samhita especially but also in other Samhitas. But the
m an da la as a place o f the d e i t y s presence is also a means for
encountering the deity: by worshipping him there in order to dispose
him favourably, but also by experiencing his favour by the mere
sight o f the mandala, by obtaining a name and a mantra that were not
chosen by a human being, but by the powers present in the mandala,
and last but not least, by the physical contact with him through the
Visnu hand.
W
l TSvara-Samhitii 2 1 .1 34c136b:
sm a rcd daksinapanau tu ca kra m b u n ih a n ia d h ya g n m // 1 3 4
p radhanadcvatabrndam s v e s v e d h a m n i pare s th ita m l
svam aricigancnaisa d yo ta ya n ta m tu c;lkhi!am H 1 3 5
tenacyu ta ka rcim iva sodakcnalabhcta tarn /
Cf. also P a u sS ] 1.28 (q u o ted on pp. 13 I f.), P a d m a -S a m h ita , caryapada 2 . 6 6 - 6 7 b ,
V isn u -S a m h ita 1 0.46c-52b, ViSvamitra-Sarnhilii 9,75c~76a.
Cf. Satvata-S am hita 19.38c39b (quoted in note 90) and P au sS ] 1,28.
1,1
V isn u -S a m h ita 10.50cd; haste visn u h sth ito y a s m in visn u h a sta s til sa sm rta h //
F or th e use o f the term visnuhasta , cf. also J a y a k h y a - S a m h ita 16.335a, P a d m a Sam hita, caryapada 2.67a, and V iivfim itra-Sam hita 9.75d.
V isn u -S a m h ita 10,51:
janm antarasahasrais tu y a t papam sancitam pura /
sarvani tat spariam iitrcna vilayam y a ty asam Sayam II 5 1
MA ND AL A S AND Y ANTRAS IN THE P AN C A R A T R A
139
The S ym b o lic M eaning o f the M andala

As mentioned several limes above, the mandala is a place where the
deity and his various aspects are made present by imposing their
mantras upon it. Invested with mantras, the mandala is very powerful
and merely looking at it has an effect.
However, the Pahcaratra Samhitas not only assign a meaning to
the mandala but also to its constituent parts. As a place o f the deitys
presence, the mandala is considered to be his body and thus, the
m a n d a l a s constituent parts are considered to be the constituents o f
the body. Since the b o d y s constituents, i.e., the principles ( tattva)
arising from the primary matter (prakrti), also constitute the universe,
the mandala is also a representation o f the universe.
Such a conception is found in the Satvata-Samhita, which is also
adopted by the Isvara-Samhita: Having thus constructed [the m a n
dala], he should consider [it] as having the nature o f the universe like
a body. K no w that the white, yellow, etc., powders are the [five]
elements ( bhuta). Th e upasobhas96 are the subtle eleme nt s ( fanm atra), the so b h a sn the organs o f action (karana), and in the same
way, all the corners and doors are the sense organs. T h e outer
enclosures are the three [constituents o f the primary matter] sattva,
etc. The wide vith P is the mind (m anas); the pit ha" is called conceit
(garva = ahamkara). The intellect ( dhi = buddhi) is the lotus. Its ruler
is the p u ru sa who consists o f pure thought in the form o f a seed
syllable ( bijatm an), and the Lord (isvara), who is characterized by
bliss [and] whose mere sight pleases the mind forever, is present
there without a concrete fo rm . 100
'J> Cf. P a u s S ] 10 .3 4 d 35b: 1 e x p lain to y o u [now ] the entire [g reat m a n d a la ],
t h r o u g h w h ic h , i f it is m e re ly lo o k e d at, th e fe tte rs o f w o r l d l y e x i s t e n c e are
d e s tro y e d . (ta in ca k rtsn a m v a d a m i t e // y c n a sandrstam atrena b h a v a b a n d h a ksa yo
b h a v c t /) Cf. further V is n u -S a n ih ita 10.64cd: I f a m a n d a la is m e re ly lo o k e d at, the
ac c u m u la ted sins arc destro y ed . (drstam atrc p ranasyanti m a n d a te p a pasancayah //)
% A p te 1973: 504 and 1987: 131 e x p la in s u p a io b h iB as re - e n tr a n ts or the
in v e rte d c o u n te rp a rts o f the o ffse ts o c c u p y in g the sp a c e in b e t w e e n o ffs e ts and
c o rn e rs o f the e n c l o s u r e s . F o r a b e tte r u n d e rs ta n d in g o f this e x p la n a tio n , cf. the
d raw in g in Hikita 1991: 319.
T h e d o h h a is the sp ace b e tw e e n the u p a so b h a s; cf. the d ra w in g in H ik ita 1991:
3 1 9 (here the fem inine fo rm s sobhii and upasobha are used).
m Cf. note 44.
T h e pitha is the part o f the m a n d a la w h ich su rro u n d s th e lotus(es); cf. F ig u re 1
in Apte 1973 and also N aradiya-S arphita 8.10c.
Satvata-S am hita 1 l,32c-36:
140
MARION RASTELLI
A similar conception can be found in the Visnu-Samhita. In this

text, the mandalas constituent parts arc seen as parts o f the body on
one hand, and as cosmic and divine powers on the other. Possibly,
two texts that were originally different arc joined here as the two
conceptions are com bi ned without a real inner co n ncc ti on . In
addition, some principles ( tnltvn) and deities appear twice: He
should consider the mandala as a human being (purusn). The lotus is
its heart. [The m and ala s] centre is between the arms. The stalk o f
the lotus is at the base o f the navel. The two back doors in the south
and the north are to be known as the two feet. The stalk o f the lotus,
which has nine holes, is the seat o f the kulns Vimala, etc.1*" Its root is
the subtle Janardana in the form of a seed syllable ( bijarupa), since
one should know that the root o f the lotus, w hi ch s u pp o rt s
everything, is based on his greatness. The phonems a, etc. , 102 became
the winds [of the body]. On the stalk arc the finger-nails.IM The knot
('Igranthi) is the great o n e (i.e., the intellect [huddhi]) consisting o f
the [three] constituents [o f the primary matter]. The eight petals are
then the various [modifications of] the intellect, dharma, adhanna,
etc. Within the stalk is the endless ego principle that carries the
subtle elements, the sense organs, and the elements. The lotus [of the
g o ds throne (asana) visualized upon the mandala] is the unevolved
[primary matter]. Some [teach] that the circlcs o f sun, moon, and
firel(M are the different phonems a, etc.,105 and others consider them
krtv a iv a m am isandhaya sarvatm atvena dclm vat // 32
rajiim si vid d h i bhutani silapitadikani ca /
lanm iitrany u p a io b h a n i sobhani karanani tu // 33
e v a m sarva n i k o m n i sa d vaninindriyani ca /
bahiravaranam y a d vai sa ttva d ya m iritayam hi y a t 1/ 34
m anah suvitata vith i garvah p it h a w u d a h rta m l
dhih p a d m a m tadadhisthata bijaim a cin m a ya h p um iin I I 35
am urta isvanid ciitra tisth u ty anandalaksanah /
y a sy a sandarsanad cva Sasvad hhavah prasidati I I 36
(~ ISvara-Sam hita 11.161 c - 165).
101 T h e s e arc V i s n u s nine sa k tis, viz., V im ala, Utkarsini, J u an a, K riya, Y o g a,
PrahvT, Satya, Ia, and A n u g rah a (V isnu-S am hita 6 .4 4 c -4 5 ).
1(12 Here, p h o n e m s im posed upon the m andala arc probably meant.
118 T h e finger-nails are identified with the thorns on the stalk, both o f w h ic h are
called kantaka in Sanskrit.
IIH T h e s e also b e lo n g to the throne (asana) that is v isu a liz e d upon the m an d ala.
F o r the m ental visualization o f an ;isana upon the m an d ala, cf. also A h irb u d h n y a S am h ita 2 8 .1 7c18. For the constituent parts o f the asana, cf. Rastelli 2002.
1115 T h e p h o n e m s arc c o n s id e re d as p arts o f the w o rld ly c re a tio n ; cf. V is n u Sam hita 9,53: [The universe] is to be know n as being pervaded by the first sound in
M ANDAL AS A ND Y AN TRA S IN THE P A N C A R A T R A
141
as born o f the eight saktis Viniala, etc. The lotus that is the great
support o f the universe is in the centre o f the egg (? a n d a )m The
Maya is in the egg-shell below the egg, Vidya is above it.107 He
should consider the border ( prativarana) as the pericardium o f the
heart-lotus. Out o f the [border], god Vairaja (i.e., Brahma) is taught
as [being present] in the shape o f the rampart. In the yellow, white,
red, black, and dark powders are the [five elements of] the earth, etc.
The Vasus are on the tips o f the lotus petals; the Rudras and Adityas
are the filaments, the Maruts are on the junctures o f the petals; the
planets and stars are the powders. On the lines o f the petals upper
[edges] and between [them] are the rivers and oceans. [Mount] M e m
is in the pericarp, the munis are on the seed syllables. Visnu is in sun,
moon and fire. (...) The thorns are Yaksa, etc., the hairs on the stalk
are known as the Apsaras. The thread is Prajapati; the roots o f the
leaves are the winds [of the body] such as prana, etc. The atm an,
who is the lord o f the universe, who is Hari, who is known as having
no parts ( niskala), [and] who abides in the s up re m e abode, is
[present] in the centre o f the caki'amandala. He w h o has thus
recognized that the god is present in the mandala leaves V i s n u s
Maya behind and attains the supreme abode. He who worships or
visualizes (pasyet) the god as present in the mandala, even if it is not
prescribed directly, beholds the lord o f the gods forever. In the
mandala, the one who has all forms is eternally near here [in this
world]. Therefore, worship in the mandala is better than [worship in]
auspicious places such as tlrthas, etc. 1 In both passages quoted
the shape o f the nfida. V i s n u s s u p re m e abode is beyond the creation o f the p honem s,
e t c . ( nadarupcna v ijn c y n in v y lip ta m a d ya ksa rcn a tu / va rn a d ika lp a n a tita m tad
visnoh p aram am p a d a m ll) and Rastelli 1999: 125f.
Here, the lotus arising fro m V i s n u s navel on w h ich B ra h m a sits and creates
the w orld, is possibly meant.
1,17 M a y a an d V id y a are co n stitu e n ts o f the universe; cf, V is n u - S a m h ita 3.4 8 cd
and 6,42ab.
II#I V isn u -S am h ita 9 .5 8 c -7 6 b :
partisan! m andalam vid ya t p a d m a m hrdayam asya tu I I 5%
bahvantaragatam m a d h y a m n a b h im u lc ' bjanalakam /
p a d c d v c p a scim a d va rc jn a ta v y e d a ksin o tta rc I I 59
pa d m a n a la m navaccbidram vim aladikalasrayaip /
tasya m u la m bhavct s u k s m o bijarupo janardanah // 60
iidlm rabhutani sarvasya m a h in m i s v e vya va stb ita m /
p a d m a m u la m v ija n iya d y a to k nr ad a y o bhavan I 61
m arutah kantaka nale g ra n th ir g u n a m a y o m ahan /
buddh ib b ed a dala n y astau dh a n n a d h a rm a d a ya s tatah // 62
nalantc b a m k rto 'n antas tanm atrendriyabhutabbrt7
142
MARION RASTliLU
here, the mandala is seen as the body o f the deity. Like any other
body, it consists o f the tnttvus, of limbs, and of organs. However, the
mandala also consists o f principles, and o f cosmic and divine powers
that constitute the universe. Thus il is also a representation o f the
universe. Again, the universe is a manifestation o f the deity.1"0 Thus,
the mandala, even if it is not yet invested with mantras, is considered
here as being a representation o f the deity."0
Ynntrns
Like mandalas, yantras consist usually o f diagram-like drawings and
mantr as made present in them. The re are, how eve r, es sential
differences between mandalas and yantras.
Yantras are generally drawn on mobile materials. For drawing and
writing, powders are not used, but rather liquid substances; metal
p a d m o ' vya kta m akaradibhedn ' rkcn d va g n im a n d a la m 1/ 63

buddhigarvam anom iitra daksinadidalastakam /
it y a n yc vim alaclyas(;isaktijam ciiparc vidul) // 64
a n d a m adhyagatam p a d m a m vttva syiiya ta n a m im lrn t/
m aya n d a d h a h kapalastha vidya cordhvam vyavasthita // 65
p u n ta ta m hrdabjasya ka lpayct prativiiranam I
tadbahyatah purakaro vairajo (leva ucyutc II 66
p lta c c h a ru n a kfsn csu dyiim c ca ksm a d a ya h sthituh /
va sa vo ' bjadalagrastha rtidradily;7,s! ca kcsarah II 67
m aruto dalasandhistha rcnavo grahatarakah /
dakintaragrarckhasu saritah sag: was tat ha II68
m c ru s tu karnikantastlio bijesu m un a ya h sthitah /
su ryen d va g n ig a to v isn u s(...)
y a k sa d y iih kanlaka nalc rn m a n y apsarasah sm rtah 111 1
sutram prajapatih p atranm lc pranadivayavah /
cakram andalam adhyastha alm a sarvcsvaro harih II 12
n iska la s tu sam akhyatah para m e v y o m n i sam sthitah I
c v a m y o vctti d cv a m tain m andalantargatam tatlul // 73
sa hitva va isn a vim m a ya m apn u ya t param am p a d a m l
a vid h a n c 'p i y o d cva n i m a n dalSntahsthitam y a jc t II 74
p a s y e d va tcna dcvcsa h sa ksa d drsto b h a ved d h n iv a m /
m a n d a te fra bhavcn n ity a m sa n n id h ya m sa rv a ru p im h 1115
tasm at tirth a d ip u n yeb h yo visista m m andate 'rca n a m /
V isn u -S a m h ita 9 .6 9 d 7 1b is an insertion that docs not m ake sense with reg ard to the
contents o f the passage.
l,wAll constituent parts o f the universe are m anifestations o f the deity (cf. Rastelli
1999: 98f.); thus the universe is also his representation.
1111 T h e S a m h ita s teach sim ila r n o tio n s with regard to the te m p le ; cf. R astelli
2003.
MA ND AL A S AND Y AN TR AS IN T H E P A N C A R A T R A
143
yantras can also be engr ave d. Th u s ya nt ra s can be carried

everywhere and also worn as amulets,
When mandalas are prepared, a diagram is drawn and only then
are the mantras made present upon it by imposition (nyasa). When
preparing a yantra, drawing and writing of the mantras are done in a
single process, and, as just mentioned, the mantras are written. This
is not the case with mandalas.
This is probably connected with the fact that in most cases yantras
are not considered to be ju st places where the deity can be made
present and worshipped as mandalas are, but are considered to be
representations of the deity himself (see below pp. 144f.).
Yantras are used mainly for worldly purposes (see be lo w pp.
146ff.).m
The writing materials most commonly used for the preparation o f
yantras are birch-bark ( bhurja) and cloth ( vastra, ka rp a ta )U2 Other
materials mentioned in the texts are gold (sometimes embellished
with jewels, corals, and pearls), silver, copp er and other metals,
wood, and s t o n e . 113 Th e writing paint is prepared from rocana,UA
saffron (ku n ku m a ), sandalwood (candana), talc (ghana), ca m p h or
( tusara), musk (kasturika), milk, agaru115 and d e w . 116 Th e writing
utensil is a golden needle (h e m a su c i).in The yantras described in
An ir uddha-Samhita 5 seem to be special cases. According to the
prescription o f this text, the yantras should be prepared and w o r
shipped on a platform (vedi) in a pavilion (m andapa)."s These yan
tras are, o f course, not mobile.
111 T h e s e d ifferences b e tw e e n m a n d a la s and yantras have e m e rg e d fro m m y study

o f the P a n c aratra-S am h itas. F o r a list o f differen ces that v aries slig h tly fro m m ine,
cf. B ru n n e r 1986: 19 (cf. B ru n n er, pp. 1 6 2 -1 6 3 ). T h e d ifferen ces n o te d b y B ru n n e r
m a y also apply to Pancaratra, but 1 have not yet fo u n d ev id en ce fo r this in the texts.
112 J a y a k h y a -S a m h ita 2 6 .97b, 2 9 .9 6 b , 164a, P a d m a - S a m h ita , caryapada 2 5 .1 0 2 c
32.46b.
113 P a d m a -S a m h ita , caryapada 32.92c, A h ir b u d h n y a -S a m h ita 2 6 ,3ab, 74cd, 85cd.

114 A c c o rd in g to A p te 1957 ( s .v . g o ro c a n a ), rocana is a b rig h t y e llo w p i g m e n t
p re p a re d from the urine o r bile o f a cow, o r f ound in the head o f a c o w .
115 A q u ila r ia a g a llo ch a . P e r f u m e s , o i n t m e n t s , a n d oil a r e o b t a i n e d f r o m its
fragrant w o o d (S y e d 1990: 31).
116 J a y a k h y a - S a m h i t a 2 6 .8 9 c , 97a, 106ab, 2 9 , 9 5 c - 9 6 a , 163cd, A h i r b u d h n y a S a m h ita 26,4ab, 7Sd, P a d m a -S a m h ita , caryapada 2 5 . 102d103, 3 2 .4 5 - 4 6 a .
117A h irb u d h n y a -S a m h ita 26.4c, P a d m a -S a m h ita , caryapada 2 5 .1 03c, 32.45a.
118 See A n iru d d h a -S a m h ita 5.1 c3 and 59.
144
MARION RASTi'LLI
According to the Ahirbudhnya-Samhita, the ya nt r as material is

determined by the qualification (adhikara) o f its user.11'1 In addition,
different materials lead to different results following the y a n t r a s
worship. For example, a yantra made o f gold, jewels, corals, and
pearls bestows sovereignty, or a yantra made o f bi rc h-bark and
drawn with saffron and sandalwood, when worn on the head, effects
the fulfilment o f all wishes.IJI
As already mentioned, when a yantra is prepared, dr awing the
diagram and writing (he mantras are done in a single process. Th e
following is an exemplary prescription for the preparation o f a yantra
that is used for paralyzing (stambhnna) divine beings: Furnished
with rocana and saffron, he should write [the divine b e i n g s] na m e
intertwined121 with the [m ukijim inlrn in the centre o f a [drawn]
kaustubha that contains sixteen sixteenth parts (knla). Previously
each sixteenth part has been provided with nectar,122 o sage. Above
the sixteenth parts o f nectar he should place [the mulamantra], which
resembles the m o o n s rays, resting on the viSvapyaya (= la). Outside
o f the [mulaimmtra], he should draw an eight-petalled lotus with a
pericarp. [Then] he should write the g o d s ;i/ig;i[jmi/7i'ras] on the
petals just as in worship. 12'1
The drawing and writing o f the yantra are seen as a unit. The
drawing is not just a place for making the deity present by means o f
mantras, but the y a n t r a s draw ing and writing as a unit is a
representation o f the deity, The deity assumes the form o f the yantra.
IWA h irb u d h n y a -S a m h ita 26.3.
A h irb u d h n y a -S a m h ita 26.74c~77.
121 T h e r e arc se v e ra l in te r p r e ta tio n s o f w hat vidarbha o r v id a rb b ita m e a n s .
G e n e ra lly , it m e a n s that parts o f the nam e and o f the m antra, i.e., one or tw o s y lla
bles o f c a c h , are alternated. Cf. Padoux 1977 and 1 9 8 6-1992: 69 f.
122 Th a i is, the letter sii. E ach letter o f the Sanskrit alphabet has several n a m e s that
arc used as co d cs in the description o f mantras; cf, Jayfikhya-Sarnhila 6 .3 2 c - 5 7 .
115
Ja y a k h y a -S a m h ita 2 6 .8 9 c -9 2 :
ro c a n a k u n k u m a ir y u k io n u m am autravidarbhilam H 89
lik h c t ka u stu b h a m a d h y c tu k a la so d a ia sa m yn tc /
cka ikii tu kala vipra pura y u k tiim rtcn a tu // 90
sam rtanam kala n a m ca y o ja y c c ca tathopari /
v isv a p y a y a sth it am viipi candraratim isam aprabham I I 9 1
tadbahyc 'stadalam p a d m a m v ilik h c c ca sakarnikam /
d c v y o 'n g a n i (cm. d ivyahgani) yatharc.jy.vp d a lesv abhyantarc lik h c t II 92
For o th e r e x a m p le s , sec Jayakhya-S am hitfi 2 6 .9 7 i I lb, 2 9 .1 6 3 c 172, 3 2 . 6 6 - 7 9 b ,
S atv a ta -S a m h ita 17.3 3 3 c35 9 b (translation in liik ita 1992: 1 9 3 -1 9 0 ; f o r a d ra w in g
o f the y a n tra d c s c r ib c d in this p a ssa g e , see Hikita 1990: 170), P a d m a - S a m h ita ,
caryapada 2 5 . 1 0 4 - 1 07b, A n iru d d h a-S am h ita 5.
M A N D AL A S AND Y AN TRA S IN THE P A N C A R A T R A
145
Thus, the installation (pratistha) o f the saudarsanayantra (wh ich is

established and worshipped like a statue; see bel ow pp. 148ff.) is
prescribed with the following words: Having established the god,
[who is] the Lord having the shape o f a yantra, in the centre (...) 124
Further, the deity is described as consisting o f all y an t r as ( sarvaya n tra m a ya )!25 Thus, a yantra is a representation of the deity on one
hand, and the deity encompasses all yantras on the other.
Tw o other kinds o f mantric safeguards (raksa) that are described
in the Ahir budhnya-Samhita are closely related to yantras. One is
called jyo tirm a yl, consisting o f light, and the other vanm ayl, c o n
sisting o f language.
Th e first is a wheel (cakra) whose constituent parts are repre
sented by divine beings and their activities. The bra h m a n is c o n
sidered as being the wh ee ls axle, the sakti as its hub, the saktP s five
activities, viz., disappearance (tirobhava). creation, maintenance,
destruction, and favour, as its spokes, the Vyuhas and Vyuhantaras
as its felly, and the Vibhavas as flames outside the felly. It is obvious
that this kind o f raksa can only be visualized mentally,126
Th e safeguard that consists o f language (v a n m a y l raksa) is a
wheel (cakra) whose shape is formed by writing mantras instead o f
drawing lines. The Ahirbudhnya-Samhita describes several forms o f
this kind o f safeguard.127
N ow h er e in the Ahirbudhnya-Samhitas text are these two kinds
o f safeguards called yantra. The term yantra is mentioned in the title
lines o f chapters 23 and 24 o f the Ahirbudhnya-Samhita and in two
subtitles on p. 218 o f the first volume o f the Ahirbudhnya-Samhitas
edition,1211 but not in the body o f the text itself. Thus the two kinds o f
raksa probably cannot be interpreted as yantras in the actual sense.
However, these safeguards, which are very similar to yantras, can
help to clarify the meaning o f yantra as representation o f the deity.
Both the raksa consisting o f lig h t and the one co n sis ti ng o f
l a n g u a g e are direct representations o f the divine power, the first
having divine beings as its constituent parts, the latter consisting
124 A h ir b u d h n y a -S a m h ita 47.21 cd:
m a d h y c d eva in pratisthiipya yantrarupadharam p ra b h u m //
125 A h ir b u d h n y a -S a m h ita 4 7 .5 6 d and 65a.
]%A h irb u d h n y a -S a m h ita 2 1 ,4 -2 9 b .
127 A h irb u d h n y a -S a m h ita 2 2 -2 4 .
I2fl A h i r b u d h n y a - S a m h i t a 23 is called vasudevadiyantranirupana, A h i r b u d h n y a Sam hita 24, yanlradcvatadhyiinanirupana.
146
MARION RASTF.LLI
merely o f mantras that arc manifestations o f the deity and his

aspects. These raksiis are not places where the deity can be made
present, but they are the deitys presence itself The same is true o f
yantras, which, in this aspect, are much more similar lo the two kinds
o f raksiis just described than lo mandalas.
However, seeing yantras as representations o f the deity is not the
only notion found in Pfulcaratra Samhitas, According to the PadmaSamhita, after the preparation o f a yantra, the deity should be invited
into it in order to worship him there.I2y Here, similar to a mandala,
the yantra is treated as being a place for the deitys worship.
What purpose do yantras have and how are they used? It is often
emphasized that mere visualization of a yantra or concentration upon
it is enough to reach a certain goal; for example: Me who recollects
it in danger, in a battle, or in a dispute has victory in his hands; here
there should not be any doubt. Or: This yantra, o excellent sage,
destroys all calamities. There is nothing that cannot be obtained by
wearing, recollecting, [and] visualizing [it]. 1'10 Also the y a n t r a s
mere drawing can have effect: Everything accrues to the people
through its mere drawing. 131 In general, however, worsh ipp ing it
with offerings, oblations and ablutions is seen as the prerequisite for
the e ffect o f a yantra.112
Yantras are frequently worn on the body as amulets, often bound
with thread and/or covered with metal. Using them in this way,
yantras are considered to protect and to have positive effects: He
should wrap this yantra with a five-coloured thread [and] put it into a
golden casket. [If] he puts [it] on [his] right arm o r a woman between
[her] breasts he/she is liked forever, even a m o ng enemies. Th e
sadhaka can easily cross female and male rivers or the oceans for
many purposes by means o f its power, and he does not sink in the
12,1P a d m a -S a m h ita , caryapada 25 .1 07cd.

130 A h irb u d h n y a -S a m h ita 26.80:
b h a ya g a m c ca sa m g ra m c vficlc va y a h sm a rc d idnm /
vijayas iasya hastastho natra karya v ic a n m a ll
and J a y a k h y a -S a m h ita 2 6 .1 04c105b:
ya n tro 'y a m munisardCila sarvopadravaniitianah II
dhiiranat sm a ra n a d dhyanan nasti tad y a n na sa d h a yct /
See also Jayakhya-S arnhita 26.93c, A h irb u d h n y a-S am h ita 2 5 . 17c,
131 A h irb u d h n y a -S a m h ita 25.2 led:
ctallckhanam atrcna sarvam sam padyalo nrnam II
132 See J a y a k h y a - S a m h i t a 2fi.93c, 2 7 .2 1 4 c , 2 8 .1 2 a , 2 9 .5 2 d , P a d m a - S a m h i t a ,
caryapada 32.1 15c1 16c, A n iru d d h a-S am h ita 5.59. Cf. also Rastelli 2000b: 350f.
M A N D AL A S AND Y AN TRA S IN THE P A N C A RA T RA
147
water. Fear o f beings o f the jungle or the forest is unknown [to him].
Inauspicious planets bestow favourable [things]. Manifold terrific
and exceedingly frightening poisons do not trouble [him]. Weapons
o f thieves, etc., do not enter the joints o f [his] body. (...) The demons
wh o injure children, etc. ( balagrahadayah), leave the child and go far
away if this yantra is present on its body, A pregnant wom an bears
easily on account o f wearing [this yantra], A barren wo m an [and] a
[woman] whose new-born children die will have children. 133
Likewise, it is considered auspicious to have a yantra in o n e s
house: He who has placed a yantra in his house accomplishes
everything. IM
The yantras protecting and auspicious effects are also used on
other occasions. So, at a funeral, the dead body is put upon a yantra
that is covered by a cloth.155 A pill (g u lika ) that has been prepared for
the acquisition o f supernatural powers must be purified with incense
in a casket upon which a yantra has been drawn (yantrasam puta).[ib
Other than protection, a yantra may bestow anything one wishes
as already indicated in the passage quoted above: elimination o f
sorrow, diseases and obstacles, attainment o f children, friends,
133 Jayakhya-Sam hitsi 2 9 .1 72d180b:

tac!yantram p a rivcsta yct II 172
pancarahgcna sutrcna k sip c t ka n a ka sa m p u tc /
dordandc d a k sin c ku ry a t stri va stanayugantarc I I 173
p riy a tv a m satatam y a ti va irisv apt ca sadhakah /
nadfnadan sam udran va lilaya p a rik w g h a y c t II 174
ba hvartham tatprabhavac ca no m a jja tija k m ta rc I
naranyavanajanam tu saka sa d v id y a te b h a ya m II 175
bhajanti sa n u ku la m ca viparitasthita grahah /
na bad hate visarn g h o n u n a n cka m catibhisanam II 176
coradiSastrasam ghato gatrasandhisu no visct / (...)
balagrahadayo du ra m ty a k tv a balain p ra ya n ti ca 11178
ya tre d a m tisthatc ya n tra m kirn tu taddchagam tu vai I
lagnagarbhii ca ya n a n su kh a m su te ca dharanat II 179
b h a vct putravati vandhya m rtavatsatha p u tr in i I
( J a y a k h y a - S a m h ita 2 9 . 1 7 7 c -1 7 8 b is c orrupt.) Cf, also J a y a k h y a - S a m h ita 2 9 . 1 0 2 c 103, 3 2 . 8 1 - 8 4 , S a tv a ta - S a m h ita 17.352357b. N o t o n ly y a n tra s can b e w o rn as
am ulets. A m antra written on birch-bark or a p iccc o f cloth can also b e u se d as suchcf, Rastelli 2000b: 350.
134 A h irb u d h n y a -S a m h ita 2 5 .1 8cd:

ya n tra m y a sy a g rh c n ya sta m tasya sarvam p ra sid h ya ti 11
135Ja y a k h y a -S a m h ita 24.8cd.
Ja y a k h y a -S a m h ita 26.69ab,
148
MARION RASTELLI
consorts, kingship and wealth.1'17 Yantras arc considered to be so

powerful that even antidotes against them could be necessary in the
case o f an enemy using them against one. Such antidotes are presen
ted in the Jayakhya-Samhita: the vanlhnm udnl and the jayam antra
are successful in destroying the power of yantras.1111
A particular yantra is the smidar&inayantra, which is described in
the Ahirbudhnya-Samhita and, based on the Ahirbudhnya-Samhita,
also in the ParameSvara-Samhita.'w It not only contains linear dia
grams and writings o f mantras, but also pictorial representations o f
various deities.Mn Furthermore, the writing material, which should be
solid such as nietal or stone, is covered with a yantra diagram not
only on the obverse side but also, with a different drawing, on the
14!
reverse.
The saudarianayantra is used especially by kings and those w ho
want to attain kingship. They should install it in a temple and
wor shi p it daily in order to secure their kingship: Hear the
peculiarity o f the protective prescript ion for kings, o Narada. He who
desires kingship, he who is deprived o f kingship, or he w h o is
o v e r p o w e r e d by [other] kings, having realized that the m o s t
distinguished teacher who bestows the yantra o f SudarSana exceeds
all, should worship this [teacher] with great wealth and then should
worship the four-armed Narayana, the god, whose eyes are as large
157 E .g., J a y a k h y a - S a m h i t a 2 6 .1 0 4 c 105b, 1 1 0 c - 1 1 3 b , A h i r b u d h n y a - S a m h i t a

2 6 . 7 3 , 3 6 .2 6 c - 3 0 b .
'
'
Lllt Jay a k h y a -S a m h ita 8.30a, 27.131 a.
111
T h e d escrip tio n o f the saudarianayantra in PFiram eivara-Sam hita 23 is based
on that in the A h irb u d h n y a -S a m h ita . T h e P aram eS v ara-S am h ita e v e n refers e x p l i
citly to the A h i r b u d h n y a - S a m h i t a as its source. C o m p a r e Pilram cS v ara-S arn h ita
2 3 .2 c - 3 : In fo rm e r tim es in order to a p p ease the great sin of'cu ttin g B r a h m a s h ead
N arayana taught, at the auspicious Badarika&rama, [the yantra] that re m o v e s all sins
to S ankara, w h o was asking for it. (p u n l n arayaiw noktam p iw y c b a d a rik te ra m c II 2
brahm anah Sirsavicchcdam ahapataka& lntayc / prcchatah ,i ankarasyatba sarvapapapan o d a n a m ll 3) with A hirbudhnya-S am hitii 2 5 .1 4 c~I 5: In o rder to a p p ease the great
sin o f c u ttin g P i t a m a h a 's h e a d N a ra y a n a taught me [the yantra] that p a c ific s all
a f f lic tio n s [and] k e e p s o f f all pain at the a u s p ic io u s B a d a r ik a S r a m a . ( m a n i a
naraya n en o kta ip p u n y c badarikatiramc II 14 p itanialiaiiira& hcdatm hapafaka& lntayc 1
sarvabadhapratiam anam sarvaduhkhaniviiranam II 15) Cf. also PFiramcSvara-Samhita
2 6 .4 3 a b c : () a c c o rd in g to the rule, as A h irb u d h n y a has ta u g h t it at length to
N arada, w h o has asked [for it] (...) (p rcch a tc naradayaitad a h irb u d h n ycn a vistarat /
y a th o p a d ista m v id h iv a t).
m T h e m o s t im p o rtan t d eities on the saudar&inayantra are Sudar&ina, V i s n u s
discus to w h o m the A h irbudhnya-S am hita is especially devoted, and N rsim ha.
A h irb u d h n y a -S a m h ita 2 6 .5 - 7 2 , ParamcSvara-Samhilfi 23.1 6 -1 17b,
M AN D A L A S A ND YANTRAS IN THE P A N C A R A T R A
149
as a lotus, who is dark-coloured, who [wears] yellow garments, and

w h o is adorned with all [kinds of7] ornaments according to the
prescriptions as taught by the teacher. He should have an excellent
ya nt ra made, which is made of gold, is decorated with jewels and
corals, and furnished with all [kinds of] ornaments. Only with this
in str u m en t docs he attain sound kingship. Having installed [this
y an tr a], which bestows the attainment o f everything, he should
worship it respectfully. Then he will obtain the earth with [its] seven
divisions (sapiadvipam ) [and its] towns, 1'12
In the Ahirbudhnya-Samhita, the temple (v im a n a ) for the saudarSanayanira and the prescriptions for wor ship to be performed
there are not described in detail.M3 Howe ver , the author o f the
y o u n g e r Paramevara-Samhita, who strongly emph as izes temple
w o r sh ip in general, has elaborated the descriptions o f the temple
(prasada) or pavilion ( mandapa) that should be built for the
saudarSanayantras worship, and also the pr es cr ip ti o ns for its
wor shi p,14'1The yantras worship is very similar to the com mon daily
ritual in a temple. This means that this yantra takes the position o f a
place for the deitys worship, similar to, e.g., a statue or a mandala.145
142 A h irb u d h n y a -S a m h ita 2 6 .8 2 -8 7 :
nljnfim raksavidhnnc hi vitc sm n iriu i narada I
ra jya rth i hrtarajyo va p a rib h u io thava nrpaih II 82
saudar& m asya yantrasya pradalaram g u ru m param t
sa rv c b h y o h y a d h ika m m a tvi1 tarn abhyarcya m abildhanaih // 83
tato narayanam d cva m p u n d a rika ya tcksa n a m /
syiim a la m pita va sa n a m sa rva b h a ra n a b h u sita m ll 84
aradhayce caturbahum a ciiryo kta vid h a n a ta h /
tap ta ja m b u n a d a m a ya m m a n ivid ru m a citrita m // 85
sa rva la in ka ra sa m yu kta m ka ra ycd ya n tra m uttam am /
eta! karanam atrcna rajyam a p n o ty ana m a ya m H 86
p ra tisth a p y a rc a y e d ctat sa d a ram sarva sid d h id a m /
tato b h u m im a va p n o ti sa p ia d vip a m sapattanam II 87
Cf. also Param eS vara-S am hila 2 3 .8 c - 9 b : K ings w ho arc d isp o se d to p ro te c t [their]
p e o p l e arc to p e r f o r m c o n tin u o u s ly th is w o rsh ip . O t h e r w i s e a lack o f f i r m n e s s
a r is e s . {prajiipalanaisilanam b hupiinam cla d arcanam II nairantarycna k a r ta v y a m
a n yatha ja y a tc 'd h rtih I)
1-0 T h e tem p le is m entioned only in A h irb u d h n y a -S a m h ita 36.35c, 36c, an d 40c.
1+1 S e e Param eS vara-S am hila 2 3 . 12 - 18b for the te m p le and P ara m c v a ra-S arn h ita
2 5 - 2 6 for the ritual prescriptions.
145 T h e re is m u c h in scrip tio n al and literary e v id e n c e th at im a g e s o f S u d a r t a n a
w e r e w o r s h i p p e d in P a n c a r a tr a te m p le s , o fte n in s p e c ia l s h rin e s. S e v e ra l su c h
im a g e s h ave been p reserv ed (cf. B e g le y 1973: 68ff.). Illustration 1 s h o w s a b ro n z e
im a g e o f SudarSana w ithin a yantra. T h is im ag e is d ated by B e g le y 1973: 90 to
ab o u t the 17th century. A lth o u g h it d o c s not look cxactly like the saudarsanayantra
150
MARION RASTIvLLI
The saudarsanayantra is not the only yantra that is worshipped in a

temple or pavilion. As already mentioned, yantras arc worshipped in
a mandapa also according to the Aniruddha-Samhita. Here, however,
they are drawn directly on a platform ( vedi) and not upon a mobile
material.uu' According to the Parama-Samhita, a yantra is used in
place o f a statue during the procession that is a part o f the
consecration (pratistha). The yantra that represents Visnu is carried
around the temple. After this, it is installed in the temple, and only
then is the statue o f Visnu established.1'17
Finally, a particular yantra that is related to the saudarsanayantra
should be mentioned. This is the dharakayantra, the yantra o f the
wearer, i.e., the wearer o f the saudarsanayantra. The power o f the
saudarsanayantra is considered to be so great that a human being
ca n n o t w e a r it wit ho ut addit io nal ly h av in g a dharakayantra:
[Narada:] Who wears this very wonderful divine ornament? If it is
worn, 1 do not notice the power (takti) o f anything. Please remove
[my] doubt [that has arisen] on account o f its excessive p o w e r .
Ahi rb ud h ny a: Truly, no one can wear this [yantra] o f great
splendour without [also wearing] the following, other yantra that is
full o f power, o divine seer. Hear now its nature [and] energy, o best
sage.
The dharakayantras most exterior part is the square
earth mandala with the seed syllable (blja) o f the earth on each
corner and two Nagas on each side. Within the earth mandala is the
fire m andala having the shape o f a hexagram with the fire seed
described in the A hirbud h n y a-S am h itii, it gives an idea o f w h at it m a y h a v e lo oked
like.
,mCf. p. 143.
147 Cf. P aram a-S arn h ita 19.62.6 3 b : He should then prepare a yantra o f the great
god and, after having asked for perm ission, have the initiated g u ard ian s o f the statue
lift the yantra that is Hari on to a co m fo rta b le p alan q u in that is e n d o w e d w ith an
arc h . ( tato d cv a sy a m ahato ya n tra m cka m vidhaya ca / d ik sita ir m u rtip a ir ju s ta m
s ib ik a m to ra n a n v ita m I ab h y a n u jn a m c;i y a c itv ii y a n tra m a ro p a y e d d h a rim I) Cf.
P ara m a -S a m h ita 19.70cd for the y a n tra s a n d 19.72b for the s ta tu e s installation.
148 A h irb u d h n y a -S a m h ita 2 7 .2 c -5 b :
[niiradah-]
ctad atya d b h u ta m divyarp dhriyatc kcn a bhusanam // 2
na casya dharanc ia k d n i ka sya cit k a la y a m y aham /
ntisa ktita yfisycm a m sa m sa ya m ch cttu m a rh a si// 3
ah irbudhnyah sa tya m na k c n a c id dhartum p aryatc ia n m a h a d y u ti!
rte yantrantariid a sm n d dcvarsc sakti& llinah I I 4
irn u tasya m u nisrestha svarupam viryam adya v a il
Cf. also A h irb u d h n y a -S a m h ita 2 7 .2 4 c -2 6 .
MA ND AL A S AND YANTRAS IN T H E P A N C A R A T R A
151
syllable. The round wind mandala with ils seed syllable149 is within
it, and within the latter, a wheel with ten spokes. Two syllables each
o f the saudarfanamantra and the narasimhamantra are written on nine
spokes, and on the tenth spoke, the word hana, kill. Obviously, this
combination keeps the saudar$anayantra's power in check. An eightpetalled lotus with the mantra om and the names o f the desired object
and the person the yantra is directed to ( sadhya) is in the centre o f
the wheel.150 The penultimate item seems strange in the case o f the
dharakayantra whose only purpose is the fitness for wear ing the
saudarsanayantra. The person it is directed to can only be its wearer.
T he example o f the dharakayantra shows how powerful yantras were
considered to be, and that their power could get out o f control if they
were not treated properly.
w F o r th e s h a p e s o f the e l e m e n t s m a n d a la s , cf. also J a y a k h y a - S a m h i t a 10.26,

3 6 c - 3 8 a , and 4 3 c - 4 4 b and G u p la /H o c n s /G o u d ria a n 1979: 1 7 2 -1 7 4 . T h e bijas o f the
e le m e n ts vary in th e d iffe re n t traditions. A c c o r d in g to the J a y a k h y a - S a m h ita , the
e a r t h s bija is slam , the fire s s r iim , a n d th e w i n d s h y a m (J a y a k h y a -S a m h ita
10.17c20b), cf. also G u p ta /H o c n s/G o u d ria a n ibid.
15,1A h irb u d h n y a -S a m h ila 2 7 .5 c16.
152
MARI ON RASTELLI
1. T h e I 6 'a r m e d Sudarsan acak rap u ru sa in the ^rT-Kalaniekapcrumal T em p le
M A N D A L A AND YA NT RA IN THE S I D D H A N T A S C H O O L OF
SAIVISM: DEFINITIONS, DESCRIPTION AN D RI TU AL USE*
Helene Brunner
(Translated from the French by Raynald Prevereau)
Introduction
It is co m m o n to refer to the ritual use o f more or less co m pl ex
drawings among the defining characteristics o f every denomination
o f Tantrism. Generally, such drawings are called mandalas, but also
yantras and cakras, with little consideration as to whether these terms
are synonyms or not.
This paper essentially aims at clarifying this terminology, and this
will be done in the first part. In the second part, I will discuss the
mandalas used in the cult o f Siva: analyzing a simple example, I will
describe their general structure and indicate ho w exactly they are
used in the rituals.
It is first out o f personal interest that I began investigating the
subject on which I here report my conclusions. Perplexed, undoub" T h is p a p e r is, broadly, a rem ake o f an article p u b lish ed in F ren ch s o m e fifteen
years ag o (B ru n n er 1986). That article rcp ro d u ccd alm o st verb atim a lecture given in
Paris in June 1984 on the occasion o f a c o n fe re n c e o rg a n iz e d by A. P a d o u x in the
c o n te x t o f the r e s e a r c h te a m no. 2 4 9 o f th e C e n t r e n a tio n a l de la r e c h e r c h e
sc ie n tif iq u e ( C N R S ) e n title d L T l i n d o u i s m c te x te s, d o c tr in e s , p r a t i q u e s . T h e
present paper is a coniplcte revision o f the p revious one. N o t only d id 1 e lim in a te the
oral c h a r a c te r o f the p re s e n ta tio n , b u t 1 also m o d ifie d se v e ra l e x p r e s s i o n s th at
se e m e d correct in 1984 but that p rogress in the study o f S aivism n o w sh o w s to be
incxact. 1 also rew o rk ed som e long p a ssa g e s by introducing useful precisio n s, added
m a n y references and inserted som e c o m m e n ts in part in spired by th e d iscu ssio n (not
re p ro d u c e d here) that follow ed the lecture in Paris. It was, h o w e v e r, not p o ssib le fo r
m e to ex ten d m y research . T h e re fo re , w o rk s on the su b jcct w h ic h a p p e a r e d a fte r
1986 arc not taken into consideration.
N o te by G. B iihnem ann: In this articlc the author uses the te rm c u l t in the sense
o f sectarian a ffilia tio n and lw o r s h i p 7 lritu a l, T h e w ord is not used in a d e ro g a to ry
sense.
154
I IliLliN l.i D RU N N FR
ledly like many other scholars, by the coexislcnce o f three terms that
modern authors rarely distinguish and often translate, in English, as
well as in French, by the same word diagram, I had developed the
habit over the years o f taking note o f the ritual contexts in which
those terms appeared. Soon enough, I realized that mediaeval authors
did not use the terms so freely as we do and thought that it would be
good if we imitated their precision instead o f creating confusion
where it did not exist by using a single word in our translations (and
I also accept this criticism). I was therefore pleased to seize the
opportunity provided by a conference held on this theme in Paris in
1984 to expand my research and submit the result o f my reflections
to the participants. Their reactions inspired some of the additions that
I have made to the original French paper.
I must insist at the outset on the fact that my research does not
cover all Hindu schools, not even all Tantric sects. Rather, 1 limited
m yse lf to the following texts:
(1) the fundamental texts o f the Siddhanta School, 1 those that have
co m e to be called S a i v a g a m a s or even sim pl y A g a m a s
(Mulagamas and Upagamas), but could just as well be called
Tantras since they often present themselves as such .2 1 looked at
1 W c must stop calling this school the S outhern S c h o o l, for w h ile it is true that
it is the South o f India that has kept its heritage alive, wc now k n o w that its m o s t
ancicnl texts c o m e from the North (in clu d in g the p n d d h n ti o f S o m a ia m b lu i, see my
introduction to S P 4 , pp. xliii-x lv ). Wc could call il the >aiva-Siddhanta S chool, but
since this te rm w as b o rro w e d from the Sanskrit School o f that d e n o m in a tio n by the
Tam il S chool that follow ed it and p ro fo u n d ly m o d ifie d it, and since the n a m e has
re m a in e d a ttach ed to the latter, wc should call il m ore precisely: S n iv a -S id d h a n ta
School o f Sanskrit e x p re s s io n or sim ply Sanskrit Saiva-Siddhfuita S c h o o l . T h a t is
what I k eep repeating (see, for exam ple, B ru n n e r 1977: 114-1 15 an d 1992: 38, note
2). T his appellation is here shortened into S id d h a n ta for the sake o f sim plicity.
2 See, for exam p le, p. xix o f the introduction to m y translation o f M rg c n d ra g a m a ,
kriya p a d a and caryapfida, and m ore recently G o o d a l l s in troduction to his edition o f
the Kiranavrtli, pp. x x x v i-x x x ix , It is useful lo note here that, a m o n g the T a n tra s o f
the Sid d h an ta that have reached us, rare are the texts that date from before the ninth
century. E x c e p t for the Kirana, the M rg en d ra and the M atarigaparam eSvara, th o s e
that w e re p u b lish e d in India, including those ex cellen tly ed ite d by N.R . Bhatt and
published by the Institut Frangais d Jndologic, b elong to a later period, even (hough
so m e o f th e m b o r r o w the n a m e o f a w o rk p re v io u s ly k n o w n a n d cited. F o r the
K a m ik a , sec the introduction to m y translation o f the M rg c n d r a g a m a cited ab o v e,
pp. x ii-x v . I will, how ever, have to refer to such works, which in fact, with regard to
the s u b j e c t h e r e u n d e r in v e s t i g a t i o n , m o st p r o b a b l y r e p e a t the t r a d i t i o n a l
instructions.
MA ND AL A AND Y ANTRA IN THE S I DD H A N T A
155
all the ones that were at my disposal, namely about ten o f them,
plus some preserved fragments of lost treatises;
(2) some Saiva Tantras o f the Trika: Svacchanda (SvT), Netra (NT)
and Malinivijaya;
( 3 ) a fair number o f handbooks (p a ddhati) o f the Siddhanta, the
most important o f which being the Somasambhupaddhati (SP),
called Kriyakandakramavall, written in K as hm ir in the 11th
century, and o f which I have published a complete translation.
The following handbooks, written in the South, depend more or
less directly on this work: the Aghorasivacaryapaddhati, called
Kriyakramadyotika, o f the 12th century; the yet unpub lished
Jnanaratnavali, the Siddhantasekhara and the Siddhantasaravali,
all three probably dating from the 13th century; and finally the
Isanasivagurudevapaddhati, a later work which is nonetheless
better known since it was edited early in the 20th century and
reprinted in 1988;
(4) some handbooks from the Trika School, such as the Tantraloka
( T A ) o f A b h i n a v a g u p t a an d the S a r a d a t i l a k a ( S T) o f
Laksmanadesika.
All these sources converge, so much so that the results o f my
research do not only concern the Saivism o f the Siddhanta School, as
the title o f this paper carefully suggests, but could probably apply to
a w i d e r range o f traditions. It is not certain, h ow ev er , that my
conclusions could, without further precautions, be extrapolated to all
Tantric schools, for example, to Saktism or to Pancaratra, nor to all
periods, for example, to the more recent Tantrism.
Nonetheless, I should note that the n o n - sy no n ym y o f the terms
mandala and yantra is accepted by the Sabdakalpadruma (s.v. yantra)
which quotes the following passage from the Yo ginl tantra,3 where
the possible supports for the cult of the goddess are discussed:
lingastham p u ja yed devlm pustakastham tathaiva ca /
m andalasthani nm ham ayam yantrastham pratim asu ca II
jalastham va silastham va p u ja yef param esvarim /
3
T e x t dating fro m the 16th c e n tu ry (sec G o u d r ia a n in G o u d r i a a n / G u p t a 1981:
85-86).
H ELEN EBRUN NER
156
I. Occurrence o f Ihc Three Terms in (he Ritual Texts

A . Mandala
Let us now look at the first point, which conccrns the occurrence o f
the terms mandala, yantra and cakra in the ritual texts. I will begin
with the one that is by far the most frequent in the standard rituals:
m a n d a l a a term that wc spontaneously associate with those
splendid drawings so characteristic o f Tibetan Buddhis m and o f
which we have seen a large diffusion over the last decades.
W he n and how do the Saiva texts use this w ord? W e must
obviously set aside right from the start the rather banal meaning o f
circ le (construction circle or any other disk) as well as that o f
territory or province, with which wc are not concerned here, at
least not directly. 1 shall therefore consider only the specific ritual
objects that the texts call mandalas. All appear as limited surfaces,
o f which I find three main types:
First type: a limited surface deprived of structure.
F or ex am pl e: the c o w - d u n g m a n d a l a enj oined on n u m er o u s
occasions to serve either as the scat for a god (for exam pl e,
Nalesvara, when he is called to preside over the dances performed by
the Devadasis in front o f Siva), for a man (the disciple, before his
initiation), or for a revered object (the cooking pot for the de it y s
rice, when it is removed from the fire and placed on the ground).4
Such mandalas arc made by smearing a generally circular portion
o f the ground with a semi-liquid paste made o f co w - d u n g or
sandalwood. I will call them seat-mandalas.>
S eco n d type: a limited surface showing a drawing generally made
o f the accumulation o f coloured powders,
This is the most interesting type o f mandala, one that we must:
most carefu ll y distin gu is h from those ot he r d r a w i n g s called
y a n t r a s , be c a u s e they bear some rese m bl anc e. Here are its
characteristics:
These mandalas serve as supports for the worship o f divinities.

They have no other use.
The y are temporary, being destroyed once the cer emony for
which they were built is completed.
4 Sec SP3, Index, p. 737, s.v. mandala (qiiclconque).
M ANDAL A AND Y ANTRA IN T H E S I DD H AN T A
157
They are constructed on a plane and purified area, and oriented.

The drawing (made with strings and compass) is geometrical; it
often shows a central symmetry (or, if one prefers, an axial
symmetry, with reference lo an axis perpendicular to the plane
and going through the centre);5 and it is entirely covered with
coloured powders (three, four or five different colours) hence
the exact name of these objects: rajom andalas/
The ir dimensions are sometimes considerable since they vary,

depending on the type or the text, from one to eight (according
to the Mrgendra) 7 or even eighteen cubits (according to the
Matarigaparamesvara), that is to say from ha lf a metre to about
fou r or even nine metres. The officiant is there described
entering and leaving through d o or s, mo ving around along
streets instructions that must be taken literally for the bigger
structures.
Le t us go over the first o f these characteristics; namely, that these
mandalas serve as supports for worship. The way to perform this cult
will be described in the second part o f this paper, but it will be good
at this point to specify the nature o f the worship in question. Here the
texts from the Siddhanta diverge from those o f the Trika. While the
latter recommend doing all the cults, including the daily cults,8 on a
mandala, the vast majority o f the texts o f the Siddhanta insist on
using the mandala only for the occasional (naim ittika) rituals9 such
5
T h e sq u are m a n d a la , o f central sy m m e try , with, at its centre, an eig h t-p e ta lle d
lotus, is by far the m ost c o m m o n m andala, at least in the norm al cult o f Siva. F o r the
o th e r g o d s, the m a n d a la m a y tak e o th e r sh ap es. F o r e x a m p le , a c c o r d i n g to th e
M r g e n d r a g a m a (kriyfipiida 8 .3 6 c - 3 7 b ) , the m a n d a la o f C a n d a is s e m i-c irc u la r, and
th at o f the g o d d e sse s in voked to sed u ce w o m e n takes the form o f a vulva, an eye or
an arc. T h e s h a p e and c o lo u r also vary a c c o rd in g to th e p u r p o s e o f t h e ritu al
( M rg e n d ra g a m a , kriyfipiida 8 .3 7c-38).
u T h e re arc, as we will see, d ra w in g s that, by their aspect and th e ir use, partake o f
th e n a tu re o f the m a n d a la , b u t th at arc not c o n s tr u c te d by the a c c u m u l a t i o n o f
pow d ers.
7 S e e M r g e n d r a g a m a , kriyfip iid a 8.30 and M a t a n g a p a r a m c s v a r a g a m a , k riy a p a d a
1.26a.
*
Sec, fo r ex a m p le , S v T i , c h a p te r 2, introduction to v erse 155. It is the s a m e fo r
the M r g e n d ra g a m a (see note 11).
lJ A c c o rd in g to the V ed ic classification, w h ic h in fact d oes not ap p ly well at all to
T a n tr ic rituals. T h e p r a tis th fi is g e n e r a l l y t a k e n as an e x a m p l e o f o c c a s i o n a l
(n a im ittik a ) rites; but since it is p e rfo rm e d on the initiative o f a p e rs o n w h o w is h e s
to a c q u ire m erit, it is s o m e tim e s classified a m o n g the op tio n al ( k a m y a ) rituals. T h e
utsavas m ay be spoken o f as o c c a s io n a l if one c o n sid e rs th eir perio d icity , b u t they
sh o u ld be called o p t io n a l on a c c o u n t o f their bein g p e r f o r m e d w ith a definite aim
158
HELENE BRUNNER
as the diksa, the pratistha, the p:\vitraroham , the utsava and the
optional ( kam ya) cults, that is to say all the rites performed for a
desire-oriented purpose.1" For the daily (nitya) cull o f Siva, even for
the private one, they prefer the linga." It is therefore with regard to
in view. T h e p a v itr a r o h a n a (see S P 2 , section II) b e lo n g s lo the 1p r f i y a s c i t t a '

categ o ry , w h ic h is a sso ciated with the occasional rituals out o f c o n v e n tio n only.
Finally, the d iksih are said to be o cc a sio n a l only from the point o f view o f the g u ru
p erfo rm in g the rite; those that are conferred upon the sadhakan to let th em a c q u ire
siddhis should lo gically count a m o n g the k a m y a rituals. O n e should note that these
long rituals (w h ich spread ov er several days, o f w hich the first days arc u sed for
preparatory rites designated by the general term adhivasa) can be p erfo rm ed only by
the acarya, if they are public rituals, or by the sadhaka, if th ey are private cults (on
the sadh a ka , see B ru n n er 1975), and that il is only these high ra n k e d initiates w h o
can tracc and use the rajomamlalas.
10
T h e only m andala described in the SardhatriAatikalottara (7. lab ) is p resen ted in
the context o f the ka /v y a k a n n a n , and this show s, according to its co m m e n ta to r, that
it c o n c c r n s o n ly the sadlm ka. T h e in te rp re ta tio n e x p re s s e d in this w o r k s e e m s
unusual, how ever, Indeed, it must be noted that som e o f the d e sire -o rie n ted rituals
that c o n ccrn a g ro u p o f people and not just one person, such as the purificatio n or
p acifying (&lnti) rites, can be and usually are perform ed by the acarya.
" W ith a few ex cep tio n s, for exam ple, the M rgendra, which d e sc rib e s the daily
cult o f Siva on a s th a n d ila b e f o r e c o n s id e rin g the p o ssib ility o f using a lin g a
( M rg c n d ra g a m a, kriyapada 3 .5 4 c - 5 6 b ) . Il is the o p p o site e ls e w h e re , sec, for e x
ample, SP1, pp. 2 2 6 - 2 2 9 , verses 102-1 0 3 , w here SomaSambhu, after describ in g the
cull on a linga, gives a list o f e q u ally acceptable supports (for Ihe private cult), but
c oncludes: liiigc 'p y a tya n ta m uttam am . T h e later w o rk s take in g e n eral a m o re
radical position. Q uoting the Purva-Kfiranagama, chapter 30, will suffice, In the first
p a ssa g e ( 3 0 . 2 c - 3 b ) , that text lists the different su p p o rts for the p riv a te cult: the
personal linga given by the g u m , the sthandila-, oneself; a tem porary (k sa n ik a ) linga\
a m a n d a la ; the w ater. Im m e d ia te ly a f te r w a r d s ( 3 0 . 3 c - 4 b ) , and in a s o m e w h a t
d ifferen t list, it assig n s a v alu e to the cults p e rfo rm e d on these supports: the cult
p e rfo rm e d on a m a n d a la is rated at 100; at 1000 if p e rfo rm e d on a sthandila-, at
10,000 on a k a u tu k a (p ro b ab ly a n arrow stripe o f cloth with draw in g s, later called
pata), and at 10 billion on a linga.
dalle ca guruna lingo sih a n d ilc sv a y a m a tm a n i / /
k sa n ik c m a n d a te to y c 'p y iitm arthayajannm sm rtatn /
im n d a lc tu Satam p u n y a m sth a n d ilc tu sahasnikam / /
ayu la m k a u lu k c lingc k o lik o tig n n a m bhavct / ( 3 0 . 2 c 4b)
T h e sa m e w ork takes up again the problem a little farther ( 3 0 ,7 - 8 ) , and establishes
the fo llo w in g scries, listing the cults in an increasing o rder in term s o f their value:
the m en tal cult; the cult on a m an d a la ; on a te m p o ra ry linga\ on a stripe o f cloth
(w ith d ra w in g s ? pata)\ on a painted im age (? a b h a sa )\ on an im a g e in the ro u n d
(bim ba); on a lin g a (w ith faces, to distinguish it from the next one); on a linga
d eprived o f a n th ro p o m o rp h ic traits ( n iskala-tinga}.
m anasan m a ndalam srcslhm n m andalat k sa n ik a m param /
ksa n ik a t p h a la m u tk fsta m patam caiva (atah param / /
p a ta d abhiisam utkrstam abhasat h im h a m ucyalc /
b im b a d vai lin g a m utkrstam lingad vai n iska la m param / / (3 0 .7 - 8 )
M ANDAL A A ND Y ANTRA IN THE S I DD HAN TA
159
th e occasional rituals, especially the diksa, that the Siddhanta texts

give a description o f the mandala: some will describe only one, like
th e Mrgen dra ( kriyapada 8.25c-53) and the Sardhatrisatikalottara
(chapter 7), but in general several kinds o f mandalas are suggested
for the officiant to choose from (see below), All these mandalas have
a complex structure, are rather long to elaborate, and remain present
for the complete duration o f the ritual for which they are used. When
o n e considers using a mandala for the daily cult o f S i v a , 12 the
mandala, which will have to be drawn each day, is o f course much
simpler. It is limited to the eight-petalled lotus that occupies the
centre o f the larger mandalas. According to the Suprabheda (kriya
pada 8 .8), it is drawn on a portion o f the ground previously smeared
with cow-dung, while according to other texts it is drawn on a square
platform made o f sand and grains named sthandila (thus the frequent
confusion between sthandila and mandala). But there is neve r any
mention o f coloured powders, The same instructions apply to the
mandalas used in the daily cult o f the secondary divinities, such as
Surya (see S P 1, p. 71, under [Id]).
In all cases, this second type o f mandala corresponds to the
following definition; it is a temporary divine image traced, with
s o m e exceptions, by the accumulation o f coloured po wd er s and
which must be beautiful to rejoice men and gods, I will call it the
image-mandala.
Third type: a limited surface that is squared but has no drawing.
W e also find under the name mandala some square surfaces suitably
squared and in the boxes o f which the officiant (rapidly) invokes
T h e idea, as wc can see, is to exalt the linga as the ideal support for the private cult.
A s for the te m p le cull, the q u e stio n d oes not crop up: it can only ta k e a p e r m a n e n t
im a g e as its support, that is to say a fixed lihga for &iva, a s c u lp te d im a g e for the
goddess.
12
T h e r e arc c ir c u m s ta n c e s w h e n o n e h as to. T h a t is w h a t I s a n a S iv a g u r u d e v a
e x p la in s (see TSanaSivagurudevapaddhati, k n y iip iid a 2 0 . 2 3 - 2 7 [= v o lu m e 3, p. 200,
6 - 1 5 ] ) : i f the adept is affectcd by a fa m ily im purity, he can n o t to u c h the linga, nor
the fire; he m u st therefore have s o m e o n e else p e rform the public cult, w h ile m en tally
re c itin g the m antras; then he m ust h im s e lf perform his daily cult ( c o m p u ls o ry ) on a
m a n d a la , a lw a y s m e n ta lly reciting the m antras. O ne co u ld ask if th a t ru le can b e
e x p la in e d b y the fact that, the m a n d ala bein g te m p o ra ry , n o im p u rity c o m in g fro m
the o ffician t can im p in g e on it definitively, w h ile it w o u ld h a v e a lasting effect on
the p e rm a n e n t linga; or if it can be ex p la in e d s im p ly by so m e essential inferiority o f
the m a n d a la c o m p a r e d to the lihga: the sim p le m a n d a la u se d h e r e w o u l d be less
p re c io u s than any linga, even the te m p o ra ry lihga that w o u ld be a llo w e d in the first
hypothesis. See also note 55.
160
I IliL E N E B R U N N IiR
some divine or evil powers in order lo make them favourable to his

cause with a food offering called bali.
These halimandalas are found in many rituals, including the daily
ritual. The best kn o wn o f these is certainly the one called
vastum andala, where 45 gods and 8 demons are worshipped (and
fed) before any construction, as well as a( some critical moments
associated with a given site. The works o f Stella Kramrisch13 made it
famous and loaded it with a symbolism that 1 for one have some
difficulty seeing but on which 1 do not have to elaborate here. What I
would like to emphasize, however, is that the term vastum nndala,
which we use systematically, is rare in the texts with which I am
most familiar.14 In the vast majority o f cases, these texts prefer the
terms pada and p adavinyasn to refer to these squared surfaces and
their construction; the same terms arc used when they describe the
division o f any square area (the ground of a sacrificial pavilion, the
site o f an agglomeration, etc.) in four concentric zones destined to
serve as guides for the ulterior arrangement o f the site.15 In all these
cases, we must take the term pada in the sense o f d o m a i n and
understand padavinyasa as the attribution o f their respective domain
to different entities. However, it will happen that the term mandala
be used in this context, and that is why I refer to these squared
surfaces as a third type o f mandala the only one, in fact, for which
the translation diag ram is appropriate.
I will include in the same category some simple geometrical
figures allowing for the distribution o f objects, for example, the
square divided in nine boxes which, according to some texts, serve to
fix (with the fall o f a flower) the name o f a Saiva initiate;1' or the
11 S ee K ram risch 1946, volu m e 1: 2 9 -9 7 .

14 It is not found in this contcxt in the following A gam as: the P iirva-K am ika, the
Suprab h cd a, the Ajita and the Kirana no r is it fo und in the M a y a m a ta o r the BrhatS am hita. W c find the w o rd m a n d a la tw icc in the d escrip tio n o f the P u rv a -K a ra n a,
oncc balim andala in the Saradiitilaka, once "m a n d a liid hnhyc" in a h a n d b o o k d e alin g
with pratisthii, and tw icc the term vasttnnandaki (oncc abbreviated to m an d ala) in the
Som aSam bhupaddhati (see SP4, pp. 46 and 386).
15 T h e n am e o f th ese four zones are, starting from the ccntral zone: briihm apada,
daivikapada, m a tm s y a p a d a and paiSacapada. A fifth z o n e ca 1led nlksasapada is
so m e tim e s added, see M ay am ata, volum e 1, p. 126, note 56 and Figure 9; and SP2,
pp. 3 3 2 - 3 3 3 and Plates I and II. T h e te rm pada, w hile referring here to the entire
zone, docs not lose the m eaning o f unit b ox.'
16 See S u p rab h ed ag am a, caryapiida 4 . 1 2 - 15b. T h e ccntral box and the four boxes
o f the principal directions arc those o f the five B rah m an s: the corncrs b e lo n g to four
MAN DAL A A N D Y ANTRA IN T HE SI DD HAN TA
161
s q u ar ed surfaces on which the p a h ca g a vya and other mixtures are

p r e p a re d .17 We can call these mandalas distributive diagrams.
Fourth use o f the term: There is finally one last use o f the term,
but it is totally heterogeneous with the preceding ones and does not
correspond to a category o f objects that could be integrated into our
classification. The mandalas o f which 1 am thinking are not, by the
wa y, material objects used for concrete rituals. Even though we can
d r a w them, they are mental objects that the imagination must create
and which, under certain specific circumstances, serve as supports o f
meditation. This is the case with the mandalas o f the five elements
m ent io ned in the descriptions o f bhutasuddhi as well as the descrip
tions o f the subtle body; this is also the case with the three mandalas
o f the moon, the sun, and fire (to which a saktimandala is sometimes
added) that appear at the upper end o f the throne o f Siva. The idea of
cos mi c domain is there inseparable from that o f a geometric symbol,
so that the inclusion o f these mandalas with the pr eceding ones
be com es impossible and all attempts at a translation fail.
W e therefore arrive at three well defined types o f mandalas: the
seats, the divine images and the distributive diagrams.18
B. Yantra and Cakra
I now resume my terminological exploration by looking at the terms
yantra and cakra, on which I will not elaborate so much. But first,
here are two preliminary remarks.
The first one is negative: to my knowledge, the mandalas that I
have ju st discussed, no matter the type, are never called yantras or
cakras in the Agamas. I, however, found one exception: the Upao f the six m e m b e r s ( a ngas). T h e p o in t o f fall o f the f lo w e r d e t e r m i n e s the
b e g in n in g o f the n am e o f the initiate, see SP3, p. 102.
17 See SP2, p. 320.
18 T h e distin ctio n s b etw een the th ree types o f m a n d a la s are certain ly n o t as clear
as this p a p e r leads lo believe. In particu lar, a qu ick o u tlin e o f a lotus or any o th e r
a d e q u a te d ra w in g can tra n sfo rm a s c a t into an i m a g e ; j u s t as a d r a w in g th at is a
little c o m p le x , m ade, for e x a m p le , with hulled grains on a ra w grains b a c k g r o u n d ,
tra n sfo rm s a sthandila (see above) into s o m e th in g that co u ld be called a m andala. On
the o th e r hand, it also h a p p e n s ( T 3 . 1 7 c - 1 8 a ) that the v a s tu m a n d a la are c o v e re d
w ith c o lo u re d p o w d e r s a fact that b rin g s th e m c lo s e r to the s e c o n d c a te g o r y o f
m a n d a l a s and co u ld c re a te c o n f u s i o n if w c fo rg et this e s s e n tia l d iffe r e n c e : th e
vastum andalas do not serve as supports for the cult o f a m ain g o d they are not even
c o n n c c tc d to any particular form o f Hin duism .
162
I IE L E N E B R U N N ER
game named Vatulasuddha describes in its third chapter, under the

title cakrabhedapatala, what is, in fact, an imagc-mandala; it calls it
cakra throughout the description, even once yantra. The explanation
for this infringement is doubtlessly contained in the final lines o f that
section, where it is said that the cakra can be drawn on a bark and
kept as an amulet. The Sricakra represents a better known exception.
My second remark will again contrast the texts o f the Siddhanta
with those o f other schools; the terms yantra and cakra are rarely
encountered in the Siddhanta (these terms do not appear in the lists
o f appropriate supports for the cult of iva), while they are frequent
elsewhere. It is therefore from the Tantras o f the Trika, in particular
the SvT and the NT, that I draw the characteristics o f these objects.19
1) Yanlras are drawings that differ in several ways from mandalas:
They serve only for the kam ya rituals, the desire-oriented rites,
and therefore conccrn essentially the sadhaka. The cult based on
yantras in fact only marks the first, stage o f their use. Indeed, the
yantras are generally kept after the cult and worn as amulets; or
buried for subsequent magic rituals; or eaten, after crushing the
support and mixing the resulting powder with milk or honey.
They are traced on durable materials: birch-bark ( bburjatvac,

bhurjapattra), copperplates, pieces of cloth, and now paper; they
are therefore small and mobile.
The representations they carry arc linear.
1,1
O n the basis o f these sam e two T an tras and their co m m e n ta rie s by K scm araja,
P ro fe s s o r A lex is S a n d e rso n c o m m e n te d , at (he c o n fe re n c e m e n tio n e d a b o v e (se c
Pad o u x 1986: 33), that they confirm ed the distinction that ! m ade b e tw e e n m a n d a la
and yantra: ...Y our prccisc distinction between yantra and m andala is c o n firm e d by
K s c m a ra ja w h o defines the f o rm e r (in its m ore com plex form ) as a c o lle c tio n o f
m a n tr a s w r itte n in a p a r tic u la r p attern (on N I'2 2 0 .5 9 c : yn n tra c a k ra m v iiis ta sam n ivc& ilikliito m a nlrasam uhah), w hile in its m ost basic form it is s im p ly a spell
written on a p iece o f birch -b ark (b h u rja p a tra m ),..." And he c o n tin u e d with a very
pertinent rem ark concerning the m ore subtle distinction betw een m andala and cakra:
As for the s u b tler distinction b etw een m a in ta in and c a kra if the m a n d a la is the
adharah (lo c u s) and th e cakra ( o f d eitie s/m a n tra s) the a d h c y a m (lo catcd ), then it
w o u ld follow that it is only the former that one can Iracc and that w h en one sp eak s
o f the m a n d a la to includc the circle o f deities (dcvaiacakrnm ) or m a n tra s (m antracakram ) w o rsh ip p ed in it, then this is by extension of'the prim ary se n se .
W c find a n u m b e r o f d raw ings o f yantras in the h a n dbooks o f p o p u la r T a n lrism ,
in H indi, a b u n d a n tly d istrib u ted by Indian b o o k sto re s. O f m o re refin ed art, the
d raw in g s o f the Balinese sorcerers (see the p o sth u m o u s book, l lo o y k a a s 1980) also
have so m eth in g o f (he yantra.
MAN DAL A A ND YANTRA IN T H E S I DD HAN TA
163
The drawing is engraved (rare) or (more often) traced with a

liquid some ink made from a variety o f often surprising
substances, such as blood and the bile o f a corpse in some cases
o f black magic.
The drawing is always completed with the inscription o f letters,

of bljas, each o f which makes a divinity present, and o f mantras
often containing imperative orders such as: Kill such and
such! , Heal such and such!
Their layout and use are secret.

T h e dominant idea o f the yantra is contained in its name, derived
from the root yam: with a yantra, the sadhaka constrains a divinity
to carry out a certain action for h i m .20 Just as those other machines
b ea r i ng the same name, the ritual yantra is first an ingenious
instrument.
T he N T keeps mentioning these yantras among the sovereign
remedies (for example, N T i 19.198b) and am on g the w ea po ns o f
magicians or sorcerers ( N T ] 18.88c).
2) The use o f the term cakra is much less precise. Apparently, it does
not refer to a category o f objects different from the mandalas and the
yantras.
Sometimes, the idea o f w h e e l is obvious, as in the case o f the
c a k r a o f t h irt y- tw o Saktis in cl u d ed in the m a n d a l a o f the
Sv acchandatantra (9.16ab and 9.24). But the term often simply
expresses the idea o f a coll ectio n or a m a s s : the mass o f the
divinities assembled on the same limited surface. Finally, cakra is
frequently used as a synonym for yantra, though we cannot always
tell if this practice is due to a lack o f rigour in the vocabulary or to a
change o f perspective. In those cases, the author may be talking o f
cakra to refer to the mass o f the divinities that are present, or to their
configuration, while using the word yantra to refer to the use o f the
object. But more research than what I was able to do w o u ld be
necessary to arrive at a convincing conclusion on this point.
Let us note, however, that the term mandala is never us ed in the
designation o f these magic figures (will we call them coercive
20
Sec an o th e r analysis o f the term in K u la rn a v a -T a n tra 6.86 cited in th e S ab d ak a lp a d ru m a (s.v, yantra) as c o m in g from another source:
kamakrodhadidosotthasarvaduhkhaniyantranat /
yantram ity lilnir ctasmin dcvah prinati pujitah //
164
11ELI'Nil BRUNNIiR
diagrams?) cxcept, o f course, for the construction circles or round

elements o f the total yant ra,
II.
D escription and Hit uni Use o f the Imngc-mnndnlns
A . Im portance
I men tion ed earlier that the mandalas were quasi in di spensable
elements in the occasional rituals. The chosen mandala is constructed
on the altar { vedi) that stands in the centre o f the pavilion (m andapa,
more exactly yagam andapa) where the ritual is taking place and it
serves as the principal31 support for the worship o f &iva during the
few days o f the ceremony. It is therefore present as a divine image,
and only as a divine image; that explains why, though that would not
be considered a good solution, the mandala can be substituted with a
mobile lihga placed on a sthandila,
B. Varieties
There are tens o f well differentiated forms o f mandalas, each being
de sig na ted with a specific term that so m et i m es e x p r e s s e s a
characteristic o f the drawing, sometimes the virtue o f the object. The
list found in ISanasivagurudevapaddhati, kriya p a d a 8. 3 1- 1 2 3 (=
vol ume 3, pp. 77, 8 ~ 85, 6 ) comes down to seven terms: bhadra,
sarvatobhadra, p a rva tika n ta , lataiihgodbhava, s v a stik a b ja d v a y a ,
svastikasarvatobhadra and cakrabja.22 But some other texts are more
prolix, such as the ArnSumat, which gives twenty names or s o .23
21 P r i n c i p a l , sincc there arc other s u p p o rts on which &iva must be w o rs h ip p e d
du rin g the c e rc m o n y that uses the m andala. In the m a n d a p a itself, aside fro m the
g u ru and e v e n tu a lly the disciple, there are, firstly, a vase o f w a te r p la c e d on the
n o rth -e a ste rn co rn er w h e re iv a is in stalled as the g u ard ian o f the sa c rific c and,
secondly, the fire (see SP2, pp. 5880; pp. 8 6 - 8 8 and Plates I-1V). If the c e r c m o n y
is o rg a n iz e d by a tem ple, the god o f the sanctuary n onetheless co n tin u e s to rcceivc
his cult, so that the priests often feel the need to rem ind through a special ritual the
essential identity o f all these apparently distinct Siva(s).
22 T h e list o f the R a u ra v a g a m a (kriyfipiida 2 5 .5 9 - 6 2 ) also counts seven term s, but
is so m e w h a t different. T h ere, N.R. Bhatt gives in the notes the co n stru ctio n o f each
o f these m andalas, as found in the hitherto unpublished Saivagam npaddhnti.
11
S ee Am&umat 4 3 . 4 0 - 4 7 , quo ted in R a u ra v a g a m a , v o lu m e I, p. 158, note 11.
T h e S id d h a n ta sa ra v a li, verses 7 8 - 9 1 , d escrib es ten m an d alas: h it;ilih g n d b h a v a (in
tw o s iz e s), n a vanabha, ananlavijaya, bhadra, p u n lk a ra (tw o sizes), latakaralinga,
subhadra, iim iikanta and sv a stik a plus a n o th er one used for the cult o f C anda; and
MANDALA AND Y ANTRA IN THE S I DD H A N T A
165
T h e s e numbers quickly multiply if we take into consideration all o f

the possible variations on a same theme, so that we finally arrive at
hundreds of di fferent mandalas.
Some Agamas maintain that the choice o f the mandala to be used
in a given ritual is not arbitrary but depends on the type o f ritual to
b e performed (for example, diksa or pratistha). The selection is even
m o re limited if we distinguish the private (atm artha) pratistha from
the public (parartha) pratistha and, in the case o f the public pratistha,
i f we take into account the nature o f the linga, which can be self
m a n i f e s t (svayam bhuva), e s t a b l i s h e d b y the god s o r o t h e r
supernatural beings (daivikadi), or established by men (m a n u s a ).24
Elsewh er e, we are asked to take into account the season or other
contingencies o f that order25 or, if it is a diksa, the social class o f the
initiate.26 But even if we accept all these restrictions (which not all
texts do mention), the definitive choice theoretically remains quite
vast, and in the end it is probably some traditions o f the schools that
w e r e decisive, each master most probably mastering the construction
o f only a small number o f these structures.27
C. D escription o f a Particular Mandala D estined for a D iksa
I f 1 just;
spoke in the past tense, it is because the mandala tradition is

not so alive in South India anymore.2* W e are left with the texts,
w e find a list o f e lev en in l& inaivagurudcvapaddhati, sa m a n y a p a d a 6 . 3 6 - 1 5 2 (=

v o lu m e 1, pp. 51, 1 0 - 6 2 , 2). At the o p p o site end, the M a ta rig a p a ra m e s v a ra g a n ia
(k riya p a d a 1 .2 6 -5 7 ) describes only tw o m an d alas for Siva and the K iran a (patala 20)
o n ly one, as is the ease for the M r g c n d ra g a m a as m en tio n ed earlier.
24 See Am Sum at, loc. cit.
25 See, for ex a m p le , P u rv a -K a ra n ag a m a 110.15c17:
m andaiam vedikordhve tu vasantadi ca sa d rtu //

vasantc svastikabjam ca grtsm c tu sarvabhadrakam /
pnlvrt ca bhadrain akhyatani lihgabjam svastikam ta th a //
darady cva tu hematite parvatikantamandalam /
padm asvastikam fikhyfitam sis ire tu viscsatah //
2,1
See the Saradatilaka, quoted in a S o u th -In d ian h a n d b o o k called D iksadarSa (p.
96, transcript no. 76 o f the Institut Frangais d ln d o lo g ie, P o n d ic h e rry ); the sta n z a is
not found in the printed editions o f the Saradatilaka :
vipratuup sarvatobhadram gauritila tirpasya tu /
v a iiy a n fm tu latfilingam iD dranam sv a stik a m b h a vet / /
21 Financial considerations w ere also present, sec note 47.
28
M o s t o f the o ff ic ia n ts n o w use s o m e d r a w i n g s p r e p a r e d in a d v a n c e on
c a rd b o a rd or cloth, w hich will seem a b e rra n t if w c think o f all those p a s s a g e s in the
HI-LIZNB BRUNNHR
which are rich in long and apparently very detailed descriptions, and
should in principle suffice. But alas! Whoever takes with enthusiasm
his ruler and pencil to translate these instructions into drawings will
soon be disappointed: the descriptions, as long as they may be, are
everything but clear. Therefore, all o f the attempts that I have seen o f
constructing a mandala strictly 011 the basis o f textual indications
have been disappointing: when they were not purely whimsical, the
drawings that were proposed were often hypothetical and always
incomplete, because a number o f constituting elements could not be
identified,w I know the problem quite well for having wrestled with
it when translating (he kriyapada o f the Mrgcndra.'1" I still have to
situate correctly the thirty-two doors o f the big mandala that is
described there, and until recently, more exactly until the conference
in Paris that 1 mentioned earlier, a series o f technical terms found in
that description remained mysterious to me. To most o f the problems
left unanswered up to that point I found the key in a very clear text
that I had ignored until then. It is the Saradalilaka o f Laksmanadesika (chapter 3) and its commentary by Raghavabhatta:11 The man
dala that I could draw (sec Illustration I) and on which I will
scriptures that explain the virtues o f the m an d alas through those o f the p o w d e rs o f
w hich they arc com posed.
w F or exam p le, the suiribjam andala o f which Gnoli gives the essential s tru ctu re'
in his tra n s la tio n o f the T a n lr a lo k a (1 9 7 2 , b e g in n in g o f p. 5 20) [note by G.
Biihnem ann; In the version published in 1999 the diagram ap p ears on p. 614. For a
diagram o f the m andala, cf, also Sanderson 1986: 171 and Illustration 2 in P a d o u x s
f irs t p a p e r in th is b o o k ] ; a n d th e mahrinr.uidah g iv e n by N .R , B h a t t in
M atarigaparam esvaragam a, volum e II, Figure 6.
S ec M rg e n d ra g a m a , kriyapfida 8.47c5 1.
31
T h is co m m en tato r, w ho w rote at the very end o f the 15th century, quotes m an y
so urces, in p articular several h an d b o o k s from the Siddhanta School.
,2 N o te by G. B iih n em an n : B r u n n e r 's rec o n stru c tio n o f the m a n d a la is a lm o s t

identical with the sarvatobhadram andala reproduced in c o lo u r (but not an a ly z e d ) in
D ak sh in a ra n ja n Shaslri 1940: 170 and 1963, opposite p. I and Banerji 1978: I76+.
B oth o f these b o o k s re p ro d u c e the sa m e m an d ala draw ing. T his must be a p o p u la r
d ra w in g since it also appears on the book co v er o f an Indian edition and translation
o f the D ev im a h a lm y a (Devi M ah a lm y a m [G lory o f the D iv in e M other]. 700 M antras
on Sri D urga, < S anskrit Text and> English Translation by Sw am i J a g a d isw a ra n a n d a,
M adras: Sri R a m a k rish n a M ath, fifth im pression, 110 date). B r u n n e r s d raw in g can
further be c o m p a r e d to the coloured print o f the sa rv a to b h a d ra in G h o s h a l Sastri
1983: 56+ and to a d ra w in g in m anuscript A 2 4 6/25 (labelled tilntrikakarm akanda)
preserved in the National Archives, K athm andu, T h e sam e m anuscript also contains
a s im p le r variant called la g h u sa rva to h h a d ra For a sim p ler sarvatobhadra/bhadraka,
see the sketch in the appendix, p. i in A p t c s edition o f the Pauskara-S nm hita (P art 1)
and the description o f the m andala in chapter 5 .2 1 - 2 8 o f the text. For a colour print
M A N D A L A A N D Y ANTRA IN THE S I DD H A N T A
167
c om m en t is one o f the mandalas enjoined there for the diksa, the

thoroughly auspicious one (sarvatobhadra). 1 chose it not only
bec aus e I had s uc cee de d in tracing it us in g only the textual
instructions and the commentary (to tell the truth, this was not so
difficult at all), but because its simple structure allowed me to show
an immediate symbolism equally applicable to the other mandalas.
The drawing starts with a squaring o f the initial square in 256 (16
x 16) boxes, indifferently called pada or kostha. These boxes are
grouped in four zones, the exact dimensions o f which are given in
padas in the text and reproduced in the legend that accompanies my
drawing.
1) The central zone (A) is called lotus (padma) because its space is
fully occ up ied by an eight-petalled lotus, the full geome tr ica l
description o f which is found in the text. Like all o f the lotuses
appearing in the ma ndalas, this one counts four parts; namely,
starting from the centre: the pericarp (karnika); the stamen (kesara),
covering the base o f the petals; the petals (patra, da7a), or rather the
region where they are visible and knitted together; and the tips o f the
petals ( d a la g ra ), not knitted to gether and who se form varies in
accordance with the goal in view.
2) The next zone (B), the width o f a pada, is called pitha, a term that
must be translated, as we will see, by throne. This pitha is made o f
four padas and four gatras, and it is the interpretation o f these terms
that will give its meaning to the whole structure.
Indeed, while pada evidently means foot, the meaning o f gatra is
far from obvious. I u n d e r s to o d it only w h e n I f o u n d a text
(Siddhantasaravali, verse 76) that gives the colours o f these parts as
follows, starting from the east: black and white; white and red; red
and yellow; yellow and black. That reminded me o f the description
o f the second section o f the throne o f Siva, the sim hasana which rests
upo n the anantasana. That asana is similar to a low square table
o f the s a m e m a n d a la , see: Prakrti: T h e Integral V isio n . V o l u m e 3: T h e A g a m ic

T rad itio n and th e Arts (edited b y B. B a u m e r, N e w D elhi: D.K. P rin tw o rld (P) Ltd .,
1995): 193+, Illu stra tio n s P.P. A p te 1,1.
33 T h e b ig m a n d a la s, such as the one fo u n d in the M r g e n d r a (see n o te 30) or the
m ahatnandala o f the Matariga, usually con tain m a n y lotuses: a central lotus fo r Siva,
eig h t p erip h e ra l lotuses for the d iv in itie s o f th e first c i r c l e (m o s t o f the tim e the
VidycSvaras), a n d s o m e tim e s still others.
168
HELENE BRUNNER
whose four tegs (padas), situated at the corners, each take a different
colour. It has on its sides four edgewise boards, n a m e d gatras
(because they are imagined as the bodies o f men or animals), that are
often referred to as bicolour because each half borrows its co lo ur
from the leg to which it is attached.M We should therefore u n d e r
stand the pitha o f the mandala to correspond to the sim hasana o f the
throne constructed for the cult o f 6 iva.'15 A full confirmation o f this
parallelism is given in the Saradatilaka, since upon following its
instructions to draw the mandala, wc find the padas and gatras
appearing where wc would expect them to appear if we accepted the
preceding hypothesis.
It is worth noting that, if the pitha is equivalent to the sim hasana,
it is the entire square that should be called this way, and not only the
zone that projects beyond the lotus. And this is indeed what we find
in many works.
Now to come back to the lotus itself, one will understand that it is
nothing other than the lotus in full bloom, with eight petals, that
forms the upper part o f the throne o f Siva, the one generally called
padmasana (see S P l , p. 154, note 1) on which the god is seated in
order to be worshipped. We therefore arrive at a first conclusion: the
central part o f the mandala (the lotus and its pitha) represents the
throne o f Siva reduced to its two essential parts; not as it could
materially be constructed, but as the practitioner mentally creates it
during the cult lo project it on the material pedestal o f the image that
he u s e s ,36 Exc ept for that a better faithfulness to the ritual
model this central part is equivalent to the pedestal (pitha) o f the
linga, in particular the lin g a o f a sanctuary. And since the plane
projection o f the linga itself would superpose on the karnika, we can
vt S e e S P l , p, 162, note 1, quoting A ghoraSivacaryapaddhati, n ity a k a rm a v id h i 35
(p, 88 o f the grantha edition),
3S In fact, to ju s tify the colours o f the g ;ltm o f the m andala, the c o m m e n ta to r o f
the S id d h a n ta sa ra v a li (a certain A nantaS am bhu) q u o tes tw o lines a p p e a rin g in the
p a d d h a ti o f A gh o raS iv a in the c o n te x t o f the a sa n a p u ja and w hich w c will find
q u o ted , w ith the half-;ilo k a that fo llo w s th em , in S P l , p. 163, u n d e r [ 5 0 b ] . A lso ,
N ara y a n a k aM h a , w hile c o m m e n tin g on M rg c n d ra g a m a , kriyapada 8 . 3 4 - 3 5 , w h ic h
discusses the ccntral lotus o f the m andala, refers, for another technical term, to a line
o f the S vT tak en from the description o f the throne o f Siva. It is th erefo re certain
that the S aiv a m a ste rs o f old w ere fully a w are o f the id e n tific a tio n at w h ic h I
painfully arrived that the p itha o f the lotus in a m andala represents the sim hasana.
S e e the a sa n a p u ja in S P l , pp. 1 5 4 - 1 7 6 or S vT [ 2 . 5 5 e - 8 2 . T h e im a g in a ry
throne, m ade o f mantras, must ov ersh ad o w the conerele pedestal, j u s t as the form o f
SadaSiva that will be visualized will ov ersh ad o w the material linga.
M A N D A L A A ND Y AN TR A IN T H E S I DD H AN T A
169
even say that the lotus and the pitha o f our mandala are equivalent,
from the point o f view o f the ritual, to the lihga o f a temple, provided
with its pitha.
3) Zone (C) is the street or i a n e ( vithi) where the officiant moves
around during his cult, It is therefore equivalent to the inside space
within the garbhagrha o f a temple, where movement is possible.
4) Finally, zone (D) represents the enclosure, constituted here o f four
kinds of elements:
a)
the doors (dvaras), that is to say the passages for entrance
and exit;
b)
the Sobhas, which are not just any embe ll ish me nt s (like 1
used to believe, and like some later co m m e n t a t o rs also
suggested), but the m onu m ent al doors themselves ( dvarasobhas in architecture);37
c)
the upasobhas, o f which I do not kn ow if there exists an
architectural m odel;38
d)
the corners (konas), first called w ea p o n s ,39 and which in
fact vaguely have the form o f a vajra.
All in all, the mandala o f the Saradatilaka represents, very sche
matically o f course, a minimal temple, with its unique enclosure.
And, ju st as the architecture o f a temple can become complicated
with the addition o f successive enclosures, so the m an da la can
b e c o m e co m p l i c a t e d , the b ig g e r ones p r e s e n t i n g up to four
enclosures (with two doors on each side, for a total o f 32) .40 In the
end we get a kind o f citadel.
37 T h e dva ra 6 o b h a is the e n tra n c e pav ilio n o f the first e n c lo s u r e o f a p a la c e or
tem ple, see A ch ary a 1946: 158, 243 and M a y a m a ta 2 4 . 2 - 2 2 (the w o r d is so m e tim e s
abbreviated to sobha).
* Since s o b h a is s o m e tim e s u s e d f o r dvarasobha, th e w o rd u p a so b h a p ro b a b ly
refers to the p av ilio n s th at top the s eco n d ary doors (for the upadvaras, see M a y a m a ta
9 .5 8 - 5 9 b ) , w h ic h should b e called upadvarasobhas.
39 In 6 T 3 . 1 12a, w h ic h a n n o u n c e s th at the tw o m o s t external z o n e s o f p a d a s are
re se rv e d f o r d v a r a s , sobhas, upasobhas and astras. C an w e in v o k e here the tridents
often seen on the w alls o f te m ples, at the corners?
* W e m u s t be c a re fu l not to p u sh the p a r a lle lis m to o far. In p a rtic u la r, the
s u c c e s s iv e e n c lo s u re s o f a m a n d a la h o st th e c i r c l e s d iv in itie s ( avaranadevadatas)
that the ritual places around Siva, not those that, acco rd in g to o u r A g a m a s , resid e in
the c n c lo s u rc s o f tem ples, A n y w a y , it is c lear that th e m a n d a la is not m a d e in the
im age o f the te m p le (the o pposite w o u ld be m ore likely): there are s im p ly b e tw e e n
the tw o a certain n u m b e r o f essential co rre sp o n d e n c es th at ha v e to b e kept in mind,
170
H H LIiN E BRUNNHR
Such an assimilation is warranted by the usual appellations o f the

mandalas that arc presented as houses (bhnvanas) o f Siva; and, for
the bigger ones, as towns or citadels (puma). II is confirmed also by
the fact that many names given to particular mandalas are also the
names o f some types o f towns. H is not likely, however, that we
would have arrived at any result in trying lo interpret our man dal a
and explain the technical terms that come up in its description if we
had started from that observation, since we would not have thought
o f trying to understand the central pari of the mandala via the ritual.
As far as I am concerned, the work o f interpretation is not
complete. However, a clarification o f the technical terms which have
not yet been explained would essentially not modify the general
vision o f the mandala that I was keen to present.
D.
Construct ion o f the Mandala
I now leave the narrow context o f the Saradatilaka to present in a

more general manner the ritual activity associated with mandalas.
First, their construction. This must be done on the same day o f the
ceremony that requires them41 and includes the following steps:
1.
2.
The aa iry a must first purify the ground'12 (leveled and prepared
beforehand) and locate appropriately the north-south and eastwest directions.
He or his assistant carefully then traces the axes o f the future
square, then its sides, and finally the chosen drawing, all o f this
with the help o f simple instruments: a cord, white powder and a
piece o f chalk. For the straight lines, one stretches between two
fixed points the cord covered with powder and, pulling it up by
its middle, immediately lets it go so that it hits the ground,
leaving a tracc; for the circles, one improvises a compass with a
cord o f the desired length and a piece o f chalk attached to one
extremity, the other being held fixed. The drawing must be
precise and respect scrupulously the given measurements.
41 So, for lhe big rituals, after the ndhivasit, sec S v T ] 3 .9 0 e - 9 1 b w ith c o m m e n
tary and 4 .3 4 - 3 5 ; or SP3, p. 228, note 155,
42 T h is instruction is not incom patible with the fact that the m a n d ala is g e n e ra lly
traced on a v e d i: the vccli is m ade o f beaten earth and must un d erg o the sam e p u rifi
cation proccss as any portion o f ihc ground destined to a ritual use,
M A N D AL A AND Y AN T RA IN T H E S I DDHANTA
3.
171
The acarya then pours some coloured powders on the drawing,

in sufficient quantity to form a notably thick layer. Each part o f
the lotus and each o f the other elements o f the whole receives a
particular colour, duly specified in the text that is followed.
Finally, everything must be covered, even the lanes, according
to some o f our texts.
Th e fingers used to pour the powders and the way in which to

proceed depend on the goal in v i e w .43 The same principle applies to
the materials used to get the three, four or five necessary colours.
While some Agamas, such as the Kirana, have modest demands in
this regard (cereal flour for white, minium or crushed cooked bricks
for red, coal or burnt chaff for black, curcuma or ochre for yellow,
crashed leaves for green),4'1 others, like the Mrgendra, accept these
substances only in the case o f ordinary dlksas, adding that i f one
wi shes for special pow er s or good fortune (and this m u s t be
applicable to other rituals than the diksa for wh ic h this is said),
precious materials must be used; namely, pearls, coral, gold and cats
eye for white, red, yellow and black respectively; whereas some
impure or harmful substances are well indicated for black m a g i c .45
Finally, some works offer different solutions (a good one, a middle
one and an inferior one) according to the financial possibilities o f
people;46 but we are then br o u g h t b a c k to the opi ni on o f the
Mrgendra since it is agreed that in all these cases the one who wishes
to get a precise favour from a divinity must not mind the expenses.47
See, for e x a m p le , TSanaSivagurudevapaddhati, k riya p a d a 8 .4 5 c - 4 7 b (= v o lu m e
3, p. 78, 1 6 -2 1 ) (em en d the first w o rd b h u k ty - to m u k ty -),
44 See K iran ag am a 2 0.1517 b :
y a v a g o d h u m a ja is cu rn a ii salitandulajais sitam /
dhatusindurajam raktam m rd b h ih p a k v e sta k a ir b h a v e t/ /
k rsn a m rajas tusair dagdhair angarair va su cu rn ita m /
haridrasam bhavam p ita m g a iriko d b h a va m eva v a / /
haritam curnitaih patrair haritais tat p ra k a lp a yet /
4i See M rg c n d ra g a m a , kriyapada 8.3940.
46 F or e x am p le, S u p ra b h e d a g a m a , caryapada 3 . 6 1 c - 6 5 su c c e ssiv e ly p ro p o se s for
white, red, black and yellow:
a) pearls, rubies, sapphires, gold;
b) shells, ja tilin g a (?), c ollyrium (krsnanjana), realgar (m anassila);
c) ricc flour, cooked bricks, burnt cereals, curcum a.
All o f this must o f course be crushed.
47 T h a t is w h y the p a s s a g e s th at e n jo in the u se o f p recio u s stones an d g o ld m u s t
not be c o n sid e re d m erely as theoretical. To be sure, m o s t o f th e adepts, in o rd e r to
m e e t w ith these textual injunctions, m u s t h ave b e e n satisfied w ith m ix in g , f o r each
172
IIliLLiNIl BRUNNER
These precisions provide us with a double teaching. The first is

that the use o f mandalas was a standard practice for the desireoriented rituals. This is amply confirmed by a nu m b er o f o t h e r
instructions concer ning (lie form o f the tips o f the petals, the
thickness o f the lines, the number o f circles o f divinities to be
worshipped around Siva, etc. The insistence on these details could
lead one to believe that despite the Agamic passages systematically
associating the mandalas with occasional rituals, it is for the ka m ya
rites that these multicoloured drawings were first conceived. But that
does not assimilate the mandalas to the yantras. Contrary to the
yantra, the mandala used in a kfunyn ritual certainly is not the direct
instrument in the action to be performed: it is not a ma gi ci an s tool,
but simply a means, for the sadlmku, to obtain the favour o f a chosen
deity. By wor ship pin g it on a splendid and costly support, he
improves his chances o f pleasing the deity and, as a consequence, o f
obtaining the boons that he craves.
The second teaching to get from these same passages is essential:
our authors believed in the intrinsic virtue o f the materials used in
the fabrication o f the powders. In general, they seem to consider as
obvious the fact that precious substances bring good fortune,^ and
harmful substances misfortune. In addition to that, some texts give a
more precise teaching concerning the correspondence between the
colour, a pinch o f a p recious m aterial with a c h c a p c r one; but th ere m ust have b een
others rich eno u g h to c o v e r at least a small, or even a big, m a n d ala w ith these co stly
pow ders. W e must recall, w hen reading our texts, that there w as no lack o f m o n e y in
M iddle A ge Indin, csp ccially a m o n g the kings or princes w ho, m o re often than not,
w ere those w h o spon so red the im portant rituals. T hat is w h y 1 tend to b e lie v e that
m a n d a la s m a d e o f p rc c io u s s to n e s w e r e a c tu ally c o n s tru c te d ; n o t f re q u e n tly o f
c o u rs c , and th e y w e re p r o b a b ly not v e ry big, but I do not b e lie v e th a t th e ir
descriptions arc purely theoretical. W e should also note that these m aterials w ere not
lost for e v erybody; they b ecam e the property o f Ihe main officiant, like all the rest o f
the m aterial used in the yiigam andapa (see SP4, p. 251, verse 72bc). T o think that
s o m e ficiirya co uld h a v e p u sh ed their rich d iscip les to e n g a g e in such s u m p tu o u s
ex p en ses is a step that we m ay or m ay not want to lake.
N ote also that financial co n sid e ra tio n s a lread y play a role in the c h o ic e o f the
m an d ala, the b ig g e r and m ore com p lex o nes requiring b ig g er quantities o f co lo u red
p o w d e rs. Sec R a u ra v a g a m a , kriyftpuda 25.60(1 w h ic h , after d e sc rib in g se v e n m a n
dalas, ad d s that one will c h o o s e a m a n d a la a c c o r d in g to his o w n m e a n s (y a th a vibhavam ).
4* It must be noted here that each o f the precious stones p o sse sse s a given virtue,
but that docs not seem to be the first reason for their use in the m andala.
M A N D AL A A N D Y AN TRA IN T H E S I D D HA NT A
173
colours, some deities and some fortunate effects;49 but these indica
tions vary too much between the sources to speak o f a solid tradition
and a real conviction. The only point on which everybody agrees is
that these powders make the mandala powerful an idea repeated
over and over.
E. W orship o f Siva on the M andala
N o mantra is enjoined during the construction o f the mandala. Once
completed, the mandala therefore is not yet a divine image no more
than a carved linga, before the pratistha ceremony. It will b ec o m e
one wh en the cult will have br ought Siva and the p o w e r s that
accompany him down on the mandala.
That cult, again, is a cult o f Siva on the mandala, not a cult o f the
mandala as such, despite the term mandalapuja sometimes used. It is
performed like the lihgapuja, a cult o f Siva on a linga:50
The stages o f the puja are the same, with the difference that, the
support be in g te m p o ra ry , the invitation ( avahana) a n d the
dismissal ( visarjana) o f the god must be u n de r st o o d in the
strictest sense. It goes without saying that the ablutions are made
mentally.
The mantras recited are also the same, m o s t especially the

phonic seeds ( bijas) that are their essential part since, properly
pronounced, they make present the divinities o f which they are
49 S e e , f o r e x a m p l e , th e p a s s a g e o f th e M a h a k a p i l a p a n c a r a t r a q u o t e d b y
R a g h a v a b h a tta (p. 123, 1 7 -2 2 ) in his c o m m e n ta r y on S T 3.124. E a c h co lo u r five,
in that text is co n n c c tc d w ith an ele m e n t, p la c e d u n d e r the influence o f a d ivinity
and s u p p o s e d to bring a specific effect. T h ese effects, in reality, are all o f the sa m e
order, that is to say the destruction o f d e m o n ic po w ers; the result is s im p ly that the
gods are h a p p y .
T h e p o s itio n o f the S u p ra b h e d a g a m a ( caryapada 3 . 5 6 c - 5 9 ) is d iffe re n t b u t n o t
m u c h m o r e c o n v in c in g . T h a t text, e v e n th o u g h it s u g g e s ts five c o lo u r s for the
m an d ala, only sp e a k s o f the sy m b o lis m o f the colours w hite, red and black , w h ic h it
n a tu rally c o n n e c ts to the th re e g u n a s and th e th re e g o d d e s s e s (V arna, J y e s th a and
R audrl). In a last p a ssa g e it says that th e y e llo w is a d d ed in o rd er to o btain the fruit
from th e yaga."
On the s y m b o lism o f c olours and their m ag ic use, se c G o u d ria a n 1978, c h ap ter 4.
S) T h a t c u lt is d e s c r ib e d in all the T a n tr a s an d h a n d b o o k s . S e e th e n u m e r o u s
references g iv e n b y N .R . B h a tt in his edition o f the Ajita, c h a p te r 20, note 1. A m o n g
the texts q u o te d there, on ly the K irana, the M a ta h g a p a r a m e s v a r a and the M r g e n d r a
are e a r lie r th an th e S o m a a m b h u p a d d h a t i w h o s e d e s c r ip tio n ( S P l , s e c tio n HI),
though concisc, is com plete, logical an d one o f the m o s t reliable o n e s w e have.
174
HELENE BRUNNER
the sound body. The officiant imposes them (by mean s o f

flowers) unto the mandala as he would do on a linga and its
pedestal. The result is that the group o f divinities that inhabit the
mandala when all the invocations are completed is identical to
the group o f divinities who inhabit the sanctuary o f a temple (or
what stands for it in a private cult) during the cult o f Siva. That
these divinities be represented or not on the mandala by a parti
cular symbol (lotus, svaslika, etc.) is o f no importance what so
ever.
The meditations and visualizations are (hose involved in any
cult; they have no special features that would link them to the
particular structure o f the mandala. Moreover, the texts do not
mention any mental co ur sc, leading, for example, from the
periphery to the centre, as is enjoined in other traditions. There is
indeed a motion, but it is on the whole a centrifugal one imposed
by the normal enacting o f the puja. Starting from the central
lotus, where the throne o f the god, then the god, arc successively
worshipped, the cult progressively includes the peripheral deities
by enlarging cach time the concerned circle (iivararui). The se
remarks remain true in the case o f a diksa: though said to be
indispensable, the mandala is treated like any other cult support,
without any particular role; and it is used as it would be in the
context o f another ritual, a pavitrarohana, for example. In other
words, the Saiva diksa does not take advantage o f the particular
form o f the mandala o f which it requires the construction.51
F. Virtues and S ym b o lism o f the Mandala

If the Saiva mandalas are neither privileged means o f reintegration,
nor direct instruments o f initiation, then what proper quality do they
possess that makes them more appropriate for some rituals than other
cult supports?
I already mentioned the particular virtue attributed to the powders
o f which they are made, and the incessantly repeated affirmation
that, because o f them, the mandala is a powerful image. Another
characteristic often advanced is its beauty, due to the brightness and
51
S o m e texts, su c h as the M rg e n d ra an d the S v T , su g g e st the u se o f th e b ig
m a n d a la to fix th e n a m e o f the initiated disciple. But this ritual can be d o n e on a
very sim p le d is trib u tiv e m andala, and it is certainly not for this p u rp o se th at the
rajom andala is constructed.
MA ND AL A A ND Y AN TRA IN THE SI DD HAN TA
175
the richness o f the colours used, The initiate w h o constructs the

mandala is asked to make it as beautiful as possible. Through the
fineness and exactitude of the drawing, the precision o f the colouring
and the good taste evinced in the confection of the ornaments that are
left to his initiative, the officiant must strive to create a perfect
image. Is it to rejoice men, as it is sometimes su ggested or to
charm the gods, as other texts would have it? 52 Probably both. Faced
with a splendid mandala, men are happy and feel their love o f the
gods growing,53 and the gods are better disposed toward men. We
must admit that for cults that are performed in an open pavilion,
exposed to the view, beauty and brightness o f the support are no
negligible qualities. However, it seems that a statue or a richly
draped lihga54 would be ju st as impressive to the spectators (and
probably also to the gods
and I tend to believe that, despite the
importance given to aesthetics in the Ag am as , the choice o f a
mandala as the support o f a cult is more dictated by faith in its
intrinsic power than by the desire to create beauty.55
Or could there be more pertinent reasons? One would hope to find
further justification for the eulogy of the man dal a throug h other
considerations than the nature o f the pulverized materials and the
brightness o f their colours, to dig out o f the arid texts the profound
signification o f these objects that other traditions have loaded with so
many virtues .
W e naturally think o f the cosmic symbolism56 on wh ic h all the
authors who discussed the question have insisted. And, certainly, we
cannot deny that even the very simple mandala that I tried to analyze
possesses one. But that same symbolism exists in the pair formed by
52 See TA 3 1 . 4 led .
53 T h e S u p r a b h e d a g a m a g o e s further: the m e re v is io n o f the m a n d a la c lean ses
from all sins (caryapada 3.2); the soul is deliv ered fro m all the fetters that tu rn e d it
into a p a i n (ca rya p a d a 3.41 ab). Such p a ssa g e s, w h ich o f co u rse m u s t n o t b e taken
literally, at least sh o w the im p o rta n c e o f the v isio n o f th e m a n d a la n e v e r e q u a te d ,
to m y k n o w le d g e , to the dartiana o f the g o d w h o inhabits it: the id e a is to see the
e xterior form itself.
54 Just as the m andala, but con trary to the fixed linga o f the s a n ctu ary o f a tem p le,
these m o b ile im ages can be seen by all.
55 W h y th e n is th e m a n d a la d is q u a l i f i e d f o r th e d a ily ritu al, at le a s t in the
S id d h a n ta ? I can only see one logical reaso n for this: the fact that the m a n d a la s used
fo r d a ily rituals are, as w e h a v e seen, n e c e s s a rily s im p le , p r o b a b ly d e p r iv e d o f
c o l o u r e d p o w d e r s , an d th e r e f o r e s h a r e n o n e o f th e v ir tu e s a t t r i b u t e d to the
rajom andalas. T h e question, h o w ev er, m erits further in vestigations.
% On the s y m b o lism o f colours, see n o te 49.
176
liHLHNF. BRIJNNl-'.R
a lihga and its pitha, in the temple, in the city; and it is not expressed
here with any more precision or enthusiasm than there. We must be
careful not to give in to our imagination or our desires and add to the
texts that we have at our disposal; and these texts do not encourage
us to do so. To my knowledge, they do not even make explicit the
immediate symbolism that makes the mandala a miniature temple or
even a city, though it is suggested by their vocabulary. They dwell
even less on that cosmic symbolism with which we Westerners are
so obsessed. It is not that they ignore it, but they leave it to the
description o f the ritual as such to bring out the correspondences
between the different parts o f the mandala and the cosmic realities,
and it seems vain or even dangerous to want to add more. It is by
orienting the research in that direction, that is, by analyzing closely
the rituals that have mandalas as their support or pretext, that we
must attempt to bring some precision to those symbolisms, instead o f
desperately trying to make them come out o f the static structure o f
these same objects.
Conclusion
I am afraid that many readers will be disappointed, or even shocked,
by my stripping the ancient Saiva mandalas o f everything that the
imagination, drawing from other sources, had superimposed on them.
However, by bringing them back to what I consider their real status,
that o f divine images, no more and no less charged with symbolism
than the others, but characterized by the special power provided by
the powders o f which they are made and by the power o f seduction
that results from their beauty, I have not deprived the mandalas o f all
signification. Rather to the contrary. However, I did separate them,
much to my regret, from our mental model of the mandala, the one
found in Tibetan Buddhism.
I will not try to explain this troubling disparity between the two
schools, but hope that future research will bring some light on this
point. My purpose here was simply to bring out the testimony o f the
Saiva texts on the nature and ritual function o f the mandala.
M A N D A L A A N D Y A N T R A IN T H E S ID D H A N T A
177
gatra (4 squares)
pada (3 squares)
dvara (6 squares)
length unit
sobha (4 squares)
square unit (pada)
upasobha (4 squares)
ko n a or astra (6 squares)
4 zones: A = main lotus (6 x 6 = 36 squares)

B = pitha (1 unit wide: 28 squares)
C = v ith i (2 units wide: 80 squares)
D = dvaras + Sobhas + upasobhas + konas (112 squares)
(The entire mandala consists o f 256 squares.)
1. T h e sarvatobhadram andala reconstructed a ccording to the Saradatilaka and R a g h a v a b h a tta s
com m entary
IC O N S O F IN C L U S IV IS M : M A N D A L A S IN S O M E E A R L Y
SAIVA TANTRAS*
Judit Torzsok
Introduction
This study is very much inspired by and indebted to A. Sandersons
excellent article (Sanderson 1986) on the way in which various texts
o f the Trika school o f Saivism encoded their superiority to other
schools in their mandalas. It aims at examining some Saiva mandalas
not examined by Sanderson, most o f which are not based on the
trident image used in the Trika. I shall try to explore h ow these
images represent the relationship o f certain branches o f Saivism with
other Saivas as well as with non-Saivas and h ow these relationships
are visually translated in the image o f the mandala. The discussion
on mandalas as icons o f inclusivism is preceded by a short ter mi
nological investigation and a summary o f some problems concerning
initiation mandalas.
Most o f the texts considered here and consequently the mandalas
they describe date from before the Kashmirian exegetes, i.e., before
the 1Oth-1 1th century A.D. Occasional reference is m a d e to later
texts such as the IsanaSivagurudevapaddhati. Although evidence has
been brought together from various branches o f Saivism, there are a
number o f demonstrably early Tantras that have been omitted from
the discussion.1 Thus, this study does not present a synthesis o f all
*
I w o u ld lik e to thank Paul and G u illa u m e C o a ta len for h a v in g prepared the
m andala illustrations, and I d ed ica te this e s s a y to them. I thank P r o f e s s o r A l e x i s
Sand erson for a printout o f a draft articlc on m and alas he g a v e m e s o m e years ago,
w h ic h I h a v e lost unfortunately and thus ca n n o t cite. 1 h a v e tried to a v o id t o p ic s I
r e m em b e r he d is c u s s c s there in detail and h o p e not to h a v e p la g ia rize d a n y th in g
u n c o n s c io u s ly . I thank P ro fes so r Gudrun B iih n e m a n n for d ra w in g m y attention to
and correctin g a w kw a rd p oints in m y argum ent and style; 1 am f u ll y r e sp o n s ib le for
w h a tev e r rem ains uncorrected, o f course.
1
From the d e m o n s tr a b ly early S idd ha n ta s, t w o important tex ts h a v e not b een
in clu d ed in the d is c u s sio n , altho u g h they contain relev a nt in form ation : the S arva-
180
J UD1T T O R Z S O K
the material one could have access to, but is to be considered the
su mmary o f a work in progress. This, to some extent arbitrary,
choice o f sources means that whatever conclusion is drawn here is
limited and needs to be tested on further evidence. Moreover, the
discussion on mandalas as icons o f inclusivism focuses only on two
texts teaching the worship o f Bhairava: the Svacchandatantra (SvT)
and the Netratantra (NT).
1 Mandalas and Cakras
The Sanskrit term mandala and its several meanings have been an a
lyzed in detail in the Saiva context by Brunner 1986: 13-18 (cf.
Brunner, pp. 156-161), and the word has been subjected to some
analysis in almost everything that has been written on mandalas.
Without reiterating the arguments and all the meanings here, there is
one point which is perhaps not unnecessary to reconsider: the
question o f the difference between the terms mandala and cakra.
Both words have the general meaning o f circle, and thus by e x
tension they can both denote a circle o f deit ies or mantras (which are
the same, since Tantric deities are mantras and spoken o f as such):
devatacakra. That in this meaning the two words are interchangeable
can be shown by a number o f passages, for instance, in the Siddhayogesvarimata,2 in which both terms arc used when the visualization
of a circle o f Yoginis or mothers (m atr) is proscribed.1' But the inter
changeability o f these terms is reflected in more than their use in the
same context. Looking at the description o f the circles o f Yoginis in
the same text, it is somewhat confusing for the reader that in the
same passage, the central deity usually a Bhairava is described as
placed on the pericarp o f a lotus or on the hub o f a wheel, and the
surrounding deities are said to be on the petals o f a lotus or on the
jnanottara, w h o s e full text is a v a ila b le o n ly in m a n u scr ip ts to w h i c h 1 h a v e no
a cccss; and the Kirana, w h o s e on ly edition (D evakottai 1932) is a lso u navailable to
m e at present. For the dating o f early Siddhantas, s e e G o o i l a ll s introduction to his
e d itio n o f the Kiranavrtti, pp. x x x v if f . 1 have a lso om itted m a nd a la s o f t w o texts
tea c h in g m ore cs o tc ric Y a m a la and g o d d e s s worship : the B r a h m a y a m a la and the
Jayadrathayamala. T h e y teach several m andalas, s o m e o f w hich h a v e been d is c u s s e d
in Sanderson 1986. M oreover, no Kaula sources are included,
2 T h e Sid d hayogeSvarrm ata is o n e o f the root-texts o f the K ash m irian Trika,
w h ich I h a v e very tentatively dated to around the seven th century A .D . in T o r zs o k
1999a: vii.
3 S ee, for instance, verses 2 2 .2 3 and 2 8 , 4 0 for mandala and 21.1 for cakra,
ICONS OF INCLUSIVISM
181
spokes o f a wheel. The words describing the circle o f deities as a

wheel or as a lotus are mixed up, showing that what matters here is
simply a circular arrangement: the lotus terminology recalling a typi
cal mandala with the lotus in its centre and the wheel terminology
confirming that the same arrangement can be called a cakra.4
However, in spite o f this confusion o f lotuses and wheels only the
w o r d man dal a is c o m m o n l y used w hen an actual dr aw i n g is
described or referred to in a text, i.e., one following a rather precise
outline and coloured with powders. This confirms what was stated by
Sanderson in a discussion recorded in Padoux 1986: 33: the fact that
mandala normally denotes the locus o f worship ( adhara) and the
cakra [of deities or mantras] is what is located on it (adheya).
Therefore, instructions to trace or draw (likh-) a diagram and to fill it
with coloured powders are given for mandalas, but not for cakr as.5
This is not contradicted by the fact that the circle o f deities is
so m et i m es called m a n d a l a as an alternative, for that ca n be
considered a metaphoric usage as noted above. But the consistent use
o f mandala and the lack o f the word cakra in contexts o f elaborate
drawings show that the former does indeed denote the locus o f
worship, at least in an early stratum of texts.6
There are nevertheless a few examples in which it seems that
instruction is given to draw a cakra. However, in these cases the
cakra is not the full circle o f deities, for what is enjoined is that one
is to draw an actual wheel with a hub, spokes and a circumference.
Such instruction is given, e.g., in S v T 2 9.16ff., prescribing that a
wheel is to be drawn outside the central lotus o f the ma nd ala .7 Thus,
this cakra is an equivalent image o f the lo tu s rather than that o f the
J S e c , for e x a m p le , a d escrip tion starting w ith v e r s e 2 2 . 2 5 m e n tio n in g a lo tus
(padma) and fin ish in g w ith the w h e e l t er m in o lo g y in verses 2 8 - 3 1 .

5 A s noted in the sa m e d is c u ss io n by Sanderson, the t e r m in o lo g y o f the Srividya
is a s p ecia l ca s e, and probably reflects a later and lo o s e r u s e f the terms cakra and
yantra.
6 T h is w a s q u e s tio n e d b y Brunner in the sa m e d is c u s s io n , referrin g to the S v T
and the N T , w h ic h a l l e g e d ly u s e the w ord cakra for the d ra w in g itself. S in c e no
r e f e r e n c e s are g i v e n there, and s in c e 1 m y s e l f h a v e not f o u n d a n y appropriate
e x a m p le s o n ly actual w h e e ls w hich are to be draw n inside a m andala or yantra and
w h ic h are d is cu sse d b e lo w it s e e m s doubtful w heth er such co n fu s io n o f cakra and
m a n d a la is present in th e se texts. N e v e r t h e le s s , no firm c o n c l u s i o n can b e drawn
until all th e se texts are ava ila b le in elec tr o n ic form to facilitate su ch t e r m in o lo g ic a l
searches.
7 T h is p a s s a g e is m en tio n ed as an e x a m p l e fo r the m e a n in g w h e e l in Brunner
1986: 2 0 (cf. Brunner, p. 163).
182
JU D iT TO R ZSO K
mandala, and when its drawing is enjoined, it forms part o f a

mandala, but does not replace it.
A similar idea may underlie the combined lotus-wheel image
described in the Isanasivagiirudevapaddhati, kriyapada 8.10 6-1 23,
which calls it a mandala o f the wheel and the lotus (cakram bujamandala). The text is rather corrupt and appears to give several
alternatives at the same time. However, it is clear from verse 109 that
at least in one o f the versions, there is a lotus in the centre, and the
lines drawn from the centre to the tips o f the petals and to where the
petals are joined are to be lengthened further, outside the lotus, to
form the spokes o f a w h e e l / Here again, one has to draw a wheel
just as one is to draw a lotus, but the result, the whole o f the image,
which includes the passageway outside it ( v lth i), the doors, etc., is
called a mandala.
Tha t the m an d al a is the wh ol e o f the draw ing itself is also
confirmed by the synonyms used for it: b h a v a n a v e s m a n ,w and
pura" words denoting house or abode, i.e., the place where the
deities reside.12 It appears that the mandala is also identified as the
seat (pitha) o f the deity or deities, probably in the sense that it is the
locus o f the deities,13 although this very word also denotes a central
part o f the m andala .14
*
N o t e that it is pointed out in verse 106 that in this c a s e there is no pitha in the
s e n s e g iv e n in A p p e n d i x 1, i.e., the ccntral circular im a g e is not su rrounded b y a
sq ua re-sha p ed scat. T h is is probably b c c a u s e the lotus is surrounded by the w h e e l
itself, w h ich m a y be co n sid ered to rcplacc the scat.
9 M e n t io n e d in Brunner 1986: 25 (cf. Brunner, p. 170) w ithou t e x a m p le s ; s e e ,
e.g., S v T 2 5 ,34d. A sim ilar w ord, bhuvana is used in the NiS vasa, fol. 2 5 v 5 f f ., w h i c h
I understand to be a varia Icctio for bhavana.
10 E.g., S v T 2 5.19.
" S e e , for e x a m p le, S v T2 5 .19.
12 W h i le the first tw o w o r d s m ean a bode or r e s id e n c e in their first s e n s e , pura
d e n o te s a to w n or city as its first m ea n ing . This m a y be the reason w h y Brunner
1986: 25 (cf. Brunner, p. 170) interprets the term to denote larger m a nd a la s (w ith o u t
g i v in g e x a m p le s ) . H o w e v e r , as Brunner remarks, larger m a nd a la s do not s h o w any
sp ecia l relation to city plans or a nything related to towns. Therefore, it s e e m s quite
p o s s i b le that pura is used in the s e n s e a b o d e rather than t o w n in the c o n te x t o f
m andalas, e s p e c ia l l y i f w e co n sid er that as the a b o v e e x a m p le s s h o w , the S v T u s e s
pura as a s y n o n y m o f other d esig n a tio n s w ithout a n y apparent distinction. M o r eo v e r,
S id d h a y o g e ^ v a rlm a ta 2 5 . 8 and Mrgcndra, k riy a p a d a 8 . 2 9 c use the term pura to
d e n o te square m andala s o f 2, 3 or 4 hastas on each side, w h ic h arc d efin ite ly not o f
extraordinary size.
13 T h is is h o w I understand the u se o f the word in a p p o sitio n to m a n d a la in a
citation o f the S idd hayogeSvarlm ata g iv e n by Jayaratha a c / T A 3 1 .8 b . T h e S id d h a yogeS varim ata s e e m s to id entify p;7/iawith mandala in a m c t o n y m i c w a y . N o t e that
ICONS OF INCLUS1VISM
183
In short, the term cakra does not seem to be particularly vague and
its use does not appear particularly inconsistent; it has primary and
secondary as well as metaphoric meanings just as the term mandala.
B u t as far as the terminology of mandalas as more complex images is
concerned, I think it can be safely affirmed that mandala usually
den ot es the whole o f a particular image onto whi ch deities are
placed. Cakra either denotes an actual wheel as part of such drawings
o r refers to the deities themselves. M o re o ve r, cakras are not
neces saril y associated with yant ra s small draw ing s on durable
material including mantra syllables, used as charms in particular. 15
T he y are only associated with yantras inasmuch as mantra-deities or
rather their seed syllables (bija) can be incised in a circular design
(cakra) on these charms.
This short terminological investigation leads us to the question o f
h o w these circles o f mantra-deities are present on a mandala. This
subject, the visualization and placement o f mantra deities on the
mandala, is usually treated as a topic distinct from the drawing o f the
mandala, for indeed the mandala is only one o f the supports onto
which deities can be projected and visualized. Moreover, the way in
wh ich deities are to be seen or meditated upon does not depend on
the support, but on the purpose of the ritual. Th e sa me deity or
deities can be visualized as more frightful for rites to acquire lower
supernatural powers and as milder for appeasement and the like.16
Even if the visualization o f deities can vary considerably for siddhis,
there appears to be a standard visualization for initiation. And in the
c o n te x t o f initiation, it sh ould be r e m e m b e r e d that w h a t the
practitioner o f a ritual is supposed to see in a mandala is not only the
geometric drawing, but the deities placed on it. Consequently, when
texts emphasize how the initiate is impressed by seeing the mandala
for the first time, especially at the time o f his preliminary initiation
(sam ayadiksa), it is not the precision o f the drawing or the beauty o f
the colours that produce this effect, but the fact that the initiate sees
this p a s s a g e is not found in the short recension edited in T o r zs o k 1 9 9 9 a and T o r z so k
forthcoming.
14 S e c A p p e n d ix 1, Illustration 1 and Colour Plates 1 6 - 1 7 .
15 A s s u g g e s t e d b y a s u m m a r y in B r u n n e r 1 9 8 6 : 1 8 - 2 0 ( c f . B r u n n e r pp
161-164).
16 S e e , e .g ., S id d h a y o g c sv a rim a ta 2 8 . 2 9 - 3 0 and 2 2 , 2 7 - 2 9 and 3 8 for the sattvika/

rajasa/tamasa v is u a liz a tio n s o f Y o g in is . S e e a lso various fo r m s o f B h a ira v a , a l o n g
s id e S v acch an dab hairava, such as Kotaraksa, ctc., in S v T 2 9.3ff. and form s o f SadaSiva in Mrgendra, kriyapada 3 . 4 Iff. and com m en tary. S e c a ls o Sand erson 1990: 68.
JUD1T T O R Z S O K
184
the deity or deities.17 This is clear from passages which do no t

prescribe an elaborate and colourful mandala at the time o f this
initiation, but one draw n up quickly with sa nd alw ood p as te
(m andalaka).'K Such a simple mandala surely cannot impress the
initiate by its exceptional beauty. Furthermore, Ksemaraja also
explains ad SvT2 3.128 that when the blindfold o f a new initiate is
removed and he sees the mandala, he is enlightened, and is thus able
to see the d e it y .19 Th e text o f the Tantra itselF suggests the
interpretation that the removal of the blindfold is symbolic, as if it
was the removal o f the darkness o f ignorance. The Tantra also states
that what the initiate sees are the deities.2 Ksemaraja goes on to say
that since this is the first time the initiate sees the Lord af ter
thousands o f births, he is very much surprised, looks at him again
and again and falls on the ground. This shows that he can be
possessed by Siva, for he has discarded the [false] perception o f
identifying his self with the body and so forth [instead o f Siva].21
W h a t is imp orta nt in the above lines o f the Sv T and its
commentary is not simply the fact that the deities are present in the
mandala (which is quite obvious), but that for initiates they are
visible there. Therefore the deities should be considered to form part
o f the visual appearance o f the mandala once they are installed on
it.22 Moreover, it must also be remembered that these deities are
normally not represented by icons because they are too powerful and
consequently too dangerous to be depicted.2' Thus, it is when they
are installed on a mandala that their visual aspect seems to be the
17
T h i s a s p c c t i s w o r t h e m p h a s i z i n g , f o r B r u n n e r 1 9 8 6 : 3 0 ( c f . B r u n n e r , p.
175)
t r i e s t o a r g u e t h a t it c o u l d b e t h e e x t e r n a l a p p e a r a n c e o f t h e m a n d a l a tha t m a k e s it
sp ecia l.
Is S e e , e . g . , S v T 2 3 . 9 0 .
|,J pragavastho yah pasuh sa idanim cva prabuddhah ... punah punar bhagavantam
iksate.
21
... m ukham udghiitya darsayct /
vidyamanfraganaih sardham karanain sasadasivam /
ajnanapatanirmuktah prabuddhah pasur iksa te/
dandavad dharanim gatva pranipatyapunah p u n a h /
21... at a cva janm asahasrapurvabhagavatsvarilpavalokanad vism ayavistah punah
punar bhagavantam iksate I dandavadgamancna dehadipramatrtapahastanat srisivasamavcanusaranc yogyatasya darsila.
n S u c h a n i m a g e , i n c l u d i n g t h e d e i t i e s , i s r e c o n s t r u c t e d in S a n d e r s o n
1986: 187
( t h e d r a w i n g i s r e p r o d u c e d a s I l l u s t r a t i o n 3 in P a d o u x s fir s t c o n t r i b u t i o n t o t h i s
book ).
25 T h i s p o i n t is d i s c u s s e d b r i e f l y b e l o w , in s u b s e c t i o n ii o f s c c t i o n 2 .
185
m o st prominent. It is there that each o f th em is represented in

drawing by his or her distinct place such as a petal o f a lotus, which
serves not only as a locus or support o f worship, but perhaps also as
a crutch for those who perform the elaborate visualization.
The moment o f seeing the mandala with the deities is considered
so important that its mention can metaphorically refer to the whole
ritual o f initiation.24 However, mandalas are not used exclusively at
the time o f initiation, although for some or p os sib ly eve n for
most initiates the ritual o f initiation may actually be the only time
they see a m a n d a l a .25 As pointed out in a n u m b e r o f s tu d ie s ,26
mandalas are mentioned as optional supports for daily worship and
they figure quite prominently in rites to acquire supernatural powers
( sid d h i). No special mandalas are prescribed for regular worship; the
mandalas used in regular worship are just small-size reproductions o f
initiation mandalas. In what follows, focus will be laid on initiation
m an da la s and mandalas used for the acquisition o f supernatural
powers.
2 Initiation M andalas and their R o le
i. The Tw o Initiation M andalas
It has been pointed out in various discussions27 that just as there are
two major parts o f Saiva initiation,28 there are two different mandalas
24 S e e , e . g ., Paratrimika 19 and -la g hu v rtti ad loc. T h e tex t s a y s that e v e n
w it h o u t s e e i n g the m a n d a la , o n e w h o k n o w s the d o ctrin e o f this text b e c o m e s
in itiated. T h e s e e i n g o f the m a n d a la is thus co n sid e r e d the p rin cip a l e l e m e n t o f
initiation, as the c o m m e n ta r y c o n fir m s, sa y in g that it in clu d es rites from the night
spent at the p la ce o f w o rsh ip (adhivasana ) up to the fire ritual, i.e., initiation proper.
T h e S iddhayogeSvarim ata has the c o lo p h o n samayamandala at the en d o f the ch apter
d e sc rib in g the w h o le samaya rite, w h ic h a lso s u g g e s t s that the prin cip al e le m e n t o f
the rite is co n sid ered to be the mandala itself.
25 T h is m a y be the ca s e o f s o m e o n e w h o aspires on ly for liberation and w h o d o e s
not take the trouble to draw up a m andala for d aily w o r sh ip , w h i c h w a s p rob ably a
rather laborious procedure (on this, se e Sanderson 1986: 170, n ote 3). A t s o m e point
this w a s perhaps the m ost c o m m o n categ o ry o f initiates.
26 E .g ., Brunner 1986: 2 0 - 2 1 (cf. Brunner, pp. 1 6 4 - 1 6 5 ) and S a n d er so n 1986:
169-170.
27 S e e , for in s ta n c e, P a d o u x 1986: 3 4 , S a n d e r s o n 1 9 8 6 : 1 9 6 , n o te 12 8 and
Brunner in S P 3 , pp. x x x - x l i i i .
a In s o m e traditions, there is a ls o a third part b e t w e e n w h a t I call h ere the
p r e lim in a r y in itiation {samayadiksa) and in itia tio n p ro p er (diksa). It is t e r m e d
sp ecial prelim inary initiation ( vitesasamayadlksa ), and is fo un d in the S o m a s a m b h u -
186
JU D IT T O R Z S O K
drawn for these occasions. It seems that according to most ritual

manuals, the first mandala, which is drawn for the pr eliminary
initiation or sam ayadiksa, is not at all elaborate. It is made with ou t
co lou re d p o w de r s, including only a basic outline, for w h i c h
sandalwood paste is used. The details of this basic drawing usually
dubbed as gandhamandala (scent m an d al a )29 or mandataka ( small
m andala)31* are not given. However, one can often read detailed
descriptions of what is called the powder mandala (rajom andala).
This is a larger drawing filled with coloured powders, to be used for
what is initiation proper, sometimes also called nirvanadiksa, for it
bestows final liberation.
While manuals seem to share their opinion on these two mandalas
as given above, canonical texts differ on several points. Th e se
differences concern not only the mandalas, but to some extent also
the way in which the two initiation rituals are performed.31
The first or preliminary initiation which is not always na m ed
initiation32 consists mainly o f a symbolic rite o f entry into the
Saiva community. It involves the seeing o f the mandala and ends
with the ann ou nce m ent o f the rules (samaya) the neoph yte is to
observe as a new member o f this community. This initiation can be
termed preliminary initiation because it is a prerequisite to initiation
proper, which follows after a night spent on the grounds o f worship
together with the guru (adhivasa)
Initiation proper is performed next day for those w ho can and
want to receive it. It involves the rites o f purification, deification o f
the body, etc., as well as a rite o f prognostication, which is based on
paddhati as w e l l as in South Indian manuals, A s Brunner points out in S P 3 , p. x x x v ,
it s e e m s to be a 'rite batard containing elem en ts o f the diksa itself,
29 S e c , e .g ., T A 15.387.
3(1S e e , e .g ., S v T 2 3.9 0 c.
31 It w o u l d require a separate study to d is c u s s all the d eta ils and p r o b l e m s
co n cern in g the samayadiksa. T h erefore I shall on ly point out p rob lem s pertain in g to
the use o f mandala s. For a m ore detailed d is c u s s io n , s e e Brunner in S P 3 , pp. x x x f f .
and T o r zso k 1999b.
32 I h a v e s h o w n e l s e w h e r e that this rite is not c a llc d initiation ( diksii) in a
con sid er a b le n um ber o f early texts in clu d in g the S v a y a m b h u v a su tr a sa m g r a h a , the
B r a hm ayam ala, the SiddhayogeSvarTmata and the M alinivijaya. For details o f alter
native ter m in o lo g y , se e T o r zso k 1999b.
33 A n im portant c x c c p t i o n is the VinaSikhatantra, w h i c h k n o w s o n ly o f o n e
initiation, preced ed by the adhiviisa In this text, the rite o f entry u s in g the m andala
is p erfo r m e d on the s a m e day as initiation proper, and the rules (samaya) are
announced o n ly at the end o f the latter (v erse 49).
187
the n eo ph yt es dreams and on the way in which their tooth twigs34

fall on the ground. Then another mandala is prepared and another
variant o f the same kind o f rite o f entry is performed as the day
before, but this time without being followed by the announcement o f
the rules. It is then that the main part o f initiation is done. This is
basically a rather complex ritual o f purification o f the bound so u l
and its detachment from lower levels o f existence. The rite involves
fire offerings at the completion o f which the soul o f the initiand is
joined to Siva.
As even this br ie f su mmar y shows, there is a repetition with
variants o f what appears to be essentially the sa me rite, wh o se
culmination is the seeing o f the mandala. Most texts agree that there
are tw o mandalas and two rites, the first o f which usually gives
peop le the right to perform Saiva worship using the mantras o f the
cult they are initiated into. However, the texts diverge on very
significant details.35
Some texts36 prescribe that the first mandala should be the simple
one drawn with sandalwood paste as mentioned above. W h e n the
initiand is led to it blindfolded, he is to throw a flower or flowers
there. He then sees the mandala, which has a profound effect on him.
The mandala drawn up the following day is larger, m ore elaborate
and decorated with coloured powders, etc. The same person or
persons are led to it blindfolded, but this time they toss one flower on
the dr aw in g each. The nam e o f the deity on w ho s e part o f the
m an da l a the flower falls will form part o f the p e r s o n s initiatory
name, thus suggesting that the initiate was chosen as it were by the
deity who attracted his flower on the empowered diagram.
Some other texts, which may prescribe a smaller mandala for the
first rite, clearly envisage a colour mandala even for the first rite o f
entry. This mandala is probably identical to the second man dal a
except perha ps for its size. More ove r, the guru is instructed to
perfo rm the name-giving as part o f the first rite, while noth in g
particular is said about the second mandala.37
34 T h e s e arc t w ig s used for c l e a n in g o n e s teeth.
35 T h e VTnaSikha differs as pointed out a b o v e, k n o w in g on ly o n e m a nd a la rite.
y' S uch as S v T , chapter 3.
37 T h e S id d h a y o g e S v a r im a ta , for in sta nce, p rescr ib e s a c o lo u r m a n d a la for the
samaya rite as 6 .1 3 s h o w s . B u t the se c o n d m and ala although s e e m i n g l y elaborate,

for it can b e d o u b le the s iz e o f the form er is not d esc rib ed in detail; the reader is
referred to a manual instead (8 .1 1 ).
188
There are yet other versions o f these twin-ritcs. They involve a

sandalwood paste mandala in the first rite, which is nevertheless used
for performing the name-giving.' 11 is tempting to argue'19 that the
second version given above, which prescribes the use o f the colour
mandala and the name-giving in the samaya rite, appears m or e
functional40 and is perhaps the primary version. However, it may be
more important to consider the fact that the first version o f the rites
appears in the SvT, a text which is relatively old a m o n g the
demonstrably early Tantras.
W i t h o u t t ry in g to esta bli sh w hi ch or de r and m e t h o d o f
performance o f the twin-rites is primary, it can be stated that the two
rites re semble each other very muc h and thus may reflect; the
doubling o f a ritual which was originally o n e ,41 Moreover, since
there is an early Tantra, the VinaSikha, which knows only o f one
combined initiation ending with the announcement o f sam ayas, it
may represent a very early stage o f development, when even the
samaya rite was not yet independent or separable.42 Judging from the
stage represented by this text, il seems a possible development that
first the mandala rite and initiation proper became separated, which
was followed by the doubling o f the mandala rite. An additional
ar gument for the theory o f doubling could be that as some later
developments show, a further extension o f the preliminary initiation
occurs43 unde r the name o f visesasamayadiksa. To this a new repe
tition o f the mandala rite is added by AghoraSivas commentator,
Nirmalamani, who prescribes it if too much time has el lapsed since
3,1 A s in S P 3 , pp. 37 and 103.

w In T o r z s o k 1999b I tried to establish a p o ssib le ch ro n o lo g y and d e v e l o p m e n t o f
this rite in various texts, but I am n o longer sure i f there is en o u g h j u stifica tio n for
all m y h y p o th e ses. A n attempt at reconstruction has been m ade by Brunner in S P 3 ,
pp. x x x i - x x x i i i in a m o re general way, om itting details o f the mandala.
4(1 I.e., it s e e m s to be m o re appropriate to draw up an elaborate m andala for a rite
w hich f o c u s e s on this instrument. Furthermore, it co u ld be argued that the s e e in g o f
the u ncolourcd m andala in the S v T , w ith o u t the n a m e -g iv i n g ritual, appears to be
odd. W h y sh o u ld o n e be im pressed by the s a n d a lw o o d paste mandala rather than by
the c o lo u r e d m andala? What is the point here in c a s tin g flo w e r s on the d iagram i f
the n a m e - g iv i n g is om itted? For a d isc u s sio n o f t h e se p ro b lem s, s e e Arraj 1988:
144ff. and Torzsok 1999b.
41 For this h y p o th e s is and argum ents in the ca se o f the Siddhanta, s e c Brunner in
S P 3 , pp. x x x i - x x x i i i .
42 T h is co n firm s Brun ners h y p o th e sis in S P 3 , pp. xxxi.
43 T h is is the ca se, for e x a m p le , in the Som aS am b hu pad d hati and the A g h o r a 3ivapaddhati.
189
th e performance o f the samaya ritual/ '1 Thus, the samaya rite seems
to be particularly liable to doubling and expansion.
In spite o f these possibilities, it must be reiterated that no early
Ta ntr a apart from the very brief Vinasikha seems to refer only to one
m a n d a l a .45 If a redoubling took place at some point, we have no way
o f knowing how exactly it happened and through what stages.
This mean s that the role o f initiation m a n d a l a s can no t be
determined in general by covering the descriptions and versions o f
all early texts. However, a number o f significant points can be
summarized concerning their nature and importance in this ritual.46
1.
2.
It is the mandala either in its simpler or its more colourful and

elaborate version and the ritual connected to it that usually
determine the initiation names o f initiates and thus not only
qualify them to use the mantras o f the cult, but also give them a
new identity in their Saiva community. The initiate comes to be
chosen as it were by one particular deity o f that tradition, who
attracts the flower he casts on the mandala.47
A ft er the d is ci p le s blindfold is removed, the seeing o f the
mandala no matter which kind is used with the deities on it is
usually considered to have a profound effect on him, for this is
" S e c AghoraSivapaddhati 254.

45 It m ust a lso b e added that the VTnaSikha d o e s not m e n tio n the n a m e - g iv i n g
itself, but the editor o f the text, T. Goudriaan, s u p p o se s that it w a s u nd erstood (cf. p.
139 and p. 16).
4,1
In w hat f o llo w s , I a s s u m e that the m ore elaborate m a n d a la is a larger and
e x p a n d e d v e r s io n o f the c o l o u r le s s or sm a lle r one. T h is is s u g g e s t e d in all tex ts
w h i c h d e s c r ib e on ly o n e o f th em . T o m y k n o w le d g e , the o n ly tex t a c c o r d in g to
w h ic h the t w o m a nd a la s arc e x p lic it ly very different is the T A . But that e x c e p tio n a l
c h o i c e is probably m otivated b y A b h in a v a g u p ta s intent to integrate v a rio u s le v e ls
o f the Trik a in a hierarchy o f initiations. For m ore in fo rm a tio n on this to p ic , s e e
Sand erson 1986: 196,
47
W h ile texts contradict ea ch other on w heth er the elaborate c o lo u r m and ala or
th e s im p le r g a n d h a m a n d a la is u sed fo r this rite, o n e c o u ld p erh a p s fin d m o r e
e v i d e n c e b y e x a m in in g initiation n a m es. From initiation n a m e s fo un d in in scriptions
in D arasuram , it s e e m s that n a m e s o f the V idyeSvaras, aiigamantra s and mantras o f
the throne w e r e used as w e ll as the brahmamantras (s e c S rin iv a sa n 1 9 8 7 , and the
e v i d e n c e su m m a rize d in G o o d a ll 2 0 0 0 : 2 0 7 ) . T h e n a m e o f S o m a S a m b h u m a y s h o w
that e v e n lokapiila s w er e perhaps in clu d ed in the initiation m and ala, w h ich w a s then
pro b a bly a m o re ela bo ra te on e. But s in c e the a b o v e in scrip tion al e v i d e n c e c o m e s
fro m the tradition o f the Siddhanta in the C o la co u n try o f the 12th centu ry , its
t es tim o n y cannot be taken for what h a ppened in other reg io n s and periods.
190
3.
4.
his first contact with the deities o f his chosen s ch oo l.'18 In the
daily rites which he is obliged to perform ever after the initiation,
the disciple is in fact supposed to recreate this first sight o f the
deities in visualizations.'19
The seeing of the deities in the course o f initiation is not the
privilege o f some, but is experienced by all categories o f
initiates. In a number o f texts, the name-giving is also performed
for all candidates as part of the initiation.
While there are Vedic parallels to initiation proper (diksa) as a
whole, the central part o f the samaya rite perfor me d at the
mandala has no such obvious Vedic predecessor.50
ii. The Initiation M andala as the Largest D etailed R epresentation o f

E soteric D eities
The paragraphs above do little more than summarize w h a t has
already been analyzed in detail in Sanderson 1986 concerning the
Trika, namely the fact that the mandala plays a particularly important
role in creating and maintaining a new, Saiva and sectarian, identity
o f the initiate. This identity is then repeatedly confirmed in the
course o f the performance o f daily worship. This must be one o f the
reasons w h y the m om en t o f seeing the mandala is considered so
important. But in addition to this, there may be yet another reason. It
is not mentioned or expressed explicitly, but is perhaps still an
important factor here, at least as far as early texts and practices o f
Bhairavatantras51 are concerned. The initiation mandala, in addition
4S
Even i f o n ly the n a m e -g iv in g versio n is describ ed for the samaya rite, as in the
Tantrasadbhava, for e x a m p le, it is m ade clear that the initiate is duly im pressed and
fa lls on his knees. S e e 9 , 124d, w here the subjcct must be the initiate or the initiate
w ith the guru.
m For this p ro ce ss in the Trika, s e e Sanderson 1986: 1 6 9 - 1 7 0 .
Sl Brun ner in SP3, pp. x x x v i draw s a parallel b e t w e e n the samayadiksa and the
upanayana rather than b e t w e e n diksa proper and V e d i c rites. H o w e v e r , w h a t is
sim ilar in the upanayana and in the samayadlksii is not the nature but the functio n o f
the tw o rites, for both b e s t o w the qualific ation to study the scriptures. M o reo v er, il is
o n l y th e visesasamayadiksa that cr ea tes a t w i c e -b o r n in the s a m e w a y as the
upanayana d o e s , and this is not c o m m o n to all v e r s io n s o f the sam ayadlksii. M y
point in draw ing a parallel b etw e en diksii proper and V ed ic rites is that fire ritual has
o b v io u s V e d ic pred ecessors, w h ile the rile in v o lv in g the mandala is rather u niq u e to
the Tantric context,
51
T h e term is used here for Tantras teachin g the w o rsh ip o f B h airava as w e l l as
for Y a m a la ta n tr a s and Tantras te a c h in g g o d d e s s w o rs h ip . For d eta ils o f t h e s e
191
to being the paradigmatic image, is probably also the largest or one

o f the largest images representing esoteric deities in detail which is
u s e d in communal worship (in the sense that several people use it,
b u t not at the same time). This point requires a brie f investigation
into the question o f what objects were used for the wo rsh ip o f
esoteric deities and how: the nature o f worship and what substrates it
m a y require, the role and scope o f lihga worship, the question o f
an th r o p o m o r p h i c images and what size various substrates were
prescribed to have.52
M o st Bhairavatantric ritual prescriptions envisage that ritual is
per formed in an abandoned place specially prepared for this purpose,
and not in a permanent building with permanently installed images in
it. This suggests that at least in the case o f some esoteric cults (in the
Bhairavatantras and abov e5) and at a relatively early period, near the
composition o f the earliest Tantras, no permanent building or image
w a s used to perform ritual.53 It must also be noted that the list o f
supports for daily worship given by Abhinavagupta54 mostly includes
va r io u s small objects as supports for visualization w h ic h do not
actually depict the deities o f the cult. He mentions, for instance, a
rosary, a mirror or a sword-blade as well as a private lihga. W he n
im ag es are mentioned, they are small ones made o f painted clay
(p erh aps what is meant is terra-cotta), deod ar w o o d or gold or
images painted on a piece o f cloth or drawn on a skull. They are
images o f small size for private worship, never larger ones mad e o f
stone. More ov er , early Bhairavatantras do not no rm all y include
r ef er enc es to rituals w hi ch e m p o w e r icons m ad e for c o m m o n
worship (pratistha).55
ca te g o r ie s , s e e S a nd erso n 1988. For the fact that Tantras tea c h in g g o d d e s s w o rsh ip
a l s o c a t e g o r iz e t h e m s e lv e s as Bhairavatantras in a broader se n se , s e e , e . g ., Sidd hayo g eS v a rim a ta 1,19d and 8 .4 cd , the latter p a ssa g e reading mahabbairavatantre sm in
s id d hnyng cs v a rim ate.

52 A n appropriate treatment o f the subject w o u ld require a m o n o g r a p h . T h erefo re,
w h a t is presen ted b e l o w is o n ly a b r ie f su m m a r y o f a f e w p o in t s r e le v a n t to the
presen t d is c u s s io n on m and alas in s o m e early S a iv a Tantras, w ith o u t a full p r e s e n
tation o f all the e v id e n c e . It is hoped that the study Id ols and Other S u b str a te s o f
W o rs h ip in the T rik a annou nced in Sand erson 1 9 9 0 shall b e so o n available.
53 T h is h y p o t h e s is w o u l d o f co u rs e n eed further in v e stig a tio n and a full p r e s e n
tation o f the early s o u rc es on the subject.
54 S e e T A 2 6 . 3 2 f f . and Tantrasara 1 7 9 - 1 8 9 and the su m m a ry o f th e se p a s s a g e s in
S a nd erso n 1986: 170.
55 T h e o n l y s u c h text I k n o w o f is th e u n e d it e d P ih g a la m a t a referred to in
S a n d e r s o n 1 9 9 0 : 4 0 and citcd on the v is u a liz a tio n o f the d e it ie s o f the Trika. Its
192
It must be mentioned that in the demonstrably early sources o f the

Siddhanta, even if the installation o f deities in permanent, durable
images is co m m on ly discussed, these images were not used for
public rites. They were used only by members o f the par ticular
community (inatha) for individual worship. As Brunner observes in
her study and translation o f the pratistha section o f the SomaSambhupaddhati (SP4, p. v), at the time o f the writing o f this manual, no
public temple rites were per formed,56 What is envisaged by So m asambhu is that when a linga is established in a matha, for example, it
is worshipped by several people one after the other.57
Private, portable (cala) lingas were also used in more esoteric
cults, as mentioned above, even if the worship o f larger, sha red
lingas, do es not seem to be m e n ti on ed in B h a i r a v a t a n t r a s , 5K
Moreover, as Abhinavagupta writes in Tantraloka (TA) 27.2-3, these
shared lingas even if envisaged for Bhairavalantric wor ship are
not to be installed with secret, i.e., esotcric, mantras; for those
mantra-deities possess their power in their esoteric nature and once
N e p a l e s e m anuscrip t is reported to be dated A .D . 1 1 6 9 - 1 1 7 0 . T h e s a m e title o cc u r s
in the list o f the Bhairava canon as citcd by Jayaratha from the Srik anthlya ad T A
1.17, but it is not citcd by A b h in a v a g u p ta or Jayaratha h im s elf. A lt h o u g h the text
cla im s to b e part o f the Bhair ava canon as Goudriaan 1981: 4 6 p oints out, it m a in l y
d eals w ith te m p le construction and installation o f lingas. G oudria an 1981: 4 6 a ls o
o b s e r v e s that Sakta te n d e n c ie s arc alm ost c o m p le t e l y ab sen t in this w ork, w h i c h
w o u ld ex p la in w h y a Saiddhftntika author, Vidyakantlia II (pupil o f Ramakantha II),
refers to it m a n y tim es in his Bha v a cu da m a n i (for details, s e c G o o d a l l s introductio n
to his edition o f the Kiranavrtti, pp. x x v i - x x v i i ) . From the a b o v e citcd a c c o u n t s o f
the co n ten ts o f this text, it s e e m s that o n ly its ch aptcr 5 on p ainting (and p o s s i b l y
s o m e p a s s a g e s in its chaptcr 4 on ic o n s ) co u ld b e relevant in a B h a ir a v a la n tric
co ntext. T h e s e p a s s a g e s sccn i to g iv e the ico no g ra ph y o f paintings m a d e on cloth, a
substrate for private w o rsh ip m e n tio n e d by A bh in a v a g up ta in the a b o v e cita tio n s.
T h e N T a lso g i v e s s o m e details in a f e w verses, w hich arc d is cu sse d b elo w .
* O n c o f the f e w , relatively detailed, surviving p a s s a g e s about the in stallation o f
linga s and sm a ll s iz e statu es is fo un d in the Mataiiga, kriyiipiida 1 3 - 1 4 , the o n l y
lo n g er p a s s a g e pointed out in the a b o v e ed itio n as a parallel. For s o m e ad dition al
texts in m anuscript form w h ic h also deal w ith the subjcct, sec, e .g ., N U v a sa ta n tra ,
G uhyasutra, chaptcr 2 and Sarvajnanottara su m m a rized in G oudriaan 1981: 3 6 , 39.
S e e a lso t w o Pratisthatantras, (he MohaSurottara and the M ayasam graha d escrib ed in
G o o d a l l s introduction to his editio n o f the Kiranavrtti, pp. x - x i and referred to as
s ou rces o f S o m a S a m b h u s account in G o o d a ll 2 0 0 0 : 216.
57 T h is is a lso what A bh in avagup ta refers to in T A 2 7 . 5 3 - 5 4 .
5 T h is lack o f interest in c o m m o n l y w o rsh ip ped large lingas m a y also b e related
to the fact that B hairavatantras s e e m to f o c u s on the acq u isitio n o f supernatural
p o w e r s , w h ic h requires private rituals in se c lu d e d p la c e s in m o st c a s e s , i.e., rites
p erform ed w ith o u t o n e s c o m p a n i o n s . T h e solitary p erform an ce o f these rituals is
enjoined, e.g., in SiddhayogcSvarim ata 12,14, 13.11, 18.18 and 19.17.
IC O N S O F IN C L U SIV ISM
193
installed, they would lose their real nature as well as their power.
Therefore, stable or larger lihgas are to be installed with the mantras
o f the Siddhanta, even if other deities can be in vo ked in th em
temporarily. Furthermore, following the Sarvajnanottara, Abhinavagu p t a adds that secret mantras should be avoided especially in case
o n e installs a so-called manifest ( vyaktarupin) image-a wa rn ing
w h i c h shows that what is to be avoided here is first o f all an
anthropomorphic or figurative image.59
N e v e r th e l e s s , there w as one, u n d ou b te dl y a n th r o p o m o r p h i c
i m a g e which was not excluded from communal esoteric worship
a n d even recommended for certain days: the body o f the guru, that o f
o th er Saivas and certain women. In the list o f 11 possible substrates
o f external worship given in TA 6 .3,60 the last one, m u rti, a word that
co ul d possibly refer to an icon in similar contexts, is glossed by the
co m m e n ta t o r as the body belonging to the guru or others (m urtir
gurvadisam bandhini). The long description o f what is called The
W o r s h i p o f E m b o d i m e n t s (m urtiyaga) or Th e W o r s h ip o f the
C i r c l e (cakrayaga), which makes this s om ew hat enigmatic gloss
clearer, is then given in chapter 28.60ff. by Abhinavagupta: it is a
rite in which the preceptor, various other practitioners, their wives
w T h e ex c e p t io n a l installation o f a B h a ir a v a g a m ic mantra in N T j 1 8 , 1 1 9 - 1 2 1 is
d i s c u s s e d b y S a n d er so n 1990: 78, w h o understands the p a s s a g e o f the Tantra to
p res cr ib e the installation o f an c c ty p c o f the es o teric mantra o f S v a cch a n d a b h a ir a v a
in the ic o n . H e then argu es that A b h in a v a g u p ta in T A 2 7 . 8 in fa ct c o n to rts the
in te n d ed m ea n in g o f the Tantra b y interpreting the p a ssa g e to refer to the in stallation
o f a n o n - B h a ir a v a g a m ic mantra su ch as that o f Netranath a, so that the prescription
s h o u ld c o n f o r m to the fu n da m en ta l rules o f pratistha. A lthou gh A b h in a v a g u p t a s
interpretation d o e s se em forccd, it must be noted that the p a ss a g e o f the Tantra it s e l f
r e fe r s b a ck (b y s a y in g pragvidhanatah) to s o m e p r e v io u s v e r s e s o n g en er a l ru les
a b o ut pratistha. T h e s e v er se s, 1 8 . 1 0 4 c 109, p res cr ib e the in sta lla tio n o f the n o n
e s o t e r i c A m rteSa/N etran ath a a lo n e or w ith the a ls o n o n - e s o t e r ic g o d d e s s M a h a la k s m i. E v e n i f this is not e n o u g h to support A b h in a v a g u p t a s interpretation, it is
s ig n ific a n t that installation is g en er a lly e n v is a g e d here u sin g n o n - e s o t e r ic mantras,
in sp ite o f the general te n d e n c y o f the N T to m ix up v a rio u s A g a m i c p rescriptio n s
(fo r w h ic h se e N T 2 1 3 . 4 5 - 4 6 ) .
H) T h e s e substrates are a m andala, le v e le d ground, a v e s s e l, a rosary, a m anuscrip t
[ o f 6 a iv a scripture], a lingo, a skull, a p ie c e o f cloth (n o details g iv e n ) , a clay/terra
cotta im a g e (not d is cu ssc d in a n y detail, but m en tio n ed in 2 7 . 1 9 as c o lo u r e d [ citra ]),
a mirror (or a n y m irror-like su rface, su ch as a s w o r d m e n tio n e d in 2 7 . 4 4 ) and a
murti:
mandalam sthandilam piitram aksasutram sapustakam /
lihgam turam patah pustam pratima m urtir cva ca //
194
JUDIT TORZSOK
and women of lower castes or prostitutes are propitiated by alcohol

and offerings which include meat and fish/ 1
Even when an apparently figurative image or an th r o p o m o r p h i c
icon seems to be mentioned in a Bhairavatantra, such as a Daksinam u r t i in the SvT, commentators understand such references as
denoting something which is different from an icon as an e m b o d i
m en t .62
As for the size o f images used, icons o f deities in the Sid dh an ta
are described, e.g., in the Matahga, kriya p a d a , chapter 14 as be i n g
between ten ahgulas, and one hast a (0.2-0.45 metre)/1 The size o f a
linga is said to be three hastas (1.35 metre) in the same text (13.9),64
which would be the size o f the smallest mandalas.
The initiation mandala used by all the in it iands is thus the largest
image (as envisaged in a number of Bhairavatantras) or one o f the
largest images (if we consider shared lingas) a Tantrika may see and
use, for its side usually measures at least three or four hastas (eight
or nine is also recommended), that is at least 1.351.8 metres. But
unlike the other communal or shared support o f a relatively large
size, the linga, it contains a clear visual mapping o f the esoteric
deities o f o n e s tradition.
A m o n g objects used as supports for the wor shi p o f esoteric
deities, por tab le images used for private w o r sh i p are small.
Therefore, even if kept over a longer period o f time, they can be
hidden from the uninitiated. Other supports such as a mirror or a
sword are not easily recognizable as religious objects. But an image
as large as an initiation mandala is not so easy to hide, and this is
perhaps one o f the reasons why it has to be temporary: a m an da la
drawn up for the occasion and effaced ritually after it has served its
purpose.
N o t e that a ccording to A bh in a v a g up ta his sum m ary o f this yaga is b ased on the

S id d h a y o g e S v a r im a ta . H o w e v e r , the short rc ce n s io n o f that text d o c s n o t c o n ta in
a n y th in g on this particula r subject. For m o re in fo r m a tio n , s e e T o r z s o k 1 9 9 9 a :
229-230.
S e c K scm araja ad S v T 2 3 . 1 2 9 , w h o rem arks on daksiniim m urtim : na tu
pasavlm dcham ayim . H o w e v e r , it is not clear to m e what im age K scmaraja had in
mind.
fLl O n e hasta is the d istance b etw e en the tip o f o n e s m id d le finger and the e l b o w ,
I h a v e taken o n e hasta to be equal to at least 0,45 metre and rounded up the figures.
M S o m a S a m b h u e n v i s a g e s lingas o f up to n in e hastas, but it is q u e s ti o n a b l e i f
su ch large lingas w ere c o m m o n or i f they w ere in use at an earlier date as w ell.
195
3 T he Inclusion o f L o w e r Revelation in the M andala from the Vedas

to the Siddhanta: The Case o f the Svacchandatantra
T h e wa y in which the trident image o f the mandalas o f the Trika
enc ode s the supremacy o f that school has been analyzed in detail by
Sand erson 1 9 8 6 He points out firstly, how scriptural sources repre
s ent the superiority o f their system by raising the throne o f their
de ities higher, and secondly, how A b h i n a v a g u p t a s exegesis adds
f u rt he r hierarchies by stretching what is implied in the scriptural
s o u rc es o f the Trika and related schools. Thus, w he n the trident
image, which includes the full cosmic hierarchy from earth up to the
three goddesses on the tips o f the trident, is installed in the line of
inn er sensation in regular worship, it reveals the Tr ik as supremacy
b y taking [the practitioner] through and beyond the mandala-thrones
o f all other Saiva claimants to the worship and assimilation o f
absolute power.
In the same article, reference is made to the ra nking o f the
doctrines o f outsiders, i.e., non-Saivas, in scriptural sources as well
as in the Kashmirian exegetical literature.66 In what follows, I shall
take up this line o f inquiry with special reference to the SvT, which
gives a particularly detailed account of its relation to other doctrines,
in or der to illustrate the following two features o f its initiation
mandala.
1.
Whi le the trident image o f the Trika creates its hierarchy and
encodes its supremacy to others in a vertical ascent, the SvT as
well as a n u m b e r o f other texts and their m a nd al as us e a
concentric image and ar ran gem ent o f deities to express their
domination:67 they place the supreme deity o f their system in the
m id dl e o f the mandala, s u rro u n de d b y o th e r deities often
representing other schools o f thought. This method o f concentric
enco din g seems to be more c o m m o n than that o f the trident
image o f the Trika, whose mandala is in fact quite exceptional in
that it is to be seen as three-dimensional, building its central
trident upon the usual concentric image o f other Tantras. For in
the trident mandala, the central lotus is not the seat o f the
w For an illustration o f o n e o f the sev era l v e r s io n s o f the trident m a n d a la o f the

Trika, s e e C o lo u r Plate 19.
S e e Sanderson 1986: 172, e s p e c ia lly note 8.
67 T h is idea is also referred to brie fly in Sanderson 1986: 172,
196
2.
principal deity but is the lotus of gnosis, from which the trident
o f the three goddesses arises and is seen as coming out o f th e
surface o f the mandala.
The exam pl e o f the SvT also shows that the ma nd al a c a n
visually represent and include not just other brandies of Saivism,
but also non-Saiva doctrines or traditions in the form o f l o w e r
revelation. In this respect, the SvT is a special case, because it
seems to be the only Bhairavatantra to include a re latively
de tai led di scu ss io n o f other, non-&aiva and early S a i v a
(Pasupata, etc.), doctrines and to include them in its c o s m ic
hierarchy.
T he SvT ranks the doctrines o f outsiders in its 11th c h a p t e r

( 11.6 8 ff.), which seems to have beco me som eth in g o f a lo c u s
classicus on the subject later on, judging from the series o f citations
given by Jayaratha ad TA 1.33.fiS In this passage o f the T a nt r a,
schools o f thought are assigned various levels o f the universe or
principles ( tattva), from intellect ( b u d d h i) to Sadaiva. The e q u i
valences with the tattvas are not explained in a fully systematic way,
for while some tattvas are not assigned to any school,69 others are
said to represent the level o f liberation or consciousness o f several
schools at the same ti m e .70 In addition, there are also principles
which are not tattvas but are nevertheless said to be the place o f
certain schools.71 It should be noted that these inconsistencies m a y
be due to the fact that it is not uncommon in early Tantras that the
number o f ta ttv as fluctuates. Such fluctuations can be explained on
the one hand as a result o f redactional cutting-and-pasting, on the
other by the fact that it was perhaps not felt to be necessary to fix the
number o f tattvas at an earl ier stage o f doctrinal development.72
m T h e c i t a t i o n s arc not i d e n ti fie d in the ed it io n . Jayaratha q u o t e s S v T 2
1 1.6 8 c 71 d.
N o n e o f the f iv e c o v e r i n g s (kancuka) is m en tio n ed e x p lic it ly in the list, nor is
the lev el o f pure k n o w le d g e (guddhavidya) a bove maya.
711 T w o sects, the M ausulas and the Karukas arc both g iv e n the 30th le v e l, that o f
mayfr, and the level o f is vara also represents several sccts.
71 T h u s , the Jainas arc said to be e s ta b lis h e d in the three strands o f m a teria l
e x i s te n c e ( guna ), w hich d o not form a tattva. H o w ev e r, guna is s o m e t im e s listed a s a
tattva, such as in the Parakhya recorded in G o o d a l l s introduction to his ed itio n o f
the Kiranavrtti, pp. liii.
71 For a detailed d is c u s s io n on the number o f tattvas in the Siddhanta and w hat
they m a y im p ly , s e c G o o d a l l s in troductio n to his ed itio n o f the Kiranavrtti, pp.
li-lv .
'
'
197
In spite o f these inconsistencies, a n um ber o f elements o f the

hi e r a rc h y are very clearly defined and some o f them correspond to
o t h e r ranking systems. Thus, while Ksemaraja adopts a different
hier arc hy in his Pratyabhijnahrdaya (ad sutra 8), he also assigns the
level o f intellect ( buddhi) to the Buddhists.
In the ranking o f the SvT, non-Brahminical schools are placed the
lo w es t, below the 24th level, that o f material cause (prakrti): the
Buddhis ts are made to reside in buddhi and the Jainas are at the level
o f the three gunas o f sattva, rajas and tamas. N o w it could be said
t ha t the gunas and material cause are at the same level, for the gunas
are the three strands o f prakrti. However, in this passage, they seem
to be treated as separate from and inferior to prakrti.73 It should also
b e noted that in verse 68ab, buddhi itself is said to be produced from
t h e three gunas, suggesting a direct relationship between these two
principles and possibly between the schools placed at these levels.7'1
Th e 24th principle, prakrti, is assigned to Promulgators o f the
V e d a , which is interpreted to allude to the Vedanta by Ksemaraja ad
Joe.,75 but could just refer to anyone for w h o m the V ed a is the
hig hes t revelation. The 25th principle, purusa, is the highest reality
for the Samkhya, which seems quite natural.76 Above them, the 26th
principle is the highest level for the Yoga. This is normally the level
o f niya ti,11 causal determination, but in this passage, the SvT does
n o t speak explicitly about any o f the five coverings, of which niyati
is normally the first.
This exposition is followed by the placement o f various schools of
th e Saiva Atimarga, schools o f Paiupatas and Lakulas, in the cosmic
hierarchy. First, the pasupata-vrata is equated with the 33rd level o f
Isvara. Following Ksemaraja, this expression refers to the doctrine
exp ou n de d by Lakullsa. Then the text returns to a lower level, the
30th. It is the level o f m a ya , which is the highest level for Mausulas
73 In the s a m e w a y , (he Kiranatantra also lists the gunas b e l o w prakrti, as reported
in G o o d a l l s introduction to his edition o f the Kiranavrtti, pp. lv,
74 N o t e that B u d d h is t s c h o o l s arc d istin g u ish ed , but in fact all are p la c c d at the
le v e l o f m a y a in T A 4 . 2 9 - 3 0 . K scm a ra ja puts the S a u g a t a s t o g e t h e r w ith the
M lm a m s a k a s , N a iy a y ik a s and Carvakas at the le v el o f buddhi, w h i le the V ed a n tin s
and M a d h y a m a k a s arc a b o v e them.
75 T h is can be inferred from the fact that he citcs the SvctaSvatara-Upanisad.
76 H o w e v e r , K s c m a r a ja u p g r a d e s t h e m to the l e v e l o f m aham aya in
Pratyabhijnahrdaya.
the
J 1That Y o g a must be then at the level o f niyati is a ls o co n fir m e d b y Jayaratha ad

T A 1.33: tesam [i.e., patanjalanam ] pum stattvordhvavartiniyatitattvapraptir ukta.
198
JUDIT TORZSOK
and Karukas, who arc followers o f disciples o f LakulTSa, ac co rd in g

to Ksemaraja. They are said lo identify this level with the deities
KsemeSa and Brahmasvfunin respectively. It is then stated t h a t
Vaimala and Pramana (or Pancartha) Pasupalas can rcacli up to the
level o f Tsvara, identified with their highest deities, Teje sa a nd
Dhruvesa. After this, aivas probably in the general sense meaning
Saiddhantikas as well as those of other currents arc mentioned and
declared to be above the rest.
Further in the same chapter, another passage discusses doctrines
o f other schools, this lime without ranking them in an un ambi guous
way. The categorization of other doctrines is based on their relation
to dharmal adharma, detachment/lack o f dclachmcnt (vairagyafavairagya), knowledge/ignorance (jhana/ajhana) and powerfulness/lack
o f power (aisvtuya! anai&varya), According to verse 186, these eight
concepts m ake the wheel o f the samsara turn round incessantly as
eight spokes. In this passage, non-âiva and other &aiva schools are
hierarchized in the following way. Verses 17 4- 179b describe trea
tises o f logic (heiusasira) and declare them to be characterized by
adharma, lack o f detachment, ignorance and lack o f power. It seems
they receive the lowest grade here; for, as the SvT says: they are
devoid o f knowledge, Yoga and deities, and arc useless for th e
attainment o f any of the four goals o f men in life.7" After this, all the
other schools are described mentioning at least one good point about
their teachings. Mundane or common knowledge' covering ag r i
culture, politics, etc., as Ksemaraja points out is characterized by
dharma, while the doctrines o f both the Pancaratrikas and the Vaidikas involve dharma as well as knowledge. Buddhist as well as Jaina
doctrine is endowed with detachment, while the school o f Sa m kh y a
possesses both detachment and knowledge. The bcst-placed o f the
non-Saivas is again the doctrine o f Yoga, which is associated with
kn o wl edg e, de tac hm ent and power at the same time. Th e only
doctrine exhibiting all the good characteristics, and which thus goes
beyond (ati-) the others, is the Saiva doctrine o f the Atimarga. For
78 T h is particularly lo w ranking o f the s c ic n c c o f lo g ic is not f o llo w e d b y the
e x c g e t c s . Ksemaraja p la ce s the N y a y a in the sa m e group as the M lm a m s a and the
B u d d h ists both in the Pratyabhijiiahrdaya and in Spandanirnaya ad 4 and 1 2 - 1 3 . In
the latter work, h e p la cc s e v e n the S a m k h y a and the V edanta at the s a m e le v e l. It
c o u l d be a r g u e d that Hetuastra is not the s a m e as the s c h o o l o f N y a y a .
N e v e r t h e le s s , in general statem ents o f the a b o v e kind, (hey m ay not be differentiated
in a very precisc manner.
ICONS OF INCLUS1VISM
199
adherents o f the Atimarga, i.e., those who practice the Skull obser
va nc e, and the Pasupatas (verse 184), there is no further creation,
th ey are established in Isvara/DhruveSa. The ranking stops here, and
no other Saivas are mentioned.
Th is way o f ranking o f other doctrines reveals two important
d i s t i n g u i s h i n g features o f the SvT. On e is that it includes all
Brahmi nica l schools o f thoughts from the level o f prakrti upwards.
N o w , it may be argued that the Pancaratrikas are omitted from the
hierarchy o f levels. However, it is arguable that they are understood
n ex t to the Vaidikas, which is demonstrated in two other passages.
O n e is the verse referred to above, which states that the doctrine o f
b o t h Vaidikas and Pancaratrikas is characterized by d h a rm a and
knowledge. Another passage (5.44- 46) prescribes that one should
not condemn Bhairava, his and other Sastras, the latter including the
S am khy a, Yoga, Pancaratra and the Vedas, for they have all come
forth from Siva h i m s e l f and bes tow li bera tion.79 The se passages
s h o w that the Vaidikas, Pancaratrikas and the adherents o f the
S am kh y a and Yoga are all consciously felt to be very closely related
to Saiva doctrine, and consequently they are placed at the level o f
p ra krti and ab o v e .80
Secondly, the SvT also gives a particularly detailed account o f
h o w it sees itself in relation to the Atimarga. Judging from the
n u m b e r o f Pasupata branches, they must have been flourishing or
r eco gn ize d at the time o f the redaction o f this Tantra. It is also
noteworthy that the SvT particularly insists on the superiority o f the
A t i m a r g a , l e a v i n g the d o c t r i n e a b o u t the s u p e r i o r i t y o f
Bhairavatantras vis-a-vis other Saivas implicit. It sees itself as the
continuation o f the Atimarga rather than o f the Siddhanta.81
The whole issue o f ranking other doctrines according to the tattvas
gains particular significance in connection with the worship o f the
deities on the mandala. The drawing o f the mandala o f nine lotuses
w K scm araja reports a reading fro m old m a n u s c r ip ts , w h i c h states in the last
lin e that ev e ry th in g c o m e s forth from &iva and b e s t o w s the fruit o f [reaching] S i v a s
abode.
80 A l t h o u g h doctrinal a ffin itie s rem ain im portant, o th e r tex ts d o n o t state this
relatio nship so ex p licitly .
81 K semaraja ( c o m m e n t i n g on 1 1 .1 8 4 c ) s e e m s s o m e w h a t u n c o m f o r t a b l e w ith the
f a c t that the ranking stops at the l e v e l o f the A tim arga. H e s u p p lie s an ad dition al
sta tem ent to the effec t that i f fo llo w e r s o f the Atim arga are liberated, then h o w m u ch
m o r e the S a ivas. He a lso understands the word ca in the s e n s e o f api in order to read
this m ea n in g into the text.
200
JUDIT TORZSOK
(navanabha) for initiation is described in chapter 5 (from ver se 19),

followed by an explanation conccrning (he deities to be w o r s h i p p e d
on it (from 5.37c). The nine lotuses are arranged in a c o n c e n tr i c
design in such a way that the lotus in the ccnlrc is su rrounded b y
eight others, and each lotus has eight petals.112 On the pericarp o f the
central lotus, one is to place and worship the supreme deity o f this
system, Svacchandabhairava, surrounded by eight Bhairavas on the
eight petals. The eight Bhairavas arc placed on the petals as ei ght
mantra-syllablcs extracted from the nnviitmabija ,S1 In what follows
(verse 40), the text points out that one should recognize these deities
as standing for [the principles] from SadaSiva down to p r a k r ti and
they are also wor shi pp ed on the pericarps o f the s u r r o u n d i n g
lotuses.84 By including the principles from prakrti up to Sadaîva in
the mandala as secondary deities, the Brahminical schools from the
Ved ava di ns upwards are also implicitly included and their levels
represented by Vidyaraja and the other Bhairavas.85
Thus, doctrinal inclusivism also appears implicitly in a cult im age
in a fairly consistent way, since iion-Brahminical schools, t h e
Buddhists and the Jainas represented by ta ttv a s below p r a k r t i , are left
out o f the mandala and its deities. The S v T s example also illustrates
the co m m o n way to express the superiority o f a tradition in a co n c e n
tric icon, in which the supreme deity o f the school is worshipped in
the centre, surrounded by its ret inue o f deities (parivani) standing for
lower levels o f the universe and lower revelation.Kf
B For a re co n stru ctio n and illustration o f (his m andala, s e c A p p e n d i x 2 an d

C o lo u r Plate 18.
10
S e e K s c m a r a j a s com rncnlary on hakiircna: It A for K apalisa, R A for S ik h ivtlhana, K S A for Krodharaja, M A for Vik arala, LA for Mnnmatha, V A for M e g h a nada, Y A for Somaraja and 0 for Vidyaraja.
M T h e y s w a p p la c c s w ith S v a c e h a n d a as the text and (he c o m m e n t a r y c la r if y
further on.
*s W c d o not n ece s sa r ily need to f o llo w Kscmaraja here, w h o a s s ig n s the tattvas
o f S adasiva, Isvara, vidyii, nvly;l kala, niyati, purusu and prakrti to the eight d eitie s,
for the text it s e lf sim p ly s p e c ific s that they represent le v els from SadaSiva d o w n to
prakrti. The aulhor(s) o f the Tantra m a y not have had an cxact distribution in m ind,
just as the distribution o f tattvas to s c h o o ls is uneven. It is also to be rem arked that
Kscmaraja (ad 5 .1 9 ) understands that the siz e o f (he mandala, w h ich m e a s u r e s 2 2 4
in ches on each side, s y m b o l iz e s the 2 2 4 bhuvana s or w o rld s o f (lie universe. T h is is
an interesting idea, but again one that the Tantra itself d o c s not tcach.
^ The disadvan tage or im precision o f (his arrangement co m p a red to the h ierarchy
e x p r e s s e d in the vertical arrangem ent o f the trident icon is that the su rr o u n d in g
deities arc not arranged in a hierarchy in relation to each other: Vid yaraja sta n d in g
ICONS OF I NCLUSIVISM
201
4 M andates bestow ing Supernatural Powers

Al th o u gh initiation mandalas whether they are used for the samaya
rite or for the diksa appear to be the basis and model o f mandalas
u s e d for acquiring supernatural powers (siddhi) as well as for daily
w o r s h i p , siddhim andalas87 seem to differ s o m e ti m e s from their
m o d e l in several ways. The ways in which mandalas are transformed
o r visualized differently for siddhi can be divided into three groups:
i. Specialization
Some mandalas become reduced in that an element and a deity is
t aken out o f the more elaborate version and the deity is then
worshipped separately for specific supernatural powers.
ii. Expansion
By contrast, some other mandalas are expanded with a set or sets
o f other deities not necessarily present on the basic version, who
s eem to increase the power o f the deity-circle without disturbing
the hierarchy o f the central deities o f the cult.
iii. Substitution
Lastly, some mandalas are retained in their form as described for
initiation except that the deities installed and worshipped on them
are changed; thus the mandala as a drawing is considered some
kind o f framework,
i. Specialization
A good example o f how a mandala is reduced, or rather, how one o f
its deities is focused on for specific purposes can be found in the
ninth chapter o f the SvT. The chapter starts with the description o f
th e worship o f Sv acc handabhairava and h o w his man dal a o f one
lotus and four doors is to be constructed (9.12ff.). The drawing o f the
f o r p ra krti and V cd ic revelation has the sa m e position as S ik h iv a h a n a e m b o d y in g the
level o f ISvara and the PaSupatas.
117T h e term s id d h im a n d a la is u se d in the c o lo p h o n o f the S id d h a y o g e s v a r im a ta ,
c h a p te r 25, w h ic h thus distinguishes b e tw e e n this m a n d a la and the trid e n t-b a se d one
u s e d fo r the n a m e -g iv in g s a n ia y a n tc . H o w e v e r, the S id d h a y o g e s v a r im a ta s e e m s to
e n v i s a g e this v e ry s id d h im a n d a la fo r diksa, j u d g i n g f r o m an a s id e in 2 5 .1 6 c d :
d ik sa y a m sadh a n e h y a sm in n e v a m m a n a vika lp a n a . T his u s a g e m a y reflect th e v ie w
o f th e S id d h a y o g e s v a rim a ta on the subject, n a m e ly that liberation is j u s t one o f the
sid d h is (see 2 9 .8 - 1 1 ) . It sh o u ld also be no ted that the NiSvasa, on th e o th e r hand,
u ses the te rm m an d ala b esto w in g lib eratio n (m u k tim a n d a la ) on fol. 2Qr4.
202
mandala begins with a single lotus, on whose pericarp Sv ac ch an d a is

later to be installed with manlra-syllables and worshipped. This lotus
is then surrounded by a wheel o f 32 spokes (9,16), on w h i c h the
practitioner worships a set o f 32 goddesses, starling with A r u n a,
The size o f this square mandala can vary. Verse 14 e n v i s a g e s
mandalas o f one, two, four or eight hastas on each side; w h i l e the
one with nine lotuses for initiation is prescribed as meas uring nine
hastas on each side.
This mandala o f Svacchanda is said to bestow all kinds o f s u p e r
natural powers, especially power over all the worlds. Some o f these
worlds are well-known from Puranic cosmography; they include the
various hells, underworlds (patala) and the seven lokas. Othe rs are
identical with principles (tattva) o f the universe in the Saiva se nse
from p ra kjii and p u r u sa up to Sadasiva and Sakti. The d i a g r a m
translates into an image o f what is elsewhere insisted u p o n in
doctrinal passages: the idea that the supreme deity u l t i m a t e l y
controls the whole universe, even if lower levels are as sig ne d to
other Bhairavas as their regents, and thus it is this supreme deity that
is able to bestow full power upon the practitioner.
However, the other eight Bhairavas o f the initiation ma nda la are
not forgotten in the context o f siddhi, cither. But while the wo rsh ip
o f Svacchanda is prescribed on a mandala as a support, the other
Bhairavas are placed and worshipped on small charms written on
pieces o f birch-bark (yantras or niksas'>). They are wo rshipped for
the attainment o f much more specific goals than the control o f the
whole universe.
The first yantras described arc those o f the first and last Bhairava,
Vidyaraja and KapalTsa, who can protect the practitioner from death.
The ir mantras are to be incised in the centre o f a wheel, and the
spokes are occupied by the 32 goddesses starting with Aruna, just as
in the mandala o f Svacchanda above. The other yantras are based on
the same model, with one Bhairava in the middle and the goddesses
around him, except that in each case, some additional details are
F o llo w in g verse 16, w hich states that the wheel is outside the lotus, K s e m a r a ja
rem arks at verse 24 that there arc four goddesses in cach o f the eight d irectio n s, and
th a t th e y arc o u ts id e the lotus. H o w e v e r, the text o f v erse 2 4 c o n t r a r i l y to
16 su g g ests that the lotus and the wheel so m e h o w overlap, b c c a u sc it says th at the
g o d d e sse s are on the petals and the spokes at the sam e time.
w As the e x a m p le s below show , ruksu is not a lw a y s a p ro tectiv e a m u le t, d e s p ite
its name.
ICONS OF 1NCLUSIV1SM
20 3
g i ve n. These include instructions to incise the name o f the person

w h o is to be protected, controlled or killed in the middle; or the
a d d it io n o f other mantras to the whole yantra which envelope or
in fl a m e it, or the use o f substances collected in the cremation ground
( 9 . 6 4 - 6 5 ) . Am ong the other Bhairavas, Sikhivahana is employed to
c a u s e o n e s enemy to suffer; Krodharaja can kill someone or make
h i m mad; Vikarala can frighten o n e s enemy; Ma nma tha is invoked
f o r subjugation; M egh ana da to exile so meone and So m ar a ja to
acquir e wealth.
T h e resemblance between the mandala o f Svacchanda and the
y an t r as o f the other Bhairavas shows that in spite o f the differences
b e t w e e n mandalas and yantras, they are closely related, especially in
t h e c o nt ex t o f siddhis. Just as Svacchanda represents the truest
d o c t r i n e o f all Brahminical doctrines which are included in the
in it iat io n mandala, so here, too, he stands for al l-e ncompassing
p o w e r . Other Bhairavas are seen as specializing in more specific
tasks. Th e initiation mandala o f the nine lotuses includes all these
B h air av as as different levels o f reality, for its purpose is to bestow
qualification upon the initiate who can subsequently employ any o f
t h e s e Bhairavas for whatever goal he may want to. But for specific
aims, he is to use only the Bhairava most appropriate for his purpose.
J u d g in g from the construction o f these mandalas and yantras in the
SvT, siddhi is viewed here as the specific application o f the power
acquired in an all-inclusive way at the time o f initiation.
ii. E xpansion
This view o f the SvT is not shared unanimously by all Tantras. In a
n u m b e r o f texts, the mandalas prescribed for siddhis contain several
circles o f deities who are not necessarily present in the initiation
m an dal a. Th e Mrgendr atantra, for instance, allows an initiation
man dal a o f just the central group o f five deities or brahmamantras,90
T h is is referred to in kriya p a d a 8.44 as an option, a lth o u g h it is e m p h a s iz e d in
th e p r e c e d in g verse that one should try and m a k e a m a n d a la w ith sev eral circ le s o f
deities. V erses 8 .5 2 - 5 3 also su g g est that all the deities m a y not b e p re se n t inside the
m a n d a la , w h ic h sh o u ld ideally include three circlcs o f d eities (avarana) aro u n d the
central group. In ease o f these sm a lle r m a n d a la s, one can w o r s h ip th e o u te r circles
on o rn a m e n ta l elem en ts, such as sv a stik a s, lotuses o r dots, o r o n e c a n j u s t w o rs h ip
Siva on the fo ur-pctalled lotus. T h is last solution is interpreted by N a r a y a n a k a n th a to
b e used in case o f lack o f time, place or material means.
204
JUDIT TORZSOK
but prescribes additional outer circlcs o f deities lo be worshipped f o r

certain supernatural powers. For the acquisition of knowledge, it is
sufficient to worship the Vidyesvaras around the central gr oup o f
five brahmamantras. For di vi ne supernatural powers one s h o u ld
add the circle o f the GancSas. For m i d d l e siddhis p r o ba bl y
meaning attracting women and the like the guardians o f directions
should be further added outside the GaneSa circlc. For l o w e r
powers such as killing all the gods should be there, e x te n d i n g
o u t w ar d to the g u a r d i a n s weapo ns on the pe ri ph ery o f t h e
mandala.91 Moreover, according to kriyapada 8.46, in case the practi
tioner has various aims o f different kinds, he is to worship the central
deity surrounded by three or four outer circlcs o f deities on a
saktimandala, which is endowed with an additional circle o f t h e
mothers.92 These mothers are not present on any version o f the ini
tiation mandala, yet they arc employed lor siddhis o f all kinds, in a n
extension o f what is or could be the same as the initiation mandala.93
These prescriptions o f the M rgendra show that the outer circles o f
deities in its mandala are responsible (or siddhis o f increasingly low
kinds. However, they are not worshipped on their own for t h e se
specific functions, but always remain in the outer circles o f t h e
mandala whose centre is occupied by Siva in the form o f the five
brahmamantras. This visual arrangement implies that various siddhis
are not specific functions o f the central deities (as was the case in the
SvT), but rather that the bestowing o f supernatural powers is seen as
an extended function o f these deities, who do not bestow siddhis
themselves, but delegate lower gods, lokapalas and the like, for these
tasks.94
1,1
Cf. kriya p a d a 8.45. N o te that according to N a ra y a n a k an th a , in cacli c a s e o n ly
one circlc is to be w o rs h ip p e d a ro u n d the ccntrc and not several circlcs e x t e n d i n g
o u tw ard to the circlc inclu d in g those deities. T h e ccntrc w ith the live m a n tra -d e itie s
and the g uardians, for instance, sh o u ld be w o rsh ip p ed for m id d le s id d h is, o m ittin g
the V idyeSvaras and GancSvaras in betw een. H ow ever, the text o f the T a n tr a d o e s
n o t su p p o rt this in te rp re ta tio n , for it u ses c o m p o u n d s such as p a tip rfin tn h a n d
ganantah (qualifying yagah).
'2 S ec also c o m m e n ta r y a d lo c: sa k tin a m m a lfn a m s a m h a n d h i y a n m a n d a la m
fa fra....
a T h e addition o f fem ale deities for siddhis. is also a feature o f the S v T as s h o w n
above, w hich positions the goddesses around the central Bhairavas.
*
It m ay be te m p tin g to speculate on the b asis o f this that the M r g e n d r a s a b o v e
a rran g em en t reflects its dualistic position, w hile the S v T s w ay o f attributing p o w e r s
c o rre s p o n d s to a n o n -d u a lis tic view. H o w ev er, it is u n lik ely that culls, e s p e c ia lly
205
iii. Substitution
T h e third way o f changing the initiation mandala for s id d h i can be
illustrated by two examples taken from the Siddhayogesvarimata.
Chap ter 25 first describes the initiation mandala in a rather cursory
way. Its side measures three or four hastas and it lias a 32-inch lotus
w i t h eight petals in the middle thus resembling the basic type
reconstructed in Illustration 1 and Colour Plates 1 6- 17 (following
the more detailed prescriptions o f the srimandala in the NT).95 After
a br ief statement o f how the p lacement o f mantras is to be performed
on the body, the text appears to shift subject to give details o f rites to
a cqu ir e supernatural powers which are to be pe rfo rm ed in the
cremation ground. The placement o f the mantras is followed by the
filling o f the mandala outline with powders. Verse 34 specifies that
white powder is to be produced from powderized human bones and
red from blood. Then, the practitioner is to place a huma n skull on
the pericarp o f the lotus and on the eight petals and should write the
mantra o f Bhairava with his consort on the central skull with blood
taken from his left arm. This Bhairava holds a trident in his right
hand, on which the three principal goddesses o f the Trika, Para,
Parapara and Apara, should be projected. On the remaining eight
skulls the eight mantra-goddesses who form the retinue o f Parapara
should be written, starting with Aghora.
So far, this siddhim andala basically follows the arrangement o f
deities pre scr ibe d in ch apt er 6 for the samaya rite; the three
goddesses occupy the prongs of the trident and the group o f eight is
ea rly ones, w ere b a se d on such p rin cip les. F o r the p r o b le m o f d u a lism an d n o n
dualism in scriptural sources, sec Sanderson 1992: 282ff.
95 T h e sa m e type o f m a n d a la is giv en in an o th e r T rik a text, the T a n tr a s a d b h a v a
( 9 . 1 0 4 f f ) , w h ich calls it the sarvatobhadra(ka). A lth o u g h the S i d d h a y o g e s v a rim a ta
se e m s to ag ree w ith the T a n tra sa d b h fiv a on this m a tte r r a th e r th a n w ith a th ird
s u rv iv in g T rik a text, the M alin lv ijay a (w hich gives a m a n d a la o f a trident an d the
lo tuses), the su b se q u e n t verses on sid d h i s h o w that there is a trid e n t p re s e n t on the
m a n d a la o f the S id d h a y o g e sv a rim a ta , too: but instead o f b e in g d ra w n on the g ro u n d
in s id e th e m a n d a l a , it is d r a w n in b lo o d on a sk u ll p l a c e d in th e m id d le .
N e v e r th e le s s , it m u s t be m e n tio n e d th at the text d o c s n o t g iv e an u n a m b i g u o u s
a c c o u n t o f the m a n d a la s. T h is c h a p te r as well as ch ap ters 7 and 8 su g g e s t that the
initiation m a n d a la ( diksamandalci) m a y well re s e m b le the T a n t r a s a d b h a v a s sa rva to
bhadra, w ith o u t the trident, even th o u g h c h a p te r 6 c learly p re s c rib e s the m a n d a la
w ith th e trid e n t fo r th e sam aya rite. M o re o v e r, A b h i n a v a g u p t a s s u m m a r y o f the
principal m a n d ala o f this text on the basis o f w h ic h I h ave a tte m p te d to reconstruct
the m a n d a la in C o lo u r Plate 19, but w h ic h is not in c lu d e d in the s u rv iv in g short
re c e n sio n u n am b ig u o u sly gives one with the trident,
206
JUDIT TORZSOK
placed on the petals o f the lotus. However, after g i v i n g the

bijam an tras for the goddesses and promising the successful invo
cation o f Yoginis, verses 5 3- 55 give a new set of eight goddesses
one can equally worship with the same or other bijas on the diagram.
Their names indicate that they arc probably recommended for lower
siddhis such as killing,% The diagram itself becomes a frame wo rk
which can accommodate various groups o f deities depending on the
siddhi envisaged.
A procedure similar to this seems to be at work in chapters 21 and
22 o f the same text. Chapter 21 describes, again very briefly, a
wheel-diagram with 12 spokes, which appears to be r eco m m en d ed
for wo rship in various months o f the year. Tw el ve Rudras o f
different names are placed on the spokes, and a Bhairava, identified
with o n e s self, performs the churning o f the nectar o f immortality in
the middle. Subsequently, the text gives a summary in a few verses
o f a samaya type ritual, in which this very whccl-diagram is e m
ployed to determine the gotra name o f the initiates (verses 2021),
implying that the wheel is used as the central image o f a mandala.
The wheel is to be drawn red, with a mixture including blood. N o w a
passage in the next chapter prescribes the visualization o f the same
kind o f wheel as the one m ent io ned for the golra attribution.
However, this time the wheel, whose basic colour is red and which is
to be visualized in the middle o f an Ocean o f Blood, is mentally pro
jected in the air. Instead o f 12 Rudras, 12 frightening Yoginis or
optionally six Yoginis accompanied by six Rudras as consorts97
are to be placed on the spokes. They churn amrtn from the Ocean of
Blood and bestow success upon the practitioner.
The wheel-diagram used for a golra name-giving initiatory rite
and r e c o m m e n d e d for monthly worship is thus em p l o y e d as a
f ram ew or k on which more fierce fill deities are installed to gain
supernatural powers, just as the lotus mandala o f the main goddesses
can also ac c o m m o d a t e goddesses associated with black magic
(abhicara).
A lthough the above examples are fairly representative o f the
major ways in which the initiation mandala can be transformed to
be sto w sid d h h , they are not exhaustive. T w o additional co m m o n
* T h e first o f th em is callcd Ja m b h a n i and the last, P ram atham .
T h is is h o w the c o n jun ctio n v ;lm a y be interpreted in verse 34, follow ed by the
m ention o f th e tw o sets o f six in verse 35.
207
t y p e s o f procedure should be mentioned in this context: certain

graphical differences which are associated with particular siddhis
a nd the change o f materials with which the mandala is prepared.
As for graphical changes, after describing the principal mandala,
w h o s e base is square-shaped, the Nisvasatantra mentions briefly how
to d r a w the outline o f circular ( vaiiula) mandalas, mandalas o f a
semi-circular ( ardhacandra) type and triangular (trikona) ones. This
is f oll o w e d by the sta te m en t that for rites o f pr op itiation or
a p pe as em en t as well as for prosperity, one should use the square or
c ir cu la r type; the semi-circle should be used for the m an da l a o f
Ca nd e sa and the triangular type, for black magic.118
A different sort o f graphical change is prescribed for the worship
o f various deities and for the acquisition o f siddhis in the Mrgendra,
w h i c h draws the shape o f the petals o f the lotus in the man dal a
di ffe re nt ly for different purposes.1 For supe rnatu ral po w er s in
general (b huti), the petals should have curled-up tips; they should be
e v e n for liberation ( m u k ti) 'no For the worship o f K am es v ar a and
other deities bestowing good fortune, the petals are to have pointed
tips, whil e if one worships the iokapalas, the grahas, the ganas,
Ca ndeSa or Ganesa, the petals must be broad and cur ved at the
e d g e s . 1111 The ganas are also given round-shaped petals in another
verse. The worship o f the Vidyesvaras requires petals shaped like
c o w s ears, and Gauri, the goddess o f speech (vak) and the Rudras
are to be worshipped on lotuses whose petals look like the leaves o f
the Asvattha tree.
In th e s u b s e q u e n t v er se s, the M r g e n d r a g iv es y e t m o re
alternatives, which concern the shape o f the mandala and its colour
according to the deities or the purpose o f the worship. Thus, agreeing
w i t h the NiSvasa, it p r es cr ib es a s e m i - c i r c u l a r m a n d a l a for
worshipping Candesa, but it adds the worship o f the A m r ta v id y a s 102
to this category and specifies that the colour o f the diagram should
98
vartulam caturasram va siin tike p a u stik e tatha /

ardhacandrani tu ca n d ese abhicarc triko n a ka m // fol, 26r3,
( T h e reading o f the m a n u sc rip t c a n d l e has b een e m e n d e d to candese.)
v>See M rg en d ra, kriya p a d a 8 .3 1 -3 3 .
I fo llo w N a r a y a n a k a n t h a s in te rp re ta tio n o f the v e rse , w h o u n d e r s ta n d s the
s o m e w h a t en ig m atic sp h u ta m to m e a n a n im n o n n a ta g ra m sa m a n l.
mi F o llo w in g N a r a y a n a k a n th a s interpretation o f the w o r d m antharagrakam .
" T h e se are fe m a le m an tra-d eities b esto w in g im m ortality and the like.
208
JUDIT TORZSOK
be while. The mandala o f the Saubhagyavidyas111'1 is said to be red

and vulva-shaped or bow-shaped. The description continues with
further options for purposes o f well-being, for rain and for the
worship o f the VidycSvaras.
The materials used for (lie preparation of the mandala should also
be different when used for liberation and when employed for lower
supernatural powers (abhiciira) according to Ihc same passage o f the
Mrgendra ( kriyapada 8.4Q):I,M the coloured powders are to be made
from pearls, corals, gold and the like for liberation, but they are to be
produced using substances from the cremation ground for the lower
powers.
However, if the same mandala is used for liberation as well as for
the acquisition o f supernatural powers,"15 the Mrgendra instructs
practitioners to make white powder from rice flour, red from red
mineral from mountains vermillion or red chalk according to the
c o m m en ta to r , yellow from yellow orpiment or t u rm e r ic 106 and
black from burnt barley and the like. Since initiation mandalas used
for several people should not be made for a specific s id d h i, this
general type is probably what the text envisages for diksa.
It must also be noted that a more esoteric text teaching goddess
w or sh ip , the Si dd hayogeSvar imata, does not presc ribe i m p ur e
substances for specific siddhis as obligatory. At the sa me time it
allows these substances as alternatives for initiation as well. Thus, in
6 .12, the text gives the choice to the guru if lie wants to use ashes
(probably meant to be collected in a cremation ground) or flour for
the sam ayam andala, and in 8.8 it also gives alternatives without
restriction for the making o f the thread to be used to prepare the
outline o f the mandala: it can be made o f human hair (n a r a k e sa sam utthena, again probably obtained from corpses) or o f cotton and
the like. On the other hand, even for rites to acquire supernatural
powers, the instructions state that Hour or ricc-powder may be used,
,IB A c c o r d in g to N a r a y a n a k a n llia th e se arc m a n t r a - g o d d e s s e s b e s t o w i n g the
pow ers o f subjugation and attraction,
l(M It m a y be no ted that as the a b o v e p a s s a g e s sh o w , the M rg e n d ra a p p e a r s to
m en tio n a n u m b e r oT details about m a n d a la s w hich do not c o n cern on ly initiation
m a n d a la s , a lth o u g h the m ain subject o f the p assag e is indeed initiation m a n d a la s.
T h e r e f o r e the d istin c tio n b e tw e e n n u i k t i and abhiciira m ay not refer to initiation
m a n d a la s o f initiates with different purposes, but rather to sid d h im a n d a la .
105 T h is is w h at the c o m m e n ta ry suggest,'; at the beg in n in g o f the p assag e, saying
bh uktim uktivisayilnam yaga nam.

I,lfi I again follow the c o m m e n ta to r on haridrakadinii: haridraharitalfidina.
209
a l t h o u g h pr e fe r e n c e is given to im pure s u b st an ce s ,
powderized human bones for white and blood for red.107
such
as
5 Substitution and Change:10* The W orship o f Visnu and the

Inclusion o f the Buddha in the Netratantra
T h e siddhim andalas e x a m in e d so far s h o w that th e p a n t h e o n
worshipped on them can vary significantly and in several ways from
the pantheon o f the initiation mandalas. However, in all these cases
th e deities worshipped remain those taught in the Saiva systems:
f orm s o f Siva or Bhairava, various Yoginis or groups o f deities
form in g their retinue such as the lokapalas. It could be argued that
for siddhis, the SvT employs Bhairavas who em body lower levels of
t h e u n i v er se in the initiation mandala, and w h o c o n s e q u e n t l y
represent lower, non-âiva revelation: Vidyaraja w ho is employed to
c o n q u e r death, for instance, stands for the level o f prakrti and by
implication embodies the level attributed to Vedavadins. However,
t he deity remains a Saiva mantra-deity with a visual appearance and
n a m e o f a Bhairava, In this respect, the N T stands apart from other
early texts in that for siddhis it prescribes the alternative worship o f
deities who clearly belong to other systems, by substituting the
central god o f the cult, M rt y u m j ay a/ N et r an at h a (as well as his
consort), with non-Saiva deities.
Aft er describing the drawing and decoration o f the mandala, the
N T lists the deities to be worshipped in its centre, on the pericarp o f
th e lotus. First, the principal deity o f the cult, M r t y u m j a y a is
m ent io ned (18.62), who is to be worshipped with the goddess o f
pro sperity, Sri, as his consort. Secondly, an alternative is given
]W S id d h a y o g e s v a r l m a t a 2 5 ,3 4 c d : sita m n r sa n k h a ja m c u rn a m ra k ta m ksa ta ja b h a v ita m . I h a v e c o n je c tu re d nrsa n kh a ja m fo r the r e a d in g o f the m a n u s c r i p t s trisankhajam .

'B y this subtitle 1 intend to e v o k e the principal a r g u m e n t in E iv in d G. K a h r s s
u n p u b lis h e d dissertation entitled S ubstitution an d ch an g e: fo u n d a tio n s o f traditional
Indian h e r m e n e u tic s (O slo, 1996). I h ave not got direct access to this w o rk , w h o s e
m a in t h e s i s w a s s u m m a r iz e d b y the a u th o r in p e r s o n a l c o m m u n i c a t i o n s a n d in
lectures at the U n iv ersity o f C a m b rid g e. T h e idea is also referred to briefly in K ah rs
1998: 2 7 8 , w h o states that c h a n g e is a c h ie v e d t h r o u g h s u b s titu tio n in th at n e w
m e a n i n g m a y b e e n c o d e d into old te rm s by m e a n s o f a s u b s titu tio n a l m o d e l . In
w h a t follow s, I h o p e to s h o w that substitution is app lied in the N T in a ritual context,
n o t as a h e rm e n e u tic device, b u t as a m e th o d to in clu d e deities o f o th e r cults in its
pan th eo n .
210
JUDIT TORZSOK
whereby Sridhara, that is Visnu, can be substituted for the principal

deity. Visualizations o f Visnu ore described at the b e g in n in g o f
chapter 13, in which a number o f Vaisnava forms arc listed whi ch
include various incarnations such ;is the Man-lion (Narasimha), the
Boar (Varaha) and the Dwarf (Vamana). But the list o f alternatives
does not stop here, for the text continues by giving the visualization
o f Surya, forms o f Rudra, Marihara, Ardhanarisvara, Brah ma , and
finally the Buddha, who is said to specialize in granting liberation to
women. The commentator, Kscmaraja, introduces this p a s s a g e 109 by
say ing that the text en um er ate s various altern ative f o rm s o f
Mrtyumjaya, They can all be worshipped on a mandala, too, which is
first shown in the prescription according to which Surya is to be
placed in the middle o f a lotus (verse 23), and later by the mention o f
various loci o f worship in 28. These deities or following K s e m a
raja forms o f Mrtyumjaya can be visualized on the ground, in fire
or water, on the top o f a mountain, or in any other place whi ch is
pleasing to the mind and shall all bestow the desired success. This is
further confirmed in another passage o f verses 3 7- 43 , which states
that all kinds o f deities lead to success if they arc wor shi pp ed as
prescribed. The text explicitly says that deities o f other Tantric
traditions can be invoked as well as those o f the N y a y a , the
Buddhists,110 the Yoga, Vedic deities, e t c . " 1 The list shows that the
B u d d h a is not visualized and regarded as a m an i fe st at i o n o f
Vijmu which could also be the case but is considered to be the
Buddha o f the Buddhists and is invoked as such.
W ha t is most striking in the inclusivism of the NT is that it does
not stop at the level o f Vedic revelation and Brahminical darsanas,
but includes the Buddhists, and that it allows the worship o f forms o f
Visnu and the Buddha as principal deities.
N o w the SvT also includes the visualization o f B ra h m in ica l
deities w ho are not Saiva strictly speaking, and who represent lower
levels o f the Saiva universe situated below forms o f Siva. An
example for such inclusion can be seen in the description o f internal
worship, in the course o f building up the Saiva universe internally.
W he n visualizing the lotus o f gnosis (vidyapadnm ) on top o f the
l(HCf. the c o m m e n ta ry before S v T 2 13.17.
1111T h e N T 2 uses the irregular or aisn form iirahata as docs the SvT.
111
T h e e t c . in the text is interpreted by K scm araja lo m ean Puranic deities. T h is
m a y in cludc the w o rs h ip o f deities such as D u r g a - V in d h y a v a s in i m e n tio n e d s u b
sequently.
211
throne, which is nailed together by the four Vedas and the four aeons
( 2 ,6 4 c -6 5 b ), first a circle o f Saktis is described, w hich is to be
placed on the petals with die goddess M anonm ani on the pericarp.
T h is is follow ed by the p lacem en t o f th ree circles (m andalas/
m andalakas) on this lotus o f gnosis: the circle o f the sun on the
petals, that o f the moon on the filaments and the circle o f fire on the
p ericarp .112 T hen the visualizations o f th ree deities ( o f P u ra n ic
a p p e a ra n c e ) as reg en ts o f th ese th re e circles or sp h e re s are
prescribed: B rahm a, Visnu and Rudra placed on the outer, m iddle
a n d in n e r c ircles o f the p e ta ls, fila m e n ts a n d th e p e ric a rp
respectively. It is on top o f Rudra, still on the pericarp, that the
la u g h in g S adasiv a /M ah ap reta is then to be p ro je c te d b efore the
visualization and worship o f Svacchandas throne and o f Svacchanda
himself.
The difference betw een the place and role o f V isnu in the N T
com p ared to the SvT is that on the one hand, the N T prescribes the
w orship o f forms o f Visnu as the central deity, and on the other that
it gives several alternative form s o f V isnu, w hich indicates its
s o m e w h a t u n u su al interest for this deity in a S aiv a c o n te x t." 3
M o re o v e r, the a p p e a ra n c e o f the B u d d h a as cen tra l d eity is
undoubtedly unique here. By prescribing the w orship o f these deities,
the N T goes much further than the SvT in including other cults. This
m a y be considered not only another elem ent sh o w in g the N T s
relative lateness," 4 but also a feature that m ay reflect a different
religious scene o f its tim e.115
112 A c c o r d i n g to K s e m a ra ja ( avataranika o f 2 . 7 2 c d - 7 3 a b ) , th e s e th re e c irc le s
re p r e s e n t the in s tru m e n t, the ob jcct an d the su b je c t o f g n o sis ( mana, meya, matr )
r e s p e c tiv e ly as well as the three p o w e rs o f kn o w led g e, action and will (iccha, jm n a ,
kriya).
113 A lth o u g h V a isn a v a inflections o f Saiva deities m a y be often e n c o u n te re d , the
N T s in terest in se v e ra l such fo rm s m a y be c o n s id e r e d u n u s u a l. F o r a V a i s n a v a
v e r s io n o f K ali w o rsh ip , sec the e x a m p le fro m the J a y a d r a th a y a m a la te a c h in g the
w o r s h ip o f K ali M adhavevari w ith N a ra sim h a , given in S anderson 1988; 154.
114 T h a t the N T b e lo n g s to a re lativ ely later lay er o f the early, p re -lO th cen tu ry ,
scriptural so u rces has b e e n arg u ed on the b asis o f a m p le e v id e n c e in B r u n n e r 1974:
126ff., w h o also cites M ad h u su d a n K a u l s in troduction to the first edition.
115 R itu a l e c l e c t i c i s m a n d c h a n g i n g a ttitu d e s t o w a r d s s u c h p h e n o m e n a w e r e
a n a ly z e d in a series o f p apers by P ro fe s s o r Phyllis G r a n o f f at the E c o le d es H a u te s
E tu d e s en S c ie n c e s S o ciales (Paris) in A p ril-M a y 2 0 0 0 , e s p ecially in her first p a p e r
en title d O th e r p e o p l e s rituals: ritual ecle c ticism in early m ed ie v a l r e l i g i o n s I am
gratefu l to P ro fe s s o r G r a n o f f for g iv in g m e a v ersio n o f the final draft o f he r stu d y
an d for d ra w in g m y atten tio n to the ritual e c le c tic is m o f the N T . In the m e a n t i m e
p a rt o f this material has b een p u b lish e d ( G r a n o f f 2 0 0 0 and 2001).
212
W hat is perhaps less striking, but almost equally surprising; is the

inclusion o f the deity o f the N y a y a -w hoever it is su p p o sed to
b e .110 The inclusion o f the Nyaya is surprising bccau se th e SvT ,
which can probably be regarded as something of a reference w o rk for
the N T ," 7 categorically rejects treatises o f logic (hctuM stra) and
condemns them in a relatively long passage.Ilh
At the same time, the NT can be said to follow a kind o f logic
already established in other Tantras. For the substitution o f principal
deities o f the cult with less central ones in mandalas em ployed for
siddhi is a practice also seen in the example o f the non-syncretic
Siddhayoge& varim ata. The NT applies the sam e p ro c e d u re o f
substitution, except that it goes a step further and includes V aisnava
and non-Brahminical deities or cult figures, who do not form part o f
its basic pantheon. The substitution can be justified in the same w ay
as the SvT justifies its recognition o f other scriptures: all scriptures
as well as all deities and doctrines arc created by and identical with
the supreme deity o f this Tantra."''
Conclusion
Instead o f sum m arizing the m ajor points o f the above analysis o f
in d u s iv is m or eclecticism points which may well change in the
light o f further evidence I would like to mention tw o particular
features o f the m andala as locus o f w orship that m ay have c o n
tributed to in d u siv ism or ritual eclecticism in B hairavatantras: the
K s c m a ra ja a d loc, interprets this rcfcrcncc to m ean that the s u p r e m e deity o f

the N y a y a is one e n d o w e d w ith q u alities o f o m n is c ic n c c and the like, but w h o is
ontological ly different from men,
117 F or the N T s relying on the SvT, see B runner 1974: 126IT.
"* A s m e n tio n e d ab o v e , e v e n if N y a y a and HetuSastra (th e la tter p o s s i b l y
m e a n in g any w o rk questioning the authority o f revelation) m ay not ex actly c o v e r the
sa m e b ra n c h o f S aslra, it is un lik ely that th ey arc strictly d if f e r e n tia te d in th e s e
scriptural p assages. T h e above m e n tio n e d passage about treatises o f logic is lo ng in
that no other school o f thought is treated or criticized in such detail in the text. See
S v T 11.167179b on w h at it calls Hctu&islra, while all other traditions are dealt with
in ve rse s 1 7 9 c - 185.
IWS o m e th in g to this effect is stated in 1 3 .44-46. This p a ssa g e c o n firm s that the
prin c ip a l d eity o f this cult is the soul o f all m an tras, an d th e re fo re th e re is no
in frin g e m e n t o f th e rules p resc rib in g that rites o f d ifferen t s c h o o ls s h o u ld not be
m ix e d up. K s e m a ra ja s c o m m e n ta ry a d loc. a dds a m ore strongly n o n -d u a listic inter
pretation o f this statement.
213
c o s m ic sym bolism o f the mandala, and the fact that the draw ing
itse lf is an empty framework.
A lthough the initiation m andala may not depict the Saiva cosm os
in a m ore explicit way than other supports o f w o rs h ip , 120 its c o n
centric or vertical image o f a hierarchy o f deities and other elements
is often seen as representing a cosmic hierarchy, too, Thus, the image
o f the m andala is identified with the cosmic hierarchy in scriptural
s o u rces: the trident o f the Trika is u n d ersto o d to rep resen t the
universe from earth to Siva, the deities o f the SvT represent levels
fro m p ra kiti up to S iva-B hairava and the five outer lines draw n
a ro u n d the m andala o f the N T stand for the five kalas w hich
constitute the &aiva universe (see A ppendix 1). This identification is
continued by Ksemaraja in his com m entary on the SvT, in w hich he
states that the 224 inches o f the side o f the m andala represent the
w o r ld s o f the Saiva u n iv erse. Since the m a n d a la is seen as
representing the cosmic hierarchy, it includes lower revelation. Thus,
at least for purposes o f siddhi, deities o f these lower revelations may
be used effectively. They o f course do not fully deprive the central
deities o f their im portance and place in the hierarchy: for example,
em ploying the B uddhas pow er is recommended m ainly for women.
A second feature o f mandalas which may have contributed to the
substitution and inclusion o f non-Saiva deities is that they do not
actually depict the deities them selves.121 Although Tantric deities can
be visualized for w orship, their identity lies first and forem ost in
th e ir m antric form, as pointed out in S anderson 1990: 78. In a
nu m b er o f texts, this means that they can be visualized in som ew hat
120 As B ru n n e r 1986: 30 (cf. B ru n n er, p, 175) points out. H o w e v e r, as I h a v e tried
to a r g u e above, the m a n d a la is a spccial case c o m p a r e d to the lin g a or a te m p le in
th at it g iv e s a m a p p in g o f the deities o f o n e s tradition.
121 I think th e re is a p ractical reaso n for this, a p a rt f r o m the d o c trin a l re a s o n s
a lre a d y m en tio n e d . S in cc th e se m a n d a la s arc m o s tly m a d e o f c o lo u r e d p o w d e rs , it
w o u ld require an ex tre m e ly !arge-sizc m a n d a la to b e able to depict d eities in detail
on it. Such fig u rativ e im a g e s in co lo u re d p o w d e rs are m a d e e v e n to d a y in K erala.
An e x a m p le is the K a ja m E luttu, w h ic h depicts B h a d ra k a ll before the p e r f o r m a n c e
o f a M utiycttu. T h is sh o w s that even a single deity req u ires a ra th e r large diag ram .
T h e co n stru ctio n o f such a d ia g ra m w o u ld be qu ite u npractical for rites such as the
n a m e - g iv in g sa m a ya ritual. H o w e v e r, w h at is m is s in g on the S a iv a T a n tr ic im a g e
can be a m p ly p ro v id e d by detailed v isu a liz a tio n , the re su lt o f w h ic h m a y n o t be as
different from so m e T ib etan B u ddhist m an d alas as B r u n n e r 1986: 31 (cf. B ru n n e r, p.
176) claims. W ith o u t trying to d raw too m a n y parallels b e tw e e n th e s e tw o traditions,
I w o u ld j u s t like to e m p h a s i z e a g a in th at th e S a i v a m a n d a l a is n o t s im p ly the
d ra w in g itself, but the d raw in g and the visualization.
214
differing forms according to o n e s purpose, and thus the g o d d e s s

Para, for instance, is given a rosary and a m anuscript as a t tr i
b u tes sim ilarly to the goddess o f speech, S a r a s v a tl122 w h e n
visualized to obtain eloquence, but she is pictured as pouring out the
nectar o f immortality to conquer death,121 Taken to the extrem e, this
principle implies that any visualization can suit a m antra-deity, w ho
may well take up the appearance o f the Buddha if needed. T h e
m a n d a la is q u ite w e ll-a d a p te d for such radical c h a n g e s in
visualization, for it does not depict the deities in their concrete forms.
If one uses the geometrical framework o f the mandala, changing the
deity s appearance in visualizations does not necessitate any change
in the traditional mandala, draw n according to Saiva sc rip tu ra l
prescriptions.
In a final remark, it could be concluded that w hat renders the
identity o f B hairavic m antra-deitics weak is in fact their p o w e r
fulness. It is because they arc too powerful and dangerous to b e
depicted with their iconographic features (TA 27.23) on objects such
as the m andala that their visualized images can be c h a n g e d or
replaced by the images o f other deities; and it is this iconographic
interchangeability that allows ritual inclusivism or eclecticism .124
122F o r this identification, see Sanderson 1990: 43.

123 S e e S i d d h a y o g e v a r i m a t a , c h a p t e r s 1 1 a n d 12.
124 It w o u ld req u ire an a lto g eth er sep arate p a p e r or m o n o g r a p h to e x p lo r e w h y

such e c lecticism takes place in the N T in particular, w h ich e x h ib its a s o m e w h a t lax
attitude tow ards m ix in g prescriptions o f different traditions (tanirasainkara in 13.46).
In addition to an investigation into the c hanging religious sccn c o f d ifferen t p eriods,
a rath er precise ch ro n o lo g y o f the scriptural sources w ould also be needed.
215
A p p en d ices
A p p e n d ix 1: The Construction o f the Srimandala
B e lo w is a b rie f description o f how a simple m andala is constructed,
follow ing the prescription o f the N T with K sem arajas com m entary.
N o te th at w ith o u t the c o m m en tary it w ould be im p o ss ib le to
reconstruct the mandala. Technical terms are given with illustrations
so th at this sum m ary can serve as a basis for the reconstruction o f
o th e r mandalas. Their descriptions seem to follow m ostly the same
gen eral term inology with a few m inor differences. Som e o f these
d ifferen ces will be pointed out below. W ays in which the cardinal
d ir e c tio n s are d e te rm in e d on the g ro u n d , d e ta ils c o n c e rn in g
m easurem ents, some problems pertaining to the actual draw ing w ith
the help o f threads and the colouring with powders are not discussed
here. H ow ever, it must be born in mind that these factors, too, form
p a r t o f the process o f m andala construction, and are som etim es
d e tailed in the m iddle o f the description o f the draw ing. A good
ex am p le is the discussion o f how to establish the directions in SvT2
5.29ff. W h a t follow s concerns only the actual d raw in g and the
colours applied according to N T 2 18.3Iff. A sim ilar m andala w as
reconstructed in Brunner 1986 ( c f Brunner, p, 177) on the basis o f a
later text, the Saradatilaka and R ag h a v ab h attas co m m en tary , but
w ith o u t explanations o f all the technical terms. For the construction
o f the srimandala, see Illustration 1, while the final result o f the con
struction can be seen in C olour Plates 16 and 17 according to two
versions.
T he drawing o f the mandala starts with the construction o f a grid,
in which the size and num ber o f the cells vary . The shape o f the grid
is square ( caturasra) and its lines are always draw n along the northsouth and east-west axes. A cell is called a kostha or kosthaka, and
the length o f one o f its sides is a bhaga. In the NT, the grid o f 324
cells has 18 bhagas on each side. There is an eight-petalled lotus in
the middle occupying eight times eight cells.
In m ost mandalas, the construction o f a lotus follows the w a y in
w hich this central lotus is produced in the NT. First, four concentric
circles are drawn in the middle o f the central square o f the lotus. T he
first one has a radius o f one bhaga and is the circle o f the pericarp o f
the lotus (karnika). The second circle has a radius o f two bhagas and
2 16
JUDIT TORZSOK
marks out where the fibres will end (kcsariigra). The third circle has
a radius o f three bluigaa to show where the petals will be jo in e d to
one another (dalasamdhi). The last circle has a radius o f four bhagas,
to mark where the tips of the petals should end (dalagra).
This is followed by the drawing o f the lines where the petals o f
the lotus will meet. This means that first, one should draw eig h t lines
from the centre in the cardinal and intermediate directions. T h e s e
lines will intersect the outermost circle at the points where the tips o f
the petals arc to be. Then one draws eight additional lines w h ic h
must be in the middle o f those eight radii. This halving is done in the
same way as at the establishing of the square o f the m andala. In this
case it is done by halving the line which one could draw betw een tw o
petal lips, starling with the lips o f the north-eastern and the northern
petals. It is obvious from the description that since these latter lines
represent the sides o f the petals, they will be visible from outside the
circle o f the pcricarp up to the third (dalasam dhi) circle. It is also
m entioned that the petals have three fibres each (draw n from th e
pericarp up to the second circle).
T he next step is the drawing o f the outlines o f the petals o utside
the d alasam dhi(i.e., the third) circle. Ksemaraja says that one should
draw two arcs with the help o f a thread, fixing the thread with the left
hand in between the line in the middle of the petal (m adhyasutra) and
the line on the side o f the petal (parSvasutra). Then one should d raw
two arcs on both sides [of the petal] with the right hand, starting
from the point where the petals should intersect (already established
by the intersection o f the third circle and the parSvasutras).
After the description o f the lotus, the text gives the colours for its
various parts. It continues by stating that a white circle is to b e
drawn, its thickness measuring one inch, around the lotus. This is the
so-called air-line ( v yo m a w kh a ). Outside this circle, a square should
be made, with a yellow line which is one inch (ariguki) thick. T his is
the inner part o f what is called the seat (pitha), w hich is a square
band occupying one bhaga outside the inner square. (The w idth is
m entioned by K sem araja in his com m entary on 43cd.) T h e n this
surrounding band is to be divided into the corners (kona) and the socalled lim bs (gatraka). These limbs are formed here by leaving tw o
bhagas on each side for the corners. Thus the lim b s occupy four
cells each, as K sem araja m akes it clear. He also gives a b r ie f
definition o f the lim bs: they are particular segments which should
217
fall in betw een corners, outside the air-line (gatrakani konantaralaga avayavavisesa vyom arekhaya bahye karyani).
Next, one should leave a band which is two bhagas w ide around
th e s e a t, This is the terrace or passage (v ith i). In K se m a ra ja s
explanation the passage is the place to conduct the p u ja . O utside o f
this is the area where the doors are to be drawn, on a surrounding
tw o bhaga wide band.
T he door has two parts, here called kantha and upakantha. The
fo rm e r is the u p p er part, i.e., the part closer to the centre o f the
m a n d a la . In this m andala it occupies four cells, as K se m a ra ja
explains. He understands the prescription o f two cells in the text to
apply on both sides. Colour Plate 16 reproduces the m andala accor
ding to K sem arajas interpretation, while Colour Plate 17 gives the
b a sic structure w ith o u t the ornam ents and w ith o u t c o n sid e rin g
K sem arajas remark about the size o f the doors.
The base or lower part o f the door is one cell w ider on each side
here. K sem araja gives a definition o f both kantha and upakantha:
ka n th a m dvarordhvagam a v a y a v a v is e s a m u p a k a n th a in kanthadhog a m avayavavisesam . N ote that in a num ber o f texts, this base is not
called upakantha, but k a p o k i125
N ex t to the door, there is an ornamental part which has the shape
o f the door turned upside d o w n .120 The smaller and outer part o f this
elem ent is called the goblia and the wider upper part is the upasobha.
A gain, their m easurem ents are based on K s e m a ra ja s com m entary,
w h o h im s e lf mentions that some details are left out and understood
to be supplied by the reader. Note that the sob h a and upasobha are
c a lle d kapola and u p a k a p o la in the Isa n a s iv a g u ru d e v a p a d d h a ti
(kriyapada 8.58), in which they are also defined as having the shape
o f doors tu rn e d u pside dow n and being p laced n e x t to th e m
( tatparsvatas tadvipantavaktras tadvat kapolopakapolakah syuh).
W hen the doors are ready, the text m entions that one is to draw
three circles (inside the doors?), leaving out the w estern door, w hich
faces the deity. This is not com m ented upon by K sem araja, but there
is a b rie f mention o f a circle in the context o f the door in TA 31.83,
in w hich the door is said to be circular optionally. In the context o f
the NT, how ever, it seem s that the function o f these circles is to
125 Sec, e.g., S v l '2 5.34ff., M alinivijaya 9.3 la b and T A 31.39, 31.84cd.
S ec K s cm araja on 44d: dvaraparsvayoh p a ra vrtta d va ra sa m n ivesa ka ren a tatha
so b h opa sobha k c " k ft rnyct.
218
JUDIT TORZSOK
close, cover or seal the doors, and this is probably the reason w hy the
western door facing the deity has no circlc. For this idea, see, e.g.,
the Laksm lkaularnava quoted by Kscmaraja ad SvT2 5.35ab: dvaratrayam pidhatavyam pascimnm m pidlmpuyct. Nothing is said about
the exact position or size o f these circlcs in the NT.
This mandala, being that o f ri, is decorated with conch shells and
lotuses in the vithi ^ in the outer corners and outside. All colours are
given in detail except those for the outer corners; and it is also not
clear if the decorations in the vithi should be black or the vithi itself.
Since all parts o f a mandala have to be covered w ith c o lo u re d
pow der so that the ground should not be seen,127 it can be assum ed
that the corners also had some colouring. I have applied w hite for the
vithi as well as for the corners. The former is said to be always white
in Isanasivagurudevapaddhati, kriyapada 8.73,12* and as b oth th e
v ith i and the corners have the decorations o f conch shells and
lotuses, 1 assume they are also o f the same colour. Consequently, the
outlines o f the conch shells and lotuses are m ainly black, u n d e r
standing that the N T refers to the outlines o f ornam ents w h en p re s
cribing the black colour and not to the vith i.m V erse 4 7 c d - 4 8 a b
prescribes the draw ing o f five lines around the m andala, w h ich
represent the five kalas, the lowest ( nirrti) being the outerm ost one.
A ccording to Ksemaraja ad b e., they are white, red, black, yellow
and transparent, starting from the outermost line.
127 See, e.g., a b r i e f aside in l6 a n a iv a g u ru d e v a p a d d h a li( k riy a p a d a 8,61b: ya th a

b h u m ir na drsyatc.
128 T h is p a ssa g e states that the do o rs and the petals sh o u ld also be w h ite , w h ic h
agrees w ith the NT,
129 It is also unlikely that it should refer to the o rn a m e n ts th e m se lv e s, for it w o u ld
be very odd to require conch shells and lotuses to be black.
IC O N S O F IN C L U S IV IS M
'1*
1
'
i
' u p a
i
x
k a n . t k a .
! ' ' '
.
,
!
219
1. The outline and construction o f the srimandala according to the Netratantra
220
A p p e n d ix 2: The Navanabhamandala
The m andala o f the nine lotuses (Colour Plate 18) has been re c o n
structed according to S vT2 5.19 34 and K sem arajas c o m m e n ta ry
thereon. The grid measures 224 x 224 ahgulaa and is divided into 7 x
7 large bhagas. Ksemaraja remarks that the number 224 reflects the
num ber o f bluivam n. Mere again, the drawing starts at the centre,
where one is to draw a lotus in the same way as described in the N T
above in the central kostha. The di fference is that the scat o f the lotus
here is the outline o f the ccntral kostha itsel f without the construction
o f the gatrakas. The eight other lotuses arc constructed in the sam e
way around the ccntral lotus, leaving one kostha in betw een them .
The various parts o f the surrounding area, the sobha, upasobha, and
kantha are all said to be h a lf the size o f the v ith i. K s e m a r a j a
understands this to refer to the v ith i around the lotuses, w hich has
been obtained by halving the space between the lotuses and the edges
(see 3 3 c d - 3 4 and com m entary). The text itself must refer to the
height o f these elements, while their varying widths arc given by
Ksemaraja. The two parts o f the doors are termed here kantha (the
thinner, i.e., inner part) and kapoJa (the wider part). The elem ents
next to the door are termed upasobha (the wider or inner part next to
the kantha) and sobha (the thinner or outer part being next to the
kapola). A ccording to Ksemaraja, the height and the width o f the
kantha is the same, measuring h a lf o f the vithi. This m eans it is a
small square whose side is equal to the quarter o f the side o f a lotusseat. A lthough the size o f the kapola is not defined by the text,
Ksemaraja referring to other scriptural prescriptions understands
it to be o f the size o f h a lf a vithi by one vithi. This means it occupies
twice the space o f the kantha as reconstructed in C olour Plate 18,
V erse 34ab informs us that, there are eight doors, and K sem araja
explains that they are in between the lotuses. Since the size and
places o f the doors are determined, what is left between two doors on
each side is the obha with the upasobha. After constructing the
sobhas and the upaobhas o f the same size on each side o f each door,
the remaining parts form the four corners.
H ow ever, it is possible that in the text, kapola m eans w h at is
added to the size o f the kantha outside the kantha. This is suggested
by the name kapola, cheek. In that case, what is outside the kantha
measures three times the square o f the kantha. This would result in
221
s lig h tly different d o o r-s h a p e s, and the form s o f the sobhas,

upasobhas and corners would also change.
As for the colours, Ksemaraja points out that since they are not
specified, one has to resort to other Agamas. On the colours, he cites
th e Saiddhantika Parakhya, which has the same prescription for the
lotuses as the NT, but adds that the space betw een lotus petals is
green and the corners are red. The fibres are slightly different, having
the colours white, yellow and red starting from inside. I have made
the doors and the v ith i white, as in the case o f the N T s mandala, and
the sobhas and upasobhas red and yellow. K sem araja mentions that
the w estern door, facing the deity, should be left open or uncovered,
b u t since the covering is not explicitly prescribed in any graphic
form, I have not tried to supply it.
222
A p p e n d ix 3: The Trident Man (hi In

For the construction o f the trident mandala in the Siddhayogesvarimata, I have followed TA 31.155 with commentary as far as 1 have
been able to. The outline is to he made on a square o f three hastas on
each side, to which a surrounding hand o f 12 angulas is added for the
doors. The square o f three hastas on each side is divided into nine
large cells, one square hasta each, and these are further divided into 6
x 6 , i.e., 36 small cells each, so that each small cell m easures four
angulas on each side. The lotus in the middle, sim ilar lo the one
described in Appendix 1, measures one hasta on each side.
The trident is drawn in the large cell in the middle, i.e., leaving
three small bhagas untouched b elo w .1111 On both sides, tw o h a l f
circles are to be drawn downwards, in the neighbouring two bhaga s.
This implies that the radius o f the smaller half-circles is h a lf a bhaga,
and the radius o f the larger half-circles is one bhaga. The sm aller
half-circles continue in small arcs o f quartcr-circles above. The tip o f
the m iddle prong should end h a lf a hasta inside the large kostha
above the kostha o f the central lotus, i.e., half a hasta below the door.
The tips o f the two other prongs should fall on the side-lines o f the
large kosthas, i.e., on the lengthened side-lines o f the square o f the
central lotus. I have made these tips by lengthening the lines that
could be drawn between the centre point o f the m andala and the
points where the two small arcs end. The two points w here these
lines intersect the side-lines o f the large kostha arc the tips o f the
side-prongs. However, the exact drawing o f the tips o f the prongs are
left to the readers decision to some extent.
In addition to problems concerning the formation o f the prongs o f
the trident, I have also had problems in interpreting the way in which
the staff is to be drawn. As for its length, it readies down as far as
the edge o f the outer sq u are.111 It is said to be three hastas long
counting from the tip o f the middle prong, if I understand the text
correctly. This means that the distance between the top o f the trident
and the bottom end o f the staff is three hastas, which is indeed the
case if the staff reaches down to the edge o f the outer square. Its
thickness is two angulas, which is probably to be applied on each
130Jayaratha states: tatra m a d h y u d adhastanam bhagatrayain lyaktva.

131 See Ja y a ra th a s statem ent: parivarjilabahyadvrida& lngiilaiitani.
223
side, i.e., its full thickness is four angulas .132 T he s ta ff should not
co v er the lotus, o f course. It is not mentioned that the bottom end o f
the staff has got the thick ring called am alasaraka as in the MalinTv ija y a s version o f the mandala, nor that the staff is pointed below.
H ow ever, it is unlikely that the staff ended abruptly at the bottom,
a n d one is p ro b ab ly to draw the poin ted tip as w ell as the
am alasaraka. I have done so, assigning one bhaga to the pointed tip
as well as to the amalasaraka.
T he lotuses on the tips o f the trident m easure h a lf a hasta, i.e.,
th e ir radius is h a lf o f the radius o f the central lotus. la y a ra th a
m entions that the doors and the rest are to be fashioned as before.
T he last uncertainty concerns the central seat or pitha, which is not
m entioned, but which I have supplied, for to my know ledge there are
no occurrences o f a central lotus without a pitha around it. 1 have
ch osen the seat to be one bhaga thick, w hich w ould b e a standard
m easu rem en t, sim ilar to the m easurem ent o f the p ith a in the N T ,
w hich is described in Appendix 1, I have not draw n the small stick
like e lem en t under the central prong called g a n d ik a , w h ic h is
prescribed for the M alinivijayas trident in 31.67cd-68ab.
The colours are set out in verses 147ff. The colours for the central
lotus are the same as in the NT. I have also followed the N T w hen
colo u rin g the doors and their surroundings. The pitha is coloured
again as in the N T , for its prescription is the sam e as that o f the
M alinlvijayottara in TA 3 1 .80-81. I have followed verse 31.82 for
the colouring o f the trident, w hich is to be red, while its staff is black
and the amalasaraka is yellow. However, the v ith i is prescribed red
in verse 149 following the lost. Trisirobhairava. For the lotuses on the
tips o f the prongs, there may be two possibilities. One is that they are
the same colours as the central lotus. The other is that their colours
agree with the colours prescribed in the T risirobhairava: red, redblack/brow n and w hite for Parapara, A para and Para respectively,
i.e., on the right, left and in the c e n tre ,133 1 have follow ed the
T risirobhairava concerning the colours o f the small lotuses, but I
0 2 1 interpret the text to p rescrib e that the tw o vertical lines w h ic h fo rm the s ta f f
are to be m ark ed out on three points: in the lower, m id d le and upper part. T h e n these
po in ts arc p ro b a b ly to be con n ected in one single line on each side, b u t 1 am not sure
if this interpretation is correct.
131
S e e T A 31.118. T h e three colours basically agree w ith the c o lo u rs o f the three
g o d d e s s e s in the S id d h ay o g cS v arim ata; sec also S a n d e rso n 1990: 5 1 - 5 3 . L e f t and
right are to be s w a p p e d in the draw ing, see S id d h a y o g e sv a rlm a ta 6.24.
224
JUD1T TORZSOK
have not followed its prescript ions concerning the forms o f th e s e

lotuses. For the Trisirobhairava envisages the three small lo tu se s
with different numbers o f petals.
There is an alternative interpretation o f the description o f the trident,
which is equally possible. This interpretation w ould c h a n g e the
shape o f the upper part o f the trident, which w ould s o m e w h a t
resemble the reconstruction in Sanderson 1986: 171, except that it
w ould be much broader than the central lotus and that the s id e prongs w ould be curved, If this interpretation is fo llo w e d , the
thickness o f the upper part o f the trident is not determ ined at all in
the text., therefore I have followed the first alternative,
M A N D A L A S IN A B H I N A V A G U P T A S T A N T R A L O K A
Andre Padoux
In the T antraloka (TA), L ight on the T a n tra s , the vast treatise
A b h inavagupta com posed during the first years o f the 1 1th century,
w h e r e he expounds his own interpretation o f the notions and
practices o f the non-dualist Saiva system o f the Trika, mandalas are
m entioned a num ber o f times. However, while the whole o f chapter
31 in this work is devoted to these ritual diagrams, no general view is
given there o f the theory and practice o f the m andalas. It is only
th r o u g h his descriptions o f how m andalas are m ade use o f in
different rituals, and especially in the initiation (diksa) ritual, that
A b h inavaguptas conception o f the nature o f these devices appears.
C h a p te r 31 (163 slokas), on the nature o f m an d alas (.m andalasvarupam ), does not describe their nature, merely how to draw them.
It consists alm ost entirely o f quotations from earlier T antras. It
describes five different types o f mandalas, four o f w hich are m ade o f
tridents and lotuses (sulabjamandala), while one includes a svastika.
T h e descriptions are those o f four different Tantras, three o f w hich
have not com e down to us: the Trikasadbhava (also called T a n tra
sadbhava), the D evyayam alatantra, and the T risirobhairavatantra.
T h e fourth d escription is taken from c h ap ter 9 ( 6 - 3 0 ) o f the
M alinivijaya, the Tantra on which according to A bhinavagupta the
teach in g o f the TA is based; this text is still e x ta n t.1 W hile the
descriptions o f the Trikasadbhava and the M alin iv ijay a are clear
e n o ugh, those draw n from the two other T antras are d ifficult to
u n derstand (even with the help o f J ay arath as com m entary). O nly
tw o forms o f the Sulabjamandala can therefore be draw n w ith any
*
T h e E n g lish o f this p a p e r has b een ch c c k e d by M rs B a r b a r a B r a y w h o s e kin d
h e lp I w ish (oncc m ore) to a c k n o w le d g e very gratefully.
1
S e c B ib lio g r a p h y . T h e r e is n o th in g h e re , say s A b h i n a v a g u p t a in the first
c h a p t e r o f the TA (1.17) that is not c learly said, o r im p lie d b y th e g o d s in the
v e n e ra b le M alinlfvijayottaratantrn].
226
A ND R E PADOUX
certainty.2 The m ethod for draw ing the mandalas given in these
Tantras is the usual one, that is, to draw (heir lines with a pow dered
string on a pure, consecrated and oriented square surface, divided
usually into small square sections. Coloured powders may be added
once the pattern is drawn, so as to make it more beautiful, w hich is
something the deities like (9.41-42): one who knows how to do this
is a real m aster o f the Trika, says sloka 5 1. There arc also mandalas
m ade o f perfum ed substances, called g a n d h a m a n d a la , and less
frequently used.
W hat strikes one when looking at these diagrams is that they do
not co nform to the pattern generally considered as norm al for
mandalas, which are usually centred geometrical structures which the
user is to contem plate and/or to use for his w orship by going
m entally from their outward portion to their middle point; that is,
ontologically, from an outer lower plane to the higher central plane
o f the deity: it is a centripetal move. Here the mandalas are o f a
different type. On a square ground the main element is S iv as trident
( sula or trisula), whose staff goes vertically from the lower part o f
the m andala to its centre, where it expands in the form o f a lotus,
above w hich its three prongs rise. On the tips o f each o f these is a
full blow n lotus this is the trisuklbjam andala, the m andala o f the
trident and lotuses described in the M alinlvijaya. Or else it may
consist o f a vertical trident blossoming, in the centre o f the diagram,
into a lotus from which emerge on top and on the sides three lotustopped tridents, thus forming the tritrisulabjamandala, the m andala o f
the three tridents and [seven] lotuses (see Illustration I ):1 The mental
m o v e m e n t o f the user thus appears as an ascending one, or as
centrifugal: the m andala does not draw the user who m editates it to
its centre, but appears to invite (and induce) a fusion through ascent
to a higher level, or absorption into the shimmering lum inousness o f
a ra d ia tin g divine surface (a m a n d a la b ein g alw ays the r e c e p
tacle the adhara o f mantras/deities who are by nature luminous).
W hile these mandalas are different in their pattern, their ritual (and
2 T h is w as done by S tep h a n ie S anderson for P ro fesso r A lexis S a n d e r s o n s 1986
article M a n d a la and A g a m ic Identity in the T rik a o f K a s h m i r . T h is very eru d ite
p a p e r is to date the on ly th o ro u g h study o f the subject: the present b r i e f s u rv e y is
very largely indebted to it. Mrs S a n d e r s o n s d ra w in g s arc rcp ro d u c c d here w ith her
p e rm issio n (see Illustrations 1-3).
3 T h e r e s e e m s also to be a four trid en ts and [eight] lotuses m a n d a la ( c a lu s tr isulabjam andala).
M AN D A L A S IN A B H I N A V A G U P T A S T A N T R A L O K A
227
m e d ita tio n a l) role is not: different from that o f other diagram s: they
a r e structures on which to focus o n es attention, in which to perceive
t h e p re se n c e o f the deity or deities, in which to w orship them and
f in a ll y unite with them: the aim is the sam e even if the mental and
r it u a l course is different. This role o f mandala as a m eans o f fusion
w i t h the godhead is underscored by A bhinavagupta w h o for the
m a n d a la s described in the TA goes as far as to identify the mandala
a n d the supreme deity in TA 37.21 where he says: because the term
m a n d a [forms the word] mandala this word expresses the essence, it
m e a n s S iva (m andalam saram uktam hi m andasrutya sivahvayam ).
A s Jayaratha explains, the mandala gives (la d because o f la) the
e s s e n c e w hich designates Siva: m andalam id m andam siva h v a ya m
lailtyarthah.
T h e T A prescribes the use o f m andalas in various rituals. T he
m a n d a la is mentioned in TA 6 .2 -4 as one o f the sthanas, the p la c e s
o n o r w ith w hich rites are perform ed or mental concentration is
p r a c tic e d (the case in this chapter being the transcending o f time).
F o r the T A as for all other T antras, the m andala is the ritually
d e lim ite d and consecrated surface w here deities and supernatural
e n titie s are installed by their mantras and on w hich rites are to be
p erfo rm ed . If, however, a mandala is to be used in various rituals and
in ritual worship (puja), its more important use, in this treatise, is in
initiation (diksa) rites. Its role is so essential to initiation that seeing
th e m an d ala may mean being initiated. T A 4.49 and 13.152 quote
th u s from slo ka 18 o f the Paratrimsika: adrstam andalo p i even if
h e has not seen the m andala, which can be understood as m eaning
e v e n if he is not initiated.5 In this case, however, as A bhinavagupta
e x p la in s in his com m entary o f the Paratrim sika, the w ord m ay be
g iv e n several interpretations: it can be taken as referring to the
system o f bodily cakras or wheels o f pow er w here deities reside; or
to the secret ritual meeting o f Tantric initiates and Yoginis (m ela ka),
w h e r e the participants usually form a circle; or to the trisuiabjam andala seen during w orship or initiation, or p erceiv ed in o n e s
b o d y (as we shall see below). These interpretations all refer to cases
w h e r e the adept experiences mental cum bodily identification w ith
th e deity or its radiating power.
Since only an initiated (male) person can perform rites, we shall
look at the role o f the m andala first in initiation (diksa), then the
2 28
A ND R E PADOUX
occasional obligatory (.naim ittika) and the regular obligatory (nitya)

ritual worship.
C h a p te r 15 o f the TA describes in its latter part (436ff.) the
samayadiksa, the first degree o f initiation, by which one becom es a
sam ayin, an initiate who abides by the rules (samaya) o f the sect but
cannot perform rites. The procedure is a comparatively simple one.
For this, the sacrificial surface, the sthandila, is prepared by placing
mantras on it and worshipping them, installing powers and offerings
in vases, and then tracing a trisiilabjam andala, a m andala o f one
trident and lotuses (described in TA 3 1 .62-85, see Illustration 2),
w hich in this case is a gandhamandala (15.387),4 not traced w ith
coloured powders but with perfumed substances, The initiating guru
is to w orship on it the three goddesses o f the Trika alone or with
their consorts, or a mantradevatH (388), which he has installed there
by their mantras. He must then blindfold the initiand with a cloth on
which mantras have been placed and lead him to the m andala, make
him fall on his knees and cast flowers on the mandala, after w hich
the cloth is sw iftly rem oved: The initiand, suddenly seeing the
sacrificial area illuminated by the supernatural pow er o f the mantras
(m antraprabhavoJlasite sthale), is possessed by them and identifies
w i t h t h e m ( ta d a vesa va sa c c c h isy a s ta n m a y a tv a m p r a p y a te )
(15.451b^452a). As a lover perceives directly the virtues o f his
beloved, in the same w ay [the initiand], made perfect by the descent
o f divine grace (saktipatasam skrtah), experiences the presence o f the
m antras (m antrasannidhi) (452). This direct luminous and purifying
vision o f a mantra pantheon5 enclosed in the mandala is only the first
step in the initiation ritual, but it can be seen as the basic one since
this initial em pow erm ent o f the initiate will not only m ake him a
m e m b e r o f the sect, but will also induce a p erm a n e n t state o f
id en tificatio n w ith the deities o f the m andala, a state that will
henceforth form the basis o f his initiatic spiritual life. Though one o f
the terms used here to denote the condition o f the initiand is avesa,
possession, he does not appear to be expected to fall in trance as is
the case in the K aula initiation described in chapter 29, w here the
4 A g a n d h a m a n d a la is_also used in the w o rs h ip o f the g u r u w h ic h ta k e s p lace

b e fo re the sa m a ya d iksa (TA 15.387).
5 T a n tric p a n th e o n s arc as m uch (or perh ap s m o re ) stru ctu red g ro u p s o f m a n tra s
as g ro u p s o f deities.
229
m a n t r a s are so powerful that the initiate, m erely by seeing them , is

p o s s e s s e d and falls unconscious on the ground.
T h e next step in the Saiva initiation is that o f the putrakadiksa,
( a l s o nam ed viSesadiksa, special initiation, or nirvanadlksa, libera
t i n g initiation) which transforms the sam ayin into a [spiritual] s o n
o f h is guru. It is exam ined in chapter 16 o f the T A , w hose d e s
c r i p t i o n (based on the teaching o f the M a lin iv ija y a and other
T a n tr a s ) is more detailed than the preceding one. The ritual begins
w i t h the drawing o f a (ritrisulabjanMiidala: W hen the m aster wishes
t o p ro m o te a sam ayin to the slate o f putraka, o f sadhaka, or o f m aster
( d e s ik a ), he must first perform the preliminary purification, then, the
n e x t day, draw the m andala in the same w ay as for the com posed
s a c rific e ( sam udayikayageY' and elsew here (12a). The use here o f
t h i s m o re com plex m andalic structure, w h e re m o re deities are
in stalled , m ay be taken as showing that this initiation is o f a higher
o r d e r than the first one. Once the mandala is drawn, the triad o f the
T r ik a suprem e goddesses, Para, Parapara and Apara, are to be placed
( b y th e ir m antras) on the prongs o f the three tridents and be fully
w o r s h ip p e d (purnam sam pujitam ). T hen, the m an d ala (w h ich is
a p p a re n tly not made with powders) must be cleaned with a perfum ed
c lo th (7b). After which the master, having bathed, worships, in front
o f th e m an d ala, the deities o f the external retinue, then, on its
d o o r s , the deities o f the doors ( dvaradevata), then, going from the
n o r th - e a s t to the south-east, he w orships G anesa and o ther gods
d o w n to the ksetrapaias ( 8 - 9 ) . W e m ay note here that such
p r e s c r ip tio n s as these show that the m a n d a la is a ra th e r large
structure. The guru is now to worship with flowers, incense and other
o f fe r in g s the deities installed in the m an d ala, starting w ith the
adharasakti, at the base o f the trident, and up to Siva at the tip o f the
trid e n ts , the ritual being perform ed on each o f the three tridents.
P a ra , Parapara and Apara with their accom p an y in g B hairavas are
th u s w orshipped on the lotuses w hich are on the tips o f the three
trid en ts, then the transcendent goddess M atrsadbhava w ho is also
P a r a in the cen tral lotus w h e re she a b id e s a c c o m p a n ie d by
B h a ir a v a s a d b h a v a . Para, the su p rem e divine p o w e r, b e in g thus
ce n trally placed on all the tridents, is conceived as p erv ad in g the
T h i s re fe rs to the m a n d a la used in the so -c a lle d in n e r s a c r if ic e d e s c r ib e d in
c h a p t e r 15 ( 2 9 5 b 365) o f the TA . T h is ritual is called c o m p o s e d , o r c o m p le x , since
it b r i n g s to g e th e r several different elem ents.
230
A N D R E PADOUX
m andala, w hich is entirely full o f her presence (susam purnas

tadadhisthanamatratab 16). Several deities present in the m an d ala
are now to be worshipped. Several other rites follow, meant to infuse
in the initiand the pow er o f the goddesses o f the mandala7 and bring
him to enter the path o f non-duality. A practice is also d escrib ed
( 2 3 -2 6 ) by w hich the initiating master, penetrating then leaving
m entally (through a pranayama practice) the deity present in the
m andala, experiences an identification o f his self with it (m a n d a latm aikyanusandhana), to use Jayarathas expression (volume 10, p.
10). T hus pervaded with the p ow er o f the m andala,8 that is, th e
pow er o f the mantras placed in this diagram, he will be all the m ore
able to transmit this transforming power to the initiand, leading him
from the low er condition o f a sam ayin (or sam ayadiksita) to the
higher one o f putraka (or to the state o f sadhaka, if he is a bubhuksu,
one who seeks pow er or supernatural rewards through the m astery o f
a mantra).
M andalas are also used in other forms o f initiation described in
the TA. A mandala is used, for instance, in the funeral rite ( antyesp),
a kind o f initiation rite, where it is to be drawn in the house o f the
dead person (2 4 .1 0 -1 2 ) before the funeral rites are perform ed. It is
used, too, in the initiation o f somebody w ho is absent (whether aw ay
or dead), briefly d escrib ed in ch ap ter 2 1 . In this case, a fte r
prelim inary purifications o f the ritual place, o f the rice used in the
ritual, o f the disciple and o f the mandala, the latter is used to bring
a b o u t th e p r e s e n c e o f th e m a n tra s and to s a tis fy th e m
(m antrasam nidhisam trpti), since these are the powers that are to be
p ro p itia te d to initiate the disciple who is absent: as J a y a ra th a
com m ents, the m andala protects the disciple even if it is not seen by
him. It is useful in spite o f the fact that it is only one am ong eleven
elements used to perform that initiation (21.13-15), the ritual having
to be performed as richly as possible so as to satisfy fully the powers
invoked in the m andala. This m andala is the trilrisuiabjaim ndala,
which, A bhinavagupta says (21.19-20), is so powerful that sim ply
by seeing it, without even propitiating the mantras placed on it, [the
disciple] becom es a sam ayin (m antraniandale anahute p i drstam sat
1 S a n d e rs o n 1986: 197 s h o w s h o w A b h in a v a g u p ta s u p e r im p o s e s on the T r ik a

d eities o f the m a n d a la the fo u rfo ld s e q u e n c e o f the K ra m a tradition and e v e n the
s y ste m o f the tw elve Kalis.
* T A 1 7.1-3 un d erlin es the identification o f the initiating guru and the m andala.
M A N D AL A S IN A B H I N A V A G U P T A S T A N T R A L O K A
231
sam ayitvasadhanani) this sentence, however, is probably not to be

t a k e n literally. The likeness o f the missing person used in the rite, as
w e l l as the mantras, must be luminous ( akrtir dlptarupa ya m antras
ta d va i). After the mandala has been drawn and the deity worshipped,
a d d s the TA (22b-24), the guru must make with kusa grass and cowd u n g an image o f the disciple he is to initiate, in which he will instil
t h a t d is c ip le s mind (citta) so as to liberate him from his fetters
b e f o r e the rest o f the ritual is performed.
T h e sam e m andala is used for the initiation described in chapter
2 9 , w h ich is that o f the Kaula Trika. This initiation is different from
t h e o n e given to ordinary disciples. It is given by the K aula guru to a
f e w chosen disciples only one in a hundred thousand, according to
T A 29 .1 8 7 that is, those who are able to perform rites (the K aula
sacrifice, kulayaga)9 where a feminine partner ( dutl) plays a role and
w h e r e the o fferings in clu d e m eat and liquor, and also sexual
secretions.'" Such rites are in contradiction to the generally admitted
r u le s o f purity the Trika adept is supposed to respect in his outer
s o c ia l behavior. It is therefore to be kept secret. This is repeated
s e v e ra l tim es in this chapter by A bhinavagupta, w h o states before
d e s c rib in g the part o f the ritual where the m andala is used (29.169):
th is c a n n o t be described clearly by me because it is secret (na
p a th y a te rahasyatvat spastaih sabdair m a ya ). T h e p a s sa g e w hich
f o llo w s (1 7 0 -1 7 4 ) is indeed quite obscure. I have not been able to
r e n d e r it very clearly in spite o f the help extended to me by Professor
A le x is S anderson whom I consulted on the subject. As Jayaratha
e x p la in s in his com m entary on a preceding sloka (p. 114), A b h in a
v a g u p ta refers here im plicitly to the doctrine that the teachings o f
B h a ir a v a have four foundations, nam ely mantra, vidya, m udra and
m a n d a la ," the case here b ein g that o f the m a n d a la , w h ic h is
identified with the body o f the perform er o f the rite. In the section o f
c h a p te r 29 (slo ka s 166177), concerning the secret practice w ith a
duti, the divinized body o f the Y ogin is the substrate on and w ith
v K u l a y a g a , a c c o rd in g to a c o m m o n u se o f the te rm y a g a in s u c h tex ts, m e a n s
a ls o th e p an th eo n o f the Kula.
10 It is the o ffe rin g (a r g h y a , argha) c alled ku n d a g o la o r k u n d a g o la k a w h i c h i n
c l u d e s the sexual secretio n s p ro d u c e d d u rin g the ritual by the a d ep t and his fe m in in e
partn er.
" M a n tra s (or v id y is , w h ic h are fe m in in e m a n tra s) are p h o n ic f o r m s o f the deity.
M u d r a s , in A b h i n a v a g u p ta s v iew (see T A , chapter 32) b rin g a b o u t the identification
o f t h e p e r f o r m e r w ith th e deity w h ic h is m a d e p r e s e n t b y a ris in g fro m th e im a g e
(p ra tib im b a ) th u s p roduced. O n this see P a d o u x 1990: 6 6 - 7 5 .
232
A N D R E PADOUX
w hich the ritual is performed. The passage we are concerned w ith

runs as follows: As ail have [a body], so have the god and the
goddess [that is, the Yogin and his duti], This [body o f theirs] is the
supreme wheel ( tac cakram paramam) by which the goddess and the
pantheon (yaga) are m ade present (170). The body is indeed the
suprem e icon (deha eva param lirigam), made o f all the tattvas.
Auspicious, it is the highest place o f worship for it is occupied by the
wheel o f all the deities (171). It is this [body] which is the suprem e
m andala, m ade o f the three tridents, the [seven] lotuses, w heels or
voids (kha).12 There and nowhere else m ust the w heel o f deities be
c o n sta n tly w o rsh ip p e d , e x tern ally and internally ( 1 7 2 ) . 13 [The
perform er] should first concentrate mentally on the mantra o f each
[deity] (svasvam antraparam arsapurvam ), then touch [himself] with
the richly blissful fluids that are produced from [the b o d y ,14 this
being done] following the order o f emission and that o f resorption
(srstisam haravidhina) (1 7 3 ).15 By these contacts, the field o f o n e s
consciousness is aw akened and, becoming the m aster o f that [field],
one reaches the highest dom ain (paramam dharnan), having satisfied
all the deities [which animate his senses and body] (174). [Then the
Yogin] should gratify these [deities] in [his body] with all the heartravishing substances and by concentrating on each o f their [mantras],
following for this the procedure laid dow n for the auxiliary w orship
( anuyagoktavidhina) (175). A bhinavagupta then sings the praises o f
the m andala-body: In the divine abode o f the body (dehadevasadane), I w orship you together, o my god and goddess, night and
day, with the blissful nectar that fills the vase o f offering o f the heart,
w ith the unm ediated flowers o f the spirit which spread their native
natural fragrance, and by sprinkling over the world, bearer o f all,
with the pure essence o f my wondering ecstasy (Gamath-ti'rasa)
The use o f m andalas is prescribed not only for initiations but also
for the perform ance o f other rituals. For instance, for the w orship o f
the new ly consecrated master, the gurupuja, w hich is to be per-
12 Ja y a ra th a glosses k h a b y vyo m a n which m e a n s space, void. W h ic h void is this?

I do n o t know .
13 T hat is, by m a k in g offerings (m eat, liquor, sexual flu id s) to the deities on the
m an d ala, and by c o n su m in g them.
14 Th is refers to the kundagolaka, see note 10.
15 T h a t is to say, b e g in n in g with the h ig h e st deity at the c ro w n o f the h e a d and
p ro g re s s in g d o w n w a r d , o r b e g in n in g w ith the o u te rm o st deities o f the m a n d a la , at
the feet, and p ro g re ssin g upwards.
233
f o r m e d at th e end o f the initiation or consecration (abhiseka) ritual,

t h e g u r u is to be seated on a seat o f gold, etc . ( h a im a d ik a sa n a m ),
p l a c e d on a mandala on which a svastika is drawn (TA 2 8.425-426).
W h a t is the pattern o f this svastikam andala is not specified there, but
i t is p ro b a b ly the pattern described in chapter 31, 132-154. Slokas
1 4 7 154 o f that chapter give precise details o f the aspects and
c o l o u r s o f th e lotus petals and svastikas o f the m andala, w hich m ust
n o t o n ly be brightly coloured but also adorned with precious stones.
T h e p assag e ends: the sanctuary o f the god o f gods who satisfies all
d e s i r e s m u st be outwardly all red and shining (jvalaruna). T hat the
m a n d a l a should be bright, shining, is often said, but it is difficult to
g a u g e h ow b rig h t these diagrams really were. To be sure, they were
b r i g h t l y co lo u red , adorned with flowers, p erh ap s also gold and
p r e c i o u s stones, but often the b rightness or even the fulguration
m e n t i o n e d in the texts is that o f the mantras placed in the m andala,
n o t th a t o f the diagram itself: a brightness, therefore, w hich was
p r o b a b l y m entally perceived (shall we say im agined?) rather than
a c tu a lly seen.
A n o th er, more interesting use o f the mandala, perceived as present
i n th e body o f the adept, is described in the 15th chapter o f the TA.
T h e r e , the m andala is not identified with the body. It is felt to rise
w i t h i n it and then to o v erg ro w it, thus b rin g in g ab o u t a total
s u rp a s s in g o f bodily existence and consciousness. This takes place
d u r in g the first part o f the daily ritual w orship o f the Saiva adept,
w h o , b e fo re perform ing the external phase (bahya) o f the w orship,
t h e p u ja , is to place mentally in his body the pantheon o f the m andala
a n d to identify h im self mentally with it: as the saying goes, n adevo
d e v a m arcayet: the officiant cannot worship a deity if he is not first
f o rm a lly deified. Here, how ever, the deification is o f a very parti
c u l a r and especially intense sort since the adept is expected to trans
c e n d m entally his identity and limited consciousness by realizing it
t o b e identical w ith the non-in d iv id u al d iv in e c o n s c io u sn e s s, a
p r o c e s s that will fuse him with the unlim ited pow er o f the suprem e
g o d h e a d . T h ro u g h this practice, to quote A lexis S an d erso n , the
p e r f o r m e r ritually internalizes a m etap h y sica l o n to lo g y . 16 T h e
d i a g r a m w h ic h is u s e d to this end is th e trisulabjamandala
(Illustration 2), to be visualized by the adept as present in his body,
lfi S a n d e r s o n 1986: 172; pp. 1 7 2 - 1 8 2 o f th is s t u d y d e s c r i b e a n d e x p l a i n
t h o r o u g h l y the process and m e a n in g o f this ritual m ental w orship.
234
A N D R E PADOUX
with all the cosmic entities and deities present in it. The procedure is
as follows: first, controlling his vital breaths, the adept fuses his
prana and apana breaths in the samana breath, which is then burnt by
the ascending udana breath blazing up from below his navel along
the susum na up to the d vadasanta (or urdhvakundalini), the subtle
cen tre d e e m e d to be p laced tw elve fin g e r-sp a c e s a b o v e the
brahmarandhra, therefore above his body. This awakening and rising
o f the kundalini is the p relim in ary condition for the in tern a l
installation o f the m andala and for the worship o f its deities. In this
Y o g ic state o f trance, w hich cuts him o ff en tire ly from th e
surrounding world, the adept is to visualize the iris id a bjama n da 1a as
present in his body (see Illustration 3). Four fingers below his navel,
he places m entally the sw elling at the base o f the trid e n t and
worships it as being the adharasakti, the pow er w hich supports the
cosmos which he feels as present within him. Then he imagines (and
worships as an ascending m ovement toward the deity) the s ta ff o f the
trident which he sees mentally as rising in his body above the navel
along the vertical axis o f the susum na up to the subtle centre o f the
palate ( talu) through the 25 tattvas constituting the world, from the
earth-tattva to those o f purusa and the kancukas, which are tiered
along it. Thus all the constituents o f the manifest, impure (asuddha)
world are present in the adept, constituting the throne o f the Trika
deities. A bove the palate, he visualizes the k n o t (granthi) o f the
trident, identified with the m aya-tattva, then he visualizes its plinth
( catuskika) together with the suddha-vidya-tattva, the first level o f
the pure universe (suddhadhvan) w hich begins there and extends
above m aya up to Siva. On this plinth he imagines an eight-petalled
lotus o f gn o sis (vidyapadm a) as the Isvara-tattva. In the centre o f
that lotus the adept now mentally installs Sadasiva (the 34th tattva),
v isu alizin g him as a blazing corpse (the so -called M ah ap reta),
em aciated because he is void o f the cosmos, gazing upw ard tow ard
the light o f the absolute and laughing b o istero u sly (attahasa).17
S a d asiv a m ust be w o rsh ip p ed as m ad e up o f tw o and a h a l f
syllables'8 and as do m in atin g everything, T h e ad e p t is no w to
visualize the three prongs o f the trident rising up through his cranial
aperture (on the phonic level o f nadanta) from Sadasivas navel and
17 T h is m a d laughter is a characteristic trait o f fearso m e T a n tric deities. It is also
to be used by adepts during certain rituals. T h e practicc goes back to the PaSupatas.
ts A s no ted before, deities are m antras or m antras deities.
235
g o in g up to the dvadasanta. These prongs are deem ed to go through

t h e su b tle levels o f resonance and co n scio u sn ess n a m e d sakti,
v y a p in l and sam ana,19 On the tips o f the three prongs o f the trident
( o n the level o f the dvadasanta and on that o f unmana, the tra n s
m e n ta l plane), he visualizes three white lotuses. On these he first
en th ro n e s the mantras o f three Bhairavas, conceived o f as lying on
th e lotuses, and then, seated on the B hairavas, the three suprem e
goddesses o f the Trika: first, on the central prong, Para, the supreme,
w h ite , luminous, benevolent, pouring amrta; then, on her left, P ara
p a r a , the interm ediate, red and w rathful; and on h e r o th er side,
A p a ra , the lower, dark-red, furious, terrifying. The three goddesses
are garlanded with skulls, hold the skull-staff, etc. N ow , the adept,
se e in g these goddesses m entally, m ust w orship them together w ith
th e ir retinues, offering them (since this is a purely m ental process)
th e transcendental consciousness he has o f the fact that the universe
is an expansion o f the divine pow er and that his ow n consciousness
is totally fused into this divine, om nipresent reality. To perform this
Y o g ic practice o f the m andala is thus to experience the identity o f
th e s e lf and o f the absolute. All the fantasm agory visualized in this
w a y leads the Y ogin to feel dissolved into the transcendental void o f
th e absolute whilst being also inhabited in his body by the cosm os
a n d its presiding deities. It is an interesting, but strange, process. I f
w e c o n sid e r that this Y ogic, visionary trance-like state o f bodily
consciousness is to be experienced every day by the Saiva adept, we
m a y well w o n d er w hat psychological condition is thus induced in
him , w hat kind o f perception o f the world he lives with. Can one feel
fu sed with the absolute after having filled o n e s m ind with such a
fantastic scenery and still behave norm ally? O f course, these ritual
practices m ay have been perform ed m erely in im agination w ithout
an y real inner participation o f the Yogin, They m ay even have been
lim ited to the mere recitation o f the mantras evoking the tattvas and
th e deities (m antraprayoga). But w h at i f they w ere really e x p e
rienced? W hat if the Kaula adept, practising the ritual at least once
every day, carried always in him this scenery? This is an interesting
question but not one to be answered here.
19
O n the s u b tle le vels o f e n u n c ia tio n ( uccara ) o f a bijamantra , fro m bindu to
unmana, see P a d o u x 1990a: 4 0 4 - 4 1 1 . H e re as in o th e r cases the p la n e s (kala) o f
p h o n i c u tte r a n c e ( uccara ) arc tak en as a su b tle p r o lo n g a tio n o f th e le v e ls o f the
c o s m o s (tattvas).
236
ANDREPADOUX
1. Outline of the mandala o f the three tridents and (seven) lotuses (tritriiulabjamandala) prescribed by
the Trikasadbhavatantra
M A N D A L A S I N A B H I N A V A G U P T A S T A N T R A L O K A
237
2 . Outline o f the mandala of the trident and lotuses (trisulabjamandala) prescribed by the
Mali nlvij ayottaratantra
238
A NDRE PADOUX
PARAPARA
on
PARA
on
APARA
on
R a i.is e k h a ra b lm ir a v a
B lia in iv a s a d b h iiv a
!\T;i v;Uni a b b a fra v a
U nm ana
3 7 ...........
S am anS
S i v a t a t l v a m (36)
Vyapinl
SaktitaUvam (35)
Sakti
Sad asivatattvam ( 3 4 ) . . . .
I S v a r a l a t t v a m ( 3 3 ) .............
S u d dh a v id ylla tlv a m (3 2)....................
MSyfilatlvam (31) ................................
S adasiva M a h a p re ta
, . . L o tu s of G nosis
.................................. P lin th {calti$kikd)
................................
K n o t (granlhih)
of b a n n e r
6-3 0
1- 5
3. V isualization o f the m andala throne o f the three goddesses o f the Trika
T H E SRICAK RA A C C O R D IN G TO TH E FIRST C H A P T E R OF
T H E YOGINTHRDAYA*
Andre Padoux
T h e sricakra the m andalic form o f the goddess TripurasundarT,
sy m b o lic o f her cosm ic activity is too well kn o w n to need d e s
cribing here. It is indeed so well known that ritual diagrams are often
believ ed to be all o f the same type i.e., centred m andalic cosm ic
sy m b o ls though, as Helene Brunner explained in her article, this is
not at all the case. The mandalas o f the Tantraloka, also described in
this book, are likewise o f a different type. C osm ic diagram s o f the
sam e sort as the Sricakra are to be found in the K ubjikam atatantra,
chapters 14-16, where they are to be visualized in the body, but not,
ap p aren tly , to be m aterially draw n and used for w o r s h ip .1 T his
resem b lan ce may be due to the links existing b etw een the ancient
K ubjika tradition and the som ew hat more recent Srlvidya, 2 to w hich
th e Y o g in lh rd a y a (Y H ) b e lo n g s it bein g , to g e th e r w ith the
V a m a k esv arim ata/N ity aso d asik arn av a, one o f its tw o basic texts.
T he description o f the sricakra in the first patala o f the YH is w orth
m e n tio n in g because, rather surprisingly, it does not say how the
diag ram looks and how to draw it,3 but describes its apparition, its
d e s c e n t ( cakravatara), as a divine c o sm ic process, an outw ard
c o sm ic m anifestation o f the pow er o f the g o d head w hich is to be
m editated, visualized, and even bodily experienced b y the adept. The
sricakra is show n here as a diagram m atic cosm ic vision rather than
T h e E n g lish o f this paper h as been- as u su al k in d ly c h e c k e d b y M rs B a rb a ra
B r a y to w h o m I am as a lw a y s very grateful.
1 S ee H eilijgcrs-S eelen 1994.
2 T h is tradition is also called T ripuradarSana o r S a u b h a g y a s a m p r a d a y a . T h e n a m e
S r l v i d y a (w h ic h is also the n a m e o f its mulamantra) is e s p e c i a l l y u s e d f o r the
m o d e m , ved an tized , form o f the tradition. O n th e links b e tw e e n the K u b jik a and the
T r ip u r a traditions, see D y c z k o w s k i 1988.
3 T h is is on ly briefly m e n tio n e d in th e third c h a p te r ( 3 . 9 5 - 9 7 ) , w h e r e the cu lt o f
th e g o d d e s s an d o f h e r re tin u e o f d eitie s th e sricakrapuja, w h i c h is also to be
p e r f o r m e d is d escribed in so m e detail.
240
A ND R E PADOUX
as a ritual diagram. The theological, metaphysical bias o f the Y H s

descrip tio n is em p h asized by the earliest o f the th ree k n o w n
com m entaries o f the YH: the DIpika o f Am rtananda (13th or 14th
cen tu ry ), w h o s e th o u g h t w as very m uch in flu e n c e d by the
K a sh m iria n n o n -d u a lis t P raty a b h ijn a system , th o u g h he w as
probably from South India. The YH itself is in all likelihood a w ork
from Kashmir, and may date from the 1 1th century.'1
The srlcakra, as is well known, is m ade up o f a central triangle
with a dot (bindu) in the m iddle, surrounded by four concentric
series o f triangles, themselves encircled by two concentric rows o f
lotus petals w hich in turn are encom passed by a threefold circle
en clo sed in a sq u are gro u n d form ing the o u ter p o rtio n (see
Illustration 1): nine parts altogether, each o f which is called a cakra.
T hese nine constitutive parts o f the srlca kra are regarded as an
expansion o f the divine pow er o f the goddess, wherein abide all her
different energies and all the deities emanating from her and forming
her retinue. (These supernatural entities embody and relay her power,
infiasing the srlcakra with it and therefore somehow constituting it.5)
The srlcakra as show n here is thus not a mere outline, nor a m ere
consecrated area, but a cosm ic event and reality, to be m editated,
realized, interiorized by the adept through the practice o f bhavana;
that is, by creative identifying meditation, a practice the fundamental
importance o f w hich in this context must be emphasized: the cosmic
event, the expansion and unfolding o f power o f the cakravatara is to
be so intensely visualized, imagined, and felt to unfold in the cosmos
as well as in the adepts mind a n d body, that he identifies with it.
A n interesting aspect o f the Y H s approach to the srlcakra is that
its th ree ch ap ters are called sam keta, the first one b e in g the
cakrasamketa. The u se o f this term , w h ich m eans a g re e m e n t,
a p p o in tm e n t, m ee tin g , u n d erlin e s the fact th at w h at is bein g
described there is not a mere diagram, but the diagram m atic aspect
and result o f the meeting, the union o f the goddess Tripurasundari
4
See P ad o u x 1994: 4 2 ff. T h e YH m a y have b een influenced b y the Pratyabhijna.
s K a sh m iria n S aiv a authors, such, for in stance, as K sem araja, tend to distin g u ish
b e t w e e n y an tra, c o n c e iv e d o f as a pattern o f m a n tra s/d e itie s, and m a n d a l a as the
actual, v is ib le stru c tu re . T h is , as s u g g e s te d b y S a n d e r s o n in his c o m m e n t s on
B r u n n e r s c o n trib u tio n to P ad o u x 1986: 33, w o u ld h a v e as a c o n s e q u e n c e that it is
only the structure that one can tr a c e , that is the m an d ala, and that (I q u o te) w h e n
one speaks o f the mandala to include the circle o f deities (devatacakram) or m an tras
( mantracakram ) w o rsh ip p e d in it, then this is by extension o f the prim ary sense,
T H E S R IC A K R A
241
a n d o f her consort, &iva/Bhairava, the com m on united presence o f

t h e s e tw o aspects o f the supreme godhead in the sricakra bringing
a b o u t its apparition and endowing it with their unlim ited glory and
pow er.
A fte r eight introductory stanzas, the first chapter begins: W hen
s h e , th e supreme power, [becoming] by her own free will em bodied
a s all that exists ( visvarupinl), perceives her own throbbing radiance
( sphuratta), the cakra is then being produced . 6 It is a cosmic event:
t h e g o d d ess is taking on her cosmic fo rm .7 The YH then describes
t h e developm ent and play o f the divine energies o f the goddess, from
t h e bindu in the centre to the outer square delim iting the sricakra.
T h e bindu is said to issue from the initial, void ( sunya) phonem e a,
w h i c h is the absolute. It is described (slokas 11-12), not as a m ere
d o t, n o r as the place in w hich to visualize the goddess, b u t as
th ro b b in g consciousness whose supreme nature is light and w hich is
u n i t e d with the flashing flow [of divine p o w er], bein g the seat
( baindavasana) which is the [birth]place o f the flow m ade up o f the
th re e m atrkas. It thus assum es the form o f the threefold m atrka
w h ic h is to say the three planes o f the word, pasyanti, m adhyam a and
v a ikh a rl which appear together with the inner triangle o f the sricakra.
T h e n appears the cakra o f eight triangles know n as n a va yo n i because
it is considered as being made up o f the central triangle plus the eight
th a t surround it, nine in all, and because it is the origin, the y o n i from
w h i c h the follow ing cakras are born. It is d escrib ed as a huge
c o m p a c t mass o f consciousness and bliss (cidanandaghanam m ahat),
ab so lu tely pure, transcending time and space: a cosmic vision, not a
m e r e outline. Then, by an inner process o f tra n sfo rm a tio n and
interaction, the other constituting cakras o f the sricakra appear, each
d e s c rib e d as luminous, in each o f w hich goddesses or m antras and
p h o n e m e s are deem ed to abide (and are to be im agined as present),
e a c h also associated with tattvas down to the level o f the earth tattva,
o n the outer square. The cakras correspond, too, to the divisions o f
th e cosm os called kala, from the santyatltakala, the highest, in the
centre, to the nivrttikala, that o f the earth, in the square part. This is
to be expected since the sricakra as it appears or unfolds is an image
h
yadii sfi parama saktih svccchaya visvarupini /

sphurattam atmanah pasyct tadii cakrasya sambhavah U
1
In the n o n -d u a listic vision o f Sa iv ism , the co sm ic m a n ife s ta tio n (srsti), th o u g h
e m i t t e d , r e s u lts f r o m an a c t o f c o n s c i o u s n e s s o f th e g o d h e a d a n d r e m a i n s
o n to lo g ic a lly w ithin it.
242
A ND R E PADOUX
o f the cosmos in statu nascendi, extending from the godhead to this

world. The descending5 cosmic structure o f the srlcakra appears also
in the fact that its nine cakras are regarded as forms or creations o f
the powers (sakti) o f Siva and from this point o f view are considered
as divided into three groups, deemed respectively to correspond to
the powers o f will (iccha), knowledge (jnana) and activity ( kriya) o f
the deity ,8 The cakra, concludes sloka 24, is thus threefold. It is
an aspect o f k a m a k a la 9 and is in reality and essence expansion (prasaraparamarthatah).'n
A fter this first cosmogonic phase, this cosmic vision, the adept
must now (slokas 2536) turn to a different sort o f mental exercise.
He is to m ed itate (bhavayet), that is, visualize the nine portions o f
the srlcakra, from the outer square to the central bindu, as present in
nine centres (here called padnia) o f his Yogic subtle b o d y . 10 These
nine centres are to be visualized with their shapes, colours and
residing deities as tiered along the susum na, from the akulapadm a,1'
situated at its base, where he mentally places the outer square (called
the trailokyamohanacakra), to the ajna, between the eyebrows, where
he im agines the central bindu. The m o v em en t is therefore now
centripetal: the adept not only feels identified with the srlcakra and
imagines or perceives it in himself; he also follows mentally, linked
with his Y ogic im aginary inner structure or b o d y , an ascending
m ovem ent towards the centre o f the srlcakra and thus to the suprem e
goddess. The m ovem ent does not, however, stop in this centre but
takes on a different and more subtle (niskala) form, for the adept is
n ow to m editate the kalas, the subtle phonic p a rts the subtle
8 T h e s e a re in this o rd er o f d e c re a s in g status the three p o w e rs or e n e rg ie s o f

S iva through w hich he m an ifests the universe.
9 T o say that the srlcakra is kamakala is to say that it is m ad e u p o f the c o m b in e d
presence o f Siva (kama) and Sakti (kala ).
10 T h e p attern o f centres (cakra or granthi) and canals (nadf) w h ich the Y o g in is
to visualize as p re se n t in his bo d y and w h ere the prana or the m antras flow , is often
c a lle d the s u b tle b o d y , b e c a u s e it is a v isio n a ry , n o t an a n a to m ic a lly e x is tin g
structure. B u t this is w ro n g becau se subtle b o d y is a translation o f suksm adcha (or
suksmasarlra), the t r a n s m ig r a tin g p o rtio n , m a d e o f tattvas, o f the h u m a n b e in g :
s o m e th in g quite different.
" T h is centre is sp e c ific to the Y H (or to the T rip u r a tradition). S in c e there are
nine co n stitu tin g cakras in the srlcakra, to be visualized in the b o d ily centres, these
m u s t o f n e c e s s ity be n in e in n u m b e r. A b o v e the a k u la p a d m a th e r e is th u s a
kulapadm a, a lam bikapadm a (on the uvula) bein g a d d e d b e tw e e n the visuddha and
the a/na (see Illustration 2),
THE S RI CAKRA
243
fo rm s o f phonic energy, that is, o f the bljamantra h rh n ,'2 starting

w ith the bindu (the anusvara following the m o f the mantra). He m ust
v isu a liz e and meditate all the other kalas: ardhacandra, nirodhini,
nada, nadanta, sakti, vyapini, samana and un m a n a13 o f the uccara o f
b rim , to w hich is added in fine the so-called m ahabindu, w hich
transcends space, time and form: an utterly transcendent aspect o f
th e p h o n ic vibration. In this m ental practice th e adept, h a v in g
m ed itated and visualized each o f the nine parts o f the sricakra as
p re se n t in each o f his bodily centres, is now to perceive the central
b in d u not as the centre o f the sricakra but as the first phonic kala o f
h rlm , thus shifting from a spatial type o f m editation to a m ore subtle,
p h onic one. This meditation is in fact not purely phonic since all the
kalas have a visual aspect or symbol w hich is also to be evoked:
these aspects are described in slokas 2734, w hich also m ention the
le n g th o f tim e during w hich they are to be m e n ta lly u tt e r e d
( uccaranakala). These fractions o f time are so m inute (ranging from
l/ 4 t h to l/256th o f a m atra'4) that they cannot possibly correspond to
an y actual utterance. T h e y rather suggest the uccara"s g ro w in g
d eg ree o f subtleness. They express or correspond to a progression o f
th e Y ogin towards a total transcending o f all empirical reality either
o f form or o f sound/word. The adept is indeed finally to reach the
tra n s c e n d e n t plane o f w hat is called m ahabindu, w h ere he is to
m e d ita te and fuse with (to quote A m rtananda) the suprem e Siva,
su p rem e light, the pow er o f suprem e awareness that is the suprem e
g o d d ess M ahatripurasundari. It is therefore a totally non-m aterial
(n iska la ), m y stical e x p e rie n c e o f the s u p re m e g o d h e a d . T his
tra n s itio n from a d iag ram m a tic, spatial or visual d im e n sio n o f
m editative practice to a phonic, m antric one, with the attainm ent o f
the suprem e plane o f the deity, is expressed in sloka 36, w hich is as
fo llo w s : W h en this s u p re m e e n erg y (param a kala) sees th e
effulgence o f the self (atm anah sphuranam ), she assum es the aspect
o f Ambika: the supreme word (para vak) is being uttered. 15
T he notion that the meditation o f a m andala should lead the adept
to see or participate in the pow er o f a deity is not u n com m on. The
T h e T r ip u r a / S r i v i d y a tra d itio n has b o th a p a r tic u la r c a k ra an d a p a r tic u la r
m an tra, the 15 p h o n e m e s grividya, the three parts o f w h ic h al l en d w ith the bija hrirn.
13 O n the kalas o f hrhn, om , etc., sec P ad o u x 1990a: 4 0 2 - 4 1 1 .
A matra or m o r a is, in Sanskrit, the duratio n o f a short vowel.
15
atmanah sphuranam p a syed yada sa parama kala /

am bikarupam apanna para vak sam udtiita It
244
AND RE PADOUX
notion that it should bring about an identification (sam arasya, says

the com m entary) with the supreme plane o f vac is less frequent; this
is perhaps a further proof o f the Saiva Kashmirian origin o f the YH,
Arnbika, being the suprem e m other and suprem e level o f the
word, is the source o f the cosmos. The sadhaka having m entally
attained this level is now (slo k tB 3 7 -4 9 ) to im agine again the
suprem e g o d d e s s s intent on m anifesting the universe w ith all it
contains, a cosmic process conceived, however, as developing along
the pattern o f the srlcakra. To quote slokas 3 7 -4 0 , W hen she is
about to manifest the universe which [she holds within herself] as a
germ, assuming an oblique aspect, [she becomes] Varna because she
vomits the universe ( visvasya vamanat). Then, as the energy o f will
(icchasakti), she has the visionary [word] (pasyantf) as her body.
W hen she is the pow er o f cognition (jhanasakti), she is Jyestha, and
the interm ediate w ord ( m adhyam a vak) is then uttered. W hen the
m aintaining o f the universe prevails, her figure spreads out into a
straight line. Then, in the state o f resorption she takes on the shape o f
the bindu. W hen the reverse process takes place, her body becom es
[shaped like] a srngataka.'6 She is then the p o w e r o f activ ity
(kriya sa kd ): she is Raudri, the corporeal [word] (vaikhari), appearing
as the universe.
W hat the adept is to realize here is the first creative m ovem ent o f
the suprem e goddess m anifesting the inner triangle, together w ith
four form s o f energy and four divine form s o f herself, w hile
retaining all this within herselfhence the fourth goddess, Raudri,
and the return to the inner bindu.
In or around this central triangle other entities are now to appear,
w ho like the preceding ones are to be conceived o f both as existing
in the cosmos and as abiding in the Srlcakra. First (slokas 4 1 -4 3 ) are
produced the four plthas, the sacred seats o f the goddess, Kam arupa,
Purnagiri, Jalandhara and Qdyana, described here not as abiding in
the central triangle (called the sarvasiddhim ayacakra) but as being in
the Y ogic im aginary body (in the m uladhara, the heart, the bhrum adhya and the brahmarandhra) o f the adept: the process, as we
have already noted, is inseparably cosmic, diagrammatic and Yogic,
these three aspects being both imagined, visualized (the colours and
shapes o f the plthas are described) and bodily experienced.
T h e srngataka is th e trapa bispinosa , the w a t c r - c h e s t n u t , w h o s e fru it is
triangular in shape. T h e word srngataka is therefore used to m ean a triangle.
THE S r ICAKRA
245
T h e n four different lingas ( svayam bhu, bana, itara and para) are
im a g in e d as being eacli in one o f the four plthas, each being o f a
d iffe re n t colour and aspect,17 and each associated w ith different sets
o f Sanskrit phonemes, so that the whole pow er o f vac in the form o f
th e Sanskrit alphabet abides in them (41-44). All these elements, the
m a trka , the plthas and the lingas, are described as being expressed
( v a cya ) by (that is, as produced by) the mantra o f the goddess, the
srividyo, which is taken as being fourfold (that is, the mantra as a
w h o le plus its three parts), and are considered as corresponding to
th e five conditions or states o f co nsciousness ( avastha), jagrat,
svapna, susupti, turya and turyatita. The adept thus has a vision o f
th e srlcakra in its cosmic diversity and power. This is expressed by
slo ka 50: [This] universe w hich has com e forth as th e cosm ic
o utline born from her own will is consciousness, the [visible] form o f
th e self, uncreated bliss and b e a u ty , 18 Then the goddess herself,
s u p re m e consciousness, is to be visualized in the centre o f the
srlcakra em bracing her consort K am esvara, 19 both holding the goad
and the noose made up o f the energy o f will (icchasaktim aya), the
b o w and the arrow which are energy o f action ( kriyasaktim aya), so
th at the two, male and female, aspects o f the supreme deity are seen
as present in the diagram which they pervade and animate, as they do
the cosmos, by their united pow er and will. To quote sloka 55: Such
is the suprem e splendour, the srlcakra as her cosmic bo d y ( vapuh),
su rrounded by the dazzling waves o f her m ultitudinous p o w e r,20 a
vision which fills the adept with wonder and awe.
H av in g thus visualized the M cakra in its co sm ic aspect o v e r
flow ing with the glory o f the goddess, the adept is now to visualize
and understand the role o f another group o f pow ers or deities resi
ding in this diagram, the M udras, w hich are ten in nu m b er (slokas
17 A S aiv a lih g a is not n e c e s s a rily o f a m o re or less p h allic shape. H ere, fo r
in s ta n c e , the itaralinga is said to be ro u n d like a kadam ba flow er. T h e b e st lihga is
often said to be a turn, an incised skull. As for the four lihgas listed here, they are the
usual fo u r types o f sivalihgas.
ls
svccchaviSvam ayollckhakhacitam visvarupakam /

caitanyam atmano rupam nisarganandasundaram //
19 T rip u r a s u n d a r i w ith B h a ira v a as h e r c o n so rt is also c o n c e iv e d as K amesvari

w ith K a m e s v a ra , these latter bein g in fact the b a sic d eities o f th e D a k s i n a m n a y a
( P a d o u x 1994: 38). On K am eS v ari, see, for in s ta n c e , B ii h n e m a n n 2 0 0 0 - 2 0 0 1 ,
v o lu m e I: 131.
20
cvam rupam param tejah sncakravapusii sthitam /

tadlyasaktinikarasphuradurmisamavrtam //
246
A NDR E P ADOUX
56-71). Though this is another phase in the vision and practice o f the
sricakra by the adept, it is not described by the YH as something he
is to do, but as a development taking place in the supreme conscious
ness, in the goddess, who now takes on the aspects o f these ten
goddesses, the M udras, who incarnate and express ten different
phases o f her cosm ic activity and power. To quote slokas 5657:
W hen [the goddess] becom es lum inously aware o f the universe
[appearing] on the screen o f her own consciousness ( cidatm abhittau
prakasam arsane), being fully possessed by the will to act, she acts by
her own free will. [Such is] the pow er o f activity which, because it
gladdens the universe and causes it to flow, is called m udra. The
last sentence o f this stanza explains the name mudra: these deities are
so called because they gladden (m odanat [V m ud]) and cause to flow
(rodanat [V ru]), hence mud-ra.
The first M udra is said to pervade ( vyapaka) the whole sricakra.
The nine others abide each in one o f the constituting cakras o f the
diagram, going from the outer square to the centre: the m ovem ent is
centripetal because these deities, born from the play o f the powers o f
the goddess, incarnate or symbolize nine stages o f the return o f the
cosm os to its unm anifest source. The adept, therefore, visualizing
them and identifying som ehow with each o f them, realizes that this
is how the play o f the g o d h ead s energy o f activity (k r iy a ), whose
nature is pure consciousness, is identical with the sricakra (k iiy a
caitanyarupatvad evam cakram ayam sthitam sloka 71). But w hat
the adept must first and foremost always intensely meditate (sarvada
bhavayet) is the supreme luminous pow er (param tejas) w hich is the
w illpow er (iccha) o f the godhead his attention is to be focused on
the suprem e, on the source o f all that exists. The YH m erely d e s
cribes these M udras as luminous deities to be visualized as they are
described, and quotes their cosmic functions. W hat is prescribed here
is therefore only one more perception and realization o f the presence
and play (here tending tow ard the resorption o f the cosm os [samhara]) o f the goddess in and as her cakra: the practice by the adept
remains a purely mental, meditative one. It is, however, w orth noting
that in his com m entary on these stanzas A m rtananda describes the
mudras as hand gestures the adept is to display so as to identify with
the role o f each o f the M udras, so that these mudras, in practice, are
at the sam e time deities to be visualized and w orshipped and handgestures an act o f mental and bodily participation and identification
t h e
r i c a k r a
247
o f th e adept with the deities: this aspect o f the practice o f the sricakra
s h o u l d not be overlooked.
T h e chapter ends by prescribing two other ways o f perceiving the
srica kra in meditation, first by considering it as divided into three
p o r tio n s each comprised o f three cakras, going from the centre to the
o u t e r part, and then as consisting o f its nine cakras, going from the
o u t e r part to the centre. The adept thus follows the process first o f
e m a n a tio n then o f resorption.
T h e chapter concludes (slokas 85-86): This is w here the great
g o d d e s s M ahatripurasundari is to be w orshipped. [Such is] in its
a b s o l u t e fullness the great cakra, g iver o f eternal y o u th and
im m o rta lity . Thus has been said, o suprem e goddess, the practice
( s a m k e ta ) o f the great cakra o f the goddess Tripura, b esto w er o f
lib e ra tio n while still in life.
T o conclude, we may note that this chapter, w hich at first sight
l o o k s rather disorderly, is in fact rationally constructed, em anation
a n d resorption succeeding each other and every successive m om ent
i n c r e a s in g the aw aren ess o f and identification w ith the co sm ic
d y n a m is m o f the sricakra. First is expounded the succession o f the
d iff e r e n t parts o f the sricakra (8- 21), then the play o f energies that
m a n if e s t the cosmos (22-24): this is an outw ard tending m o vem ent
m a n ife s tin g the world. A fter which, the bhavana o f the kalas o f hrim
( 2 5 36) and eventually the M udras (56ff.) turn the attention o f the
a d e p t tow ards resorption, towards the centre o f the diagram that is,
t o w a r d s liberation. T h e description o f all the deities and entities
p re s e n t in the sricakra also manifests its p o w er o f creation and o f
re so rp tio n . The adept, by visualizing and m editatively identifying
w ith this cosmic play, progresses towards liberation. As an active
c o s m ic symbol (or as a display o f the goddesss creative and salvific
a c tio n and power), the sricakra appears very efficacious.
W e m a y finally note that the origin o f the sricakra rem ain s
sh ro u d e d in mystery. This ritual diagram is to be found in the older
tex ts o f the Traipuradarsana (which is the less ancient o f the K aula
traditions), but where does it com e from and when did it appear? W e
d o n o t know. The source is probably to be looked for in the older
K u b jik a tradition, though this is not certain.21 A South Indian origin,
21
M a r k D y c z k o w s k i ( p e rs o n a l c o m m u n i c a t i o n ) b e lie v e s th a t s o m e a s p e c ts at
le a st o f this c ak ra c o m e fro m the S rita n tra sa d b h a v a , a long (a b o u t 5 0 0 0 stan zas) and
y e t u n e d ite d text w h ich he is currently studying.
2 48
ANDRE PADOUX
too, has been suggested by some, but this theory has so far not been
conclusively proved. W hatever its origin, the srlcakra, w hich we
have ju s t seen as it is shown in a chapter o f an ancient text, is still
very much in use (in ritual and meditation) in India today: not only in
the South Indian de-tantricized Srlvidya but also in the centre and
the north o f the sub-continent.22 It therefore deserves some attention .
22
T h e m o s t c o m p lete study o f the srlcakrapujfns that o f M ad h u K h a n n a (1986) in
an u n fo rtu n a te ly as y e t u n p u b lish ed O x fo rd thesis. Available, p recisc and co m p le te
is S a n j u k t a G u p t a s d e s c r ip tio n o f th e c u lt in G u p t a / H o c n s / G o u d r i a a n 1979:
139-157.
249
T H E S R iC A K R A
L T - ..:....::::::
111 .... :...i

1, T h e srlcakra
250
ANDRE PADOUX
mahabindu
unm and
sahasrara
<o>
ajila (bindu) bhrumadhya
la m b ik a
17isitiidha
(uvula)
k a n th a
(throat)
anahata hrdaya (heart)
manipuraka nabhi (navel)
svadhisthana
sakta
m u la d h a ra
kulapadma
visa
ka n d a
a k id a p a d m a
2 . T h e bodily cakras according to A m r ta n a n d a s c o m m e n ta ry on the Y o g im h rd a y a
VASTUPURUSAMANDALAS; PLANNING IN THE IMAGE OF

MAN
Michael W. Meisler
D ia g ra m s for planning and meditation permeate South Asia, as tools
f o r praxis, practical and religious. Psychedelic or pragm atic, they
r e m a in utilitarian at their core. They do not constitute a single reality,
b u t h a v e a history that makes o f each a palim psest. By the sixth
c e n tu ry A.D., these layers had been com bined to provide a tool, both
religious and practical, for the constitution o f a shelter for deities and
w o rsh ip p e rs the tem ple as a new form o f Hindu w orship began.
T h is paper explores evidence found by recent scholarship in built
m o n u m e n ts for the application o f such diagrams to the construction,
validation, proportioning, and designing o f such shelters for the first
e arly centuries o f their use.
Indus Valley cities, with their gridded street plans dating from the
th ird to second millennium B.C., have been cited as early exam ples
o f the city as pivot o f the four q u a rte rs (W h eatley 1971) even
th o u g h their rhomboidal layouts and orientation only approxim ate a
cardinal grid. They should perhaps remain in the pre-history o f South
A sia n urban planning (Kenoyer 1998).
T he mystic cosm ogony o f the A tharva-V eda, on the other hand,
f ro m early in the first m illennium B.C., does pro v id e us w ith a
p a r a d ig m for cosm ic planning in South Asia. In b o o k 15 o f the
A tharva-V eda, cosmic speculation and the body o f m an were m ade
into a form al hom ology, as well as being described as i f a th re e
d im en sio n al m a n d a la .1 T h e re a vratya ascetic b e lo n g in g to an
un o rth o d o x o rd er2 is described as confronting his ow n divinity as
1 W h itn e y 1905: 7 6 9 c o m m e n ts that the C u lik a - U p a n is a d r e c k o n s the vratya as

o n e a m o n g the m a n y form s in w h ich B ra h m a n is cele b ra ted in A V ., m e n tio n in g in
t h e s a m e v e rs e w ith vratya ( c e le b r a te d in A V . x v .) also th e b rahm acarin a n d the
s k a m b h a and the palita."
2 H e e s t c r m a n 1962: 36, on the o t h e r han d , c o n c l u d c d th at th e vrafyas w e r e
a u th e n tic V cdic A ryan ... p r e d e c e s s o r s ] o f the d ik sita.
252
M IC H A E L W.
M EIST ER
Ekavratya, the sole V ratya (Kramrisch 1981: 472, 486). In Stella

K ram rischs retelling, [t]he transfiguration o f the Vratya has three
phases: the birth o f the god, the vision o f that god, and the building
o f his m o n u m en t (Kram risch 1981: 89). She describes the Sole
V r a ty a as a choreographed m onum ent o f deity built up by the
words o f hym ns, having a mandala for a pattern (Kramrisch 1981:
95, 93): He moves out on his vehicle, the mind, first toward the
east, then toward the south, toward the west, and finally toward the
north.... [He] incorporates into his presence the four directions o f the
extended universe (Kramrisch 1981: 93).
K ram risch saw this vision o f the vratya ascetic-already in the
second millennium B.C. as a forecast o f her dictum that the Hindu
tem ple o f a much later period could be described as a m onum ent o f
m an ife statio n (Kram risch 1946: passim), characterizing w hat the
vratya saw as the lord o f the space-tim e universe, h im s e lf the
central pillar o f a four-sided p y ram id (K ram risch 1981: 96). Yet
such a v is io n did not then constitute architecture nor represent a
developed practice.
O f several versions o f the origin o f the uni verse in the Vedas, the
sim plest is that the creator built the universe with tim ber, as a
carpenter builds a house (Encyclopaedia Britannica on-line; B row n
1942, 1965). Indeed the rituals surrounding the m aking o f V edic
shelters provide a vocabulary for wood and reed construction (Renou
1998) and suggest the presence o f a cosm ography (B odew itz 1979)
by h a v in g central and cardinal orientations, but do not define a
generative to o l (Bafna 2000: 45) for architecture, as may mandalas
o f a later period.
In the building o f altars (Staal 1984), bricks were laid to form an
orthogonal frame, and altered to m ake a variety o f shapes to suit
different ritual purposes. Perhaps the grid o f later m andalas has one
source in the piled bricks o f such sacrificial surfaces. W e have few
texts to suggest this. W e do, however, have quite early texts, SulbaSutras (D atta 1932), o f ca. the third-fourth centuries B.C., that
provide the geom etric construction, using cords to draw circles,
needed to locate the square and cardinal orientation essential for the
plot o f a sacrificial altar (Apte 1926; Bag 1971) (Illustration 1).
These geom etric m anuals scrupulously avoid interpretation, yet we
know that their function was in part to assure that an altar w ould
V A ST U PU RU SA M A N D A LA S
2 53
c o n fo rm as a homologue to an oriented and therefore square universe

(Ivlenon 1932: 9495),
It is in fact only in the time o f V a ra h a m ih ira s B rhat-Sam hita,
w r itte n in the sixth century A.D., that the use o f som ething like a
vastupurusam andala to plan cities and buildings was first designated,
in his chapter 53 On A rch itectu re.3 A distinction m ust be m ade
b e tw e e n a diagram as a ritual tool or a con stru ctio n al d e v ic e
( M o s te lle r 1988) for architecture; V a ra h a m ih ira , h o w e v e r, had
c o m p ile d in this text m any earlier layers o f know ledge as well as
contem poraneous practice, as he had also assem bled rival systems o f
a s tro n o m y in his SOryasiddhanta.4 As he intro d u ce d his p ro je ct
(53.1), [t]o gratify clever astrologers, I now proceed to com pose a
w o r k on the art o f building, such as it has been transmitted from the
C re a to r to our days, through an unbroken series o f sages. He first
invokes a Vedic description o f the original act o f sacrifice (53.2-3):
T h e re was ... some Being obstructing ... both worlds ... [who] was
su b d u ed by the host o f gods and hurled do wn. O f the several parts o f
his body, each is subjected to the particular deity by w hich it was
attacked. It is that Being o f immortal substance, w ho ... was destined
to b e the dw elling-house personified [the vastupurusa].5 H e then
g o es on for a num ber o f verses (5 3 .4 -4 1 ) to describe a variety o f
h o u se structures, their class linkage, orientation, storeys, balconies,
etc., as i f this architecture were an elaboration o f the v ern acu lar
shelters o f Vedic India (Renou 1998).
O n ly at 53.42 does he introduce the d ivision o f the plan into
squares to fit the rite o f sacrifice and the placem ent o f deities from
the older m yth (Illustration 2): In order to divide (the ground-plan
o f a house) into eighty-one squares, draw ten lines from east to west,
and ten others from north to south.6 He discusses the placem ent o f
45 deities over the body o f the vastupurusa for 13 verses (5 3 .4 2 -
3 1 refer to K e r n s translation throughout.

4 K r a m r i s c h 1946: 79 c o m m e n t s th at th e s y m b o l i s m o f t h e V a s t u p u r u s a
m a n d a l a w as a residue o f trad itio n s still k n o w n and p ra c tic e d th o u g h n o lo n g e r
re a liz e d in all their im port.
5 K ra m r is c h 1946: 73, 78 m a k e s the im p o rta n t d istinction th at V a s tu p u r u s a as
s u p p o r t o f the build in g ... is d e sc rib e d as lying with his fa c e d o w n ... w h e re a s A gni
P ra ja p a ti o f the V ed ic altar lies f a c in g u p w a rd s His h e a d sh o u ld lie to the n o r th
east,
'' H e does not describ e the location o f the sq u are re q u ire d b y the g e o m e try o f the
Sulba-S utras.
254
M IC H A E L W.
MEISTER
54),7 then introduces an alternative practical m andala for c o n

struction: One may also, should one prefer it, divide the area into
sixty-four compartments (53.55).
F or the rem ain d er o f this chapter (5 3 .5 7 -1 2 5 ) V arah am ih ira
discusses vulnerable crossings (m arm an), displacem ent o f pillars,
and a range o f magical associations and consequences still part o f
traditional wood architectural practice today (Libersat 1988), w ith
the significant dictum that [t]hc householder ... should carefully
preserve Brahm an, who is stationed in the centre o f the dw elling,
from injury ... (53.66).8
I take this time to lay out V araham ihiras order o f presentation
b ecause it is he who first puts together vastu (building), purusa
( m a n , but as a trace of sacrifice), and mandala (diagram), sum m ing
up m any centuries o f speculation on the rituals o f building (53.98):
At a period indicated by the astrologer, let the householder go to a
piece o f ground which has been ploughed, abounds with seed grow n
up, has served as a resting-place of cows, or has got the approval o f
the B rahm ans. He gives us a sense o f the range o f caste patrons
(53.100): Then touching his head, if he be a Brahman; the breast,
if a Kshatriya; the thigh, if a Vaigya; the foot, if a Sudra let him
draw a line, the first act when a house is to be built ... and o f the
im portant role o f a proficient holder o f the m easuring line, the
architect or Sutradhara (53.110): By the measuring line snapping
asunder may be predicted death; by the plug drooping its top, great
sickness; by the h o use-ow ner and architect falling short in th e ir
m em ory, death. Kern, V arah am ih iras translator, accused him o f
having the habit o f un-critically copying his authorities (1872: 292,
note 1), and yet he is not m erely sum m ing up a m illen n iu m o f
building, but m arking a m ajor transition. A new practice o f stone
construction to make temples to shelter images o f deities was ju st
beginning (M eister 1986) and the utility o f the vastupurusamandala
was about to be given a new life (Meister 1979).
7 A l t e r n a t i v e a r r a n g e m e n t s o f d e i t i e s do e x is t in o t h e r an d l a te r te x ts
( A p tc /S u p e k a r 1983). K ram risch 1946: 1 9 -9 8 best synthesizes the m ultiple layers o f
sig n ific a n c e laid o v e r the m a n d a la by a variety o f so u rces: sacrificial, z o d ia c a l,
c h ro n o m e tric , astro n o m ic , m y th o lo g ic a l, etc., a p a lim p s e s t or m o s a ic , at b e st, not
ever a w h o le fabric.
*
P a n c a r a tr a d ia g r a m s u sed for w o rsh ip sig n ific a n tly e x c h a n g e V iSvarupa for
B ra h m a n (A pte 1987: 143).
V A STU PU RU SAM A N D A LAS
255
B efore w riting o f tem ples explicitly, h o w ev er, V a ra h a m ih ira

provided two chapters (54-55) on the exploration o f w ater-springs
and culture o f tre e s , things essential to the e sta b lish m en t o f a
sacred landscape. Then, in a chapter (56: 1-31) with only 31 verses,
he provided a brief description o f various tem ples, which he begins
(56.1): Having made great water reservoirs and laid out gardens, let
one build a temple, to heighten o n e s reputation and m erit. The
gods used to haunt those spots w hich by nature or artifice are
furnished with water and pleasure-gardens (56.3), he writes, then
describes these in loving detail (56.4-8). He com m ents that the soils
he had indicated w hen treating o f house-building ... are likew ise
recom m ended to persons o f the different classes, w hen they w ish to
erect tem ples (56.9), and then, in a significant verse, he specifies
(56.10): Let the area o f a temple always be divided into sixty-four
squares, while it is highly com m endable to place the m iddle door in
one o f the four cardinal points. 9 It is this dictum that seem s to
define a new m illennium o f vastupurusam andalas, to be u sed in
practice (Illustration 3) as well as ritual.
M ost rem arkable to me, at the time o f V arah am ih iras writing in
the sixth century, is how few stone temples-and how experimental
th e ir a r c h ite c tu re had by then b e e n b u ilt ( M e is te r 1981a).
V araham ihira was on the forward cusp o f a new, even m o d e rn ,
architecture meant to shelter newly manifest images (Meister 1990).
T he rem ainder o f V arah am ih iras chapter is a listing o f twenty
kinds o f shrines (56.17), w ith varied plans, storeys, turrets, and
dorm er w indow s that probably existed not in stone but in w o o d .10 As
he casually concluded (56.31): H erew ith are the characteristics o f
tem ples described in com pendious form.... O f the volum inous works
by M anu, etc., have I, in writing this chapter, only taken notice in as
m uch as I rem em bered.
T he distinction betw een a proportional system , w hich the ritual
vastupurusamandala is for the universe, and a constructional device,
such as V araham ihira stipulates for the temple, was drawn some time
ago (Panofsky 1955).n The notion that m agical diagram s called
9 B afn a 2000: 38 is not correct in stating that V a r a h a m ih ir a s c h a p te r on te m p le s
does not e v e n m en tio n the d ia g ra m s.
10 At least one e n g in e e r ( P r a m a r 1985) h as a tte m p te d to a n a ly z e an d ap p ly a
m a n d a la designed for w o o d e n structures to the b u ild in g o f stone te m ples.
11 P a n o fsk y d istin g u ish es b e tw e e n a th eo ry o f p ro p o rtio n s and a p ractical system
o f construction.
256
M IC H A E L W.
M E IST E R
m a n d a la s, according to one recent analysis (Bafna 2000: 26),

underlie most traditional Hindu [I would prefer Indie] architectural
production has becom e well entrenched within current thought....
[F]ew scholars have attem pted to describe the precise m anner in
which the mandala could have acted as a generative diagram . Stella
Kramrisch, whose T h e Hindu T e m p le (1946) had collected a w ide
body o f references to the Sastric texts on the building o f temples that
f o llo w e d V a r a h a m i h i r a s, had in fact c o n c lu d e d th a t th e
V astum andala is the m etaphysical plan o f the temple primarily; its
c o s m o lo g ic a l and m agical im p licatio n s are d e riv e d from it
(Kramrisch 1946: 37, note 40).
The attempt at the literal mapping o f the vastupurusamandala onto
buildings by m odern architects, both Western (Volwahsen 1969: 44)
and Indian (Kagal 1986; Correa 1996), or the recent resurgence o f
Shilpa S h astrin s orienting houses as an Indian astrological e q u i
valent o f Chinese fe n g shui (for example, Rao 1995), however, are
no test to the use o f vastum andalas as a constructional tool in the past
(M eister 1997). Bafna 2000: 31 has put it another way: [W ]hat
co n n ectio n could be posited w ithin diagram s associated w ith a
m arg in a l religious cult [he is referring to T antrism ] and those
associated with a practical profession [architecture]? Yet it is preci
sely about the division o f the plan that Varahamihira is m ost explicit
and practical (53.42): [D]raw ten lines from east to west, and ten
others from north to south. He is w riting about a constructional
device related to a proportioning system.
K ram risch h e rs e lf was unsure how such a device m ight have
w o rk e d (K ram risch 1946: 58): T he V a s tu m a n d a la is a p r o g
nostication, a forecast and tonic o f the contents that will be built up
in the tem ple; it is in a literal sense, its program m e. This does not
imply an identity o f the actual plan o f the temple, with the m andala.
W hile she found some relationship between the simplified 16-square
grid in the M atsya-Purana (Kramrisch 1946: 228), here, it seems to
have been suggested by the simplicity o f the shrine; its plain, thick
w alls, w ith o u t buttresses, b elong to small structural tem ples in
central India o f the Gupta A ge ..., o f later tem ples she had little
doubt: W h en the great temples w ere built, after the ninth century
and w hich still stand, the drawing o f the V astupurusam andala had
b ecom e an architectural rite without necessarily coinciding with the
laying out o f the ground plan o f the Prasada. Her conclusion was
257
n o t so different from one scholars recent attempt to com pare square

T a n tr ic painted m andalas to the elaborated ground plan o f a tem ple
in Orissa, [i]f we were to accept that the mandala was typically used
a s a design tool, then actual built examples m ust show evidence o f
p la n n in g based on the m andala (Bafna 2000: 38).
It is, o f course, this sort o f evidence, collected in the field, that has
g r a d u a lly been accum ulating over the past h a lf century (M eiste r
1979, 1985; Pichard 1995; Thakur 1996).12 Buildings have perhaps
p ro v e d more reliable than texts as historical docum ents recording the
m eth o d o lo g ies that built them (M eister 1989). This is in part because
o f the multiple uses over time to which the vastupurusam andala was
p u t ritual, astrological, m editational, devotional, but also c o n
structional.13
Referring to the Pauskara-Samhita, Apte 1987: 129 com m ents that
M andala worship, in those days was not a part o f the tem ple ritual
o n ly , but often m aintained its independent existence like the sacri
ficial institution (Yajna). And for that ritual ... a special pandal used
to be erected on a chosen site may be on a m ountain or in a forest
o r by the side o f a herm itage or on the bank o f a river or inside
te m p le p rem ises (Pauskara-S am hita 2 ,4 -5 ). This is not so m uch
different from the shelter set up for the ritual painting o f dhulicitra in
K e ra la a form o f bhaum ika citra or earth p a in tin g (Jones 1981:
71): [T]he ceremonial draw ing in pow ders m ay also be perform ed
a t night in an appropriate space within the precincts o f a Mambutiri
B rah m an a household or in a palace o f a Sam anta or K shatriya ruling
fam ily. T he designated area ... is traditionally covered b y a canopy
co n s tru c te d o f four slim areca logs, w rapped in n e w u n b le a c h e d
clo th , supported by pillars o f the sam e wood, sim ilarly w rapped,
w h ic h form the boundaries o f the sacred draw ing. The image o f the
g o d d e s s or an o th er deity is b uilt up in co lo u red p o w d ers, th en
destroyed through ecstatic dance.
A sm all but significant side current o f stone tem ple architecture
th a t m ay reflect such te m p o ra ry p a n d a ls are the th in -w a lle d
niandapika shrines set up as funereal m em o rials in C entral India
12 S e e also, h ow ever, m y cautionary review (M e is te r 1999).

13 A recen t r e v ie w (B a fn a 200 0 : 4 7 ) is co rre c t to c o n c lu d e th at w e m u s t lo o k
u p o n the V a s tu p u r u s a m a n d a la as an idea th a t h as b e e n c o n s ta n tly re d e f in e d and
e x p l o i t e d th ro u g h h i s to r y b u t his c a v e a t th at w h a t w e h a v e m a d e o f it n o w is
m e r e ly a recently co n stru cted u n d e rsta n d in g m ay best be applied to him self.
258
M IC H A E L W.
M E IST E R
from the sixth to tenth centuries A.D. (M eister 1978).14 These stand
in sharp contrast to the thick-walled temples ... o f the G upta age
(K ram risch 1946: 228), their inner sancta only twice in width the
thickness o f their walls. Those, at the very beginning o f the stone
tradition, measured their inner sacred space by the dem ands o f the
m andala, not their walls by the efficiencies o f stone (Illustration 3
A).
Actual physical yantras or metal mandala plaques are buried in the
foundation o f structures as tools to sanctify the building. As the label
to tw o such metal plaques recently on display in the A m erican
M u se u m o f N atural H istory, N ew Y ork, put it (H u y ler 1999),
Yantra are specific m andala created to attract Divine Energy o f a
deity into a sacred space. M ade o f metal or stone, yantra are buried
beneath a tem ples inner sanctum during construction. Such yantras
are also draw n at significant points o f the plan in late O rissan
p ra c tic e ( B o n e r 1975), but should not be co n fu sed w ith the
vastupurusamandala as a w hole.15
To look for the application o f the vastupurusamandala introduced
by V araham ihira as a constructional tool or planning device in the
sixth century, it is necessary to m easure standing tem ples. Bafna
2000: 4 1 - 4 2 is m ista k e n in c la im in g a tro u b lin g lack o f
corroborating evidence from surviving built structures. T here have
been both detailed studies o f individual p la n s as well as w hat he
calls c o m p arativ e m orp h o lo g ical s tu d ie s that go w ell beyond
im p o sin g a co n strain in g o rth o g o n a lity (M eiste r 1982, 1983a;
T hakur 1990).
F ro m m y ow n ex p erien ce, how ever, certain p ro c e d u re s are
im p o rtan t (M eiste r 1979, 1999). The square G upta tem p les o f
K r a m r is c h s reference Sanchi 17, T igaw a, N ach n a, even Deogarh use thick m asonry walls to surround an inner sanctum ca. h a lf
the width o f the whole (M eister/D haky/D eva 1988). Early seventhcentury shrines in Orissa (Illustration 3 A) with developed latina
nagara superstructures and a cross-plan with closed doorw ays on
14 A n inscription o n the s e v e n th -c e n tu ry m a n d a p ik S shrine at M a h u a refers to a

stone m a n d a p ik a set up in h o n o u r o f a local ru lers d eceased parents.
15 T h e sq u are grid w ould then sim ply be an in stru m en t a y a n tra u sed fo r the
d e p ic tio n o f the vastu d e itie s (B a fn a 200 0 : 45). B a f n a s re fe re n c e to y a n tr a s
e m b e d d e d w ithin the traditional d raw in g s o f tem p le p la n s, h o w ev er, begs the issue
that no such d raw ings survive before the m odern period.
259
th re e w a lls,16 precisely fit a constructing grid o f 64 squares w hen

m easu red at the root o f their wall m ouldings (the khura h o o f o f the
vedibandha) that is, w here the m andala co u ld be draw n on the
sto n e foundation that formed the floor level o f the sanctum. T hese
shrines well fit V arah am ih iras dictum to let the area o f a tem ple
a lw ays be divided into sixty-four squares and to place the m iddle
d oor in one o f the four cardinal points (56.10),
In m y own fieldw ork (M eister 1979) I first b egan to m easure
te m p le plans thinking they would only confirm K ram risch s intuition
th a t the vastum andalas specified by V araham ihira and by later texts
h a d b e c o m e an a rch itectu ral rite , as te m p le s b e c a m e m ore
e lab o rate after the ninth century, w ithout necessarily coinciding
w ith the laying out o f the ground p lan (K ram risch 1946: 228).
H o w ev er, directly m easuring the m ouldings o f a group o f seventha n d eighth-century temples in M adhya Pradesh in particular the
s eventh-century nagara tem ple at M ah u a first gave me evidence
th at architects o f this region and in this period w ere using a new
p ro c e d u re (Illustration 3 B). Standing above two levels o f a stone
foundation, the sanctum walls o f this Siva tem ple m easured ca. 556
cm from corner to corner (ca. 114 cm for karna and bhadra piers; 57
cm for intermediate pratirathas) at the khura h o o f o f th e vedibandha
m ouldings, which was the floor level o f the sanctum. The inner space
o f the sanctum measured ca. 228 cm in width.
T h e se m e a s u re m e n ts em b o d y a n ew p a ra d ig m for b oth the
c o n c e p t and c o n s tru c tio n o f the m u ltip ly in g w all offsets that
distinguish nagara tem ples in this period (Illustration 3 C). Central
bhadras on the outer walls project the m easure o f the brahmasthana,
flanking pratiratha offsets mark the dimensions o f the inner sanctum.
Such a system 1 found also rigorously applied to other tem ples in the
G w alior region (M eister 1979). Such a use o f the m andala was new,
practical, and expanded the m a n d a la s m eaning to the temple, as a
physical expression o f its plan.17
As a test o f the constraining orthogonality, as B afna 2000: 41
p u t it, I also m easured and analyzed rectan g u lar tem ples in this
region and century, as w ell as those beginning to ex perim ent w ith
lf K ra m r is c h 1946: 271 interprets sadasra to be this type: [T ]h e g r o u n d p lan ...
has six faces, for each o f its th ree sides has a central b u ttress w h ic h is set o f f from
the w all....
17 S in h a 2 0 0 0 has e x te n d e d this an aly sis o f the bhadra c lu s te r to vesara te m p le s
in the D eccan.
260
M IC H A E L W. M EIST ER
octagonal and turned-square plans (M eister 1982, 1983a, 1983b,

1984), In both cases, the m andala continued to control the w idth o f
walls, location o f corners, and to project sanctum and brahm asthana
dimensions through the walls as measured offsets (Illustration 6).
N o t all tem ples across all o f India yield sim ilar results, but
increasingly regional understandings and m isunderstandings o f this
system o f planning becom e clearer, as m ore tem ples have been
adequately m easured and analyzed (M eister 1985; Pichard 1995;
Thakur 1996), In South India, for example, use o f an odd-numbered
grid, centring the sanctum on a square, made rings o f expansion
possible (Illustration 4), In the north, separate sacred spaces might
overlap (Illustration 5).
Bafna 2000: 41 confuses the role o f measure and proportion in his
conclusion that such variation makes o f the vastumandala not so
m uch a constructional aid as a tool for the designer, one that was
used to control the proportions o f the design rather than its m easure.
M easure in the Indian context was relative, determined by the height
or hand o f the donor, architect, or image; p ro p o rtio n was the
constructional a id (Meister 1985).
H e also, it seem s to me, is w rong to c o n c lu d e th at the
Vastupurusamandala c a n n o t be rotated w ith o u t lo sin g all its
sig n ifican ce (B afna 2000: 41). R otating plans su p erim p o se one
turned square on the other, moving toward the circle (Illustration 7)
that constitutes a m an d alas pre-existent form (and which surrounds
the s q u are gated p alaces in the p a in te d m an d alas o f B a f n a s
article).18
K ram risch 1946: 62, while pointing out that earlier texts ... do
not record circular V astus reported that Utpala, the tenth century
co m m en tato r o f the Brhat S a m h ita describes in detail the c o n
struction o f circular sites.... This appears a developm ent around the
principal Vastu, which is and must remain square.... She also cited
A gni-Purana 93.40: In. the m iddle o f the six sided, three sided, and
circular plan, should be the square. Referring to Vastuvidya 7.6 and
10.15, she also observed (Kram risch 1946: 62, note 105): as the
months advance the Vastupurusa moves ... The spatial order o f the 8
directions simultaneously denotes a temporal order; the Vastu is the
tim e piece,... T his ro ta tin g V astu is called C a ra v a s tu and is
K ra m risc h 1946: 41 rem arked that the square s y m b o l o f the ex te n d e d w o rld in
its o rder has pre c e d e n cc over the circle o f time, the second o rn a m e n t.
V AST U PU RU SAM A N D A LAS
261
d is tin g u is h e d from the Sthira-vastu, w h o se p o sitio n is fixed....

T e m p le s are meant to last and are always built [i.e., founded] on the
Sthiravastu.
Experim ents with em bedded octagons from the octagonal stone
te m p le at M undesvari and brick tem ples in D aksina K osala o f the
s e v e n th cen tu ry (M e iste r 1981b, 1984) to the g re a t m israka
su p e rs tru c tu r e o f the C h o la te m p le at G a h g a ik o n d a c o la p u ra m
(Pichard 1995) might suggest that architects at an early period also
th o u g h t in such terms o f their temples (and their construction).
M y own w ork for a period o f tim e focused on m easu rin g and
analyzing m onum ents that could provide test cases for the limits o f
m an d ala planning. I had thought that the application o f the grid o f
th e m andala and its significance to the proportioning o f tem ples in
th e seventh and eighth centuries could not explain the variations
found in the ninth and tenth. W hat 1 discovered, however, in Central
In d ia w as a shift in construction o f the tem ple that p reserv ed the
relatio n o f bhadra and pratirathas to san ctu m and brahmasthana
w hile pulling the bulk o f the tem ple w ithin the grid (Illustration 3
D ) ,19 This bhadravyasa m easure allowed the fabric o f the tem ple to
be reduced, proportions in the wall to be m ore balanced, and new
plans to em erge (M eister 1979, 1985). That the grid o f the m andala
could continue to have a practical utility, even in com plex and huge
te m p le s o f th e 11th c e n tu ry , as at K h a ju r a h o , w as s ta rtlin g
(Illustration 5).20 N o longer fixed at the foundation as in earlier
shrines (Illustration 3), yet still governing the walls enclosing the
inner sanctum, the continuing presence o f these proportions in the
fabric o f these stone m onum ents is perhaps our strongest surviving
e v id e n c e for the notion o f a geom etrical device w ith sym bolic
d im ensions u n derlying all architectural p ro d u c tio n (B afna 2000:
42).21
I
think Bafna 2000: 43 is right that [p]ractically speaking, a grid
is a cum bersom e and com plicated tool for the laying out o f plans; it
19 B afn a 2000: 41 m ista k e n ly attributes this ch a n g e to the sev en th c en tu ry in stead

o f the ninth.
20 M e i s t e r s a r g u m e n t is s o p h i s t i c a t e d a n d p u r s u a s i v e : th e g rid is o n ly a
r e g u la tiv e tool an d the very act o f e m b e d d in g it is auspicious.... B u t in f o rm u la tin g
th is idea, he se e m s to h a v e m o v e d a g o o d deal a w a y fro m the strict orien tatio n and
h ie ra rc h y o f the V astupurusam andala" (B afna 2000: 41).
21 B a f n a , h o w e v e r , r e s o lv e s that th e idea o f th e g o v e r n i n g m a n d a l a ... is
m e re ly a recently constructed u n d e rs ta n d in g (B afn a 2000: 42, 47).
262
M IC H A E L W. M E IST E R
is extremely susceptible o f errors unless checked by diagonals, yet

find his altern ativ e the centerline system ... still used ... to
com pute the proportion o f statues an odd choice (see M osteller
1991). From the time o f the Sulba-Sutras, the geometry and tools for
laying out a plan were known. If the Sutradhara was controller o f the
cord let him draw a line, the first act ... as Varahamihira 53.100
had put it he also controlled the geom etry that the use o f the
com pass m ade possible (Illustrations 1, 7). Such geom etric c o n
struction gives precision; a grid establishes proportion; reference to
the vastupurusamandala maintains ritual authority.
B afn a 2000: 39, 41, how ever, calls such c o n stru c tio n an
alternative peg -an d -strin g geom etry and com m ents that [t]here
are no indications within the literature on the history o f Indian
m athematics, or within vastusastra texts, that there were two separate
techniques o f geometrical constructions prevalent at any tim e. Yet
the constructive geom etry prevalent from the tim e o f the SulbaSutras m ust itse lf be seen as the source both o f the m a n d a la s
con strain in g o rth o g o n a lity , in B a fn a s terms, and o f a certain
freedom from it. B afna 2000: 41 adm its, som e silpa manuals
specifically record peg-and-string operations to ensure a precisely
oriented construction o f the square perim eter o f the Vastupurusa
mandala itself.
M y analysis o f temples with turned-square plans (M eister 1982,
1983b, 1984, 1989) beginning w ith the rem arkable m id-eighthce n tu ry G argaj M ah ad ev a tem ple at Indor in M adhya P radesh
(Illu stratio n 6 B), can d em o n strate both the c o n tin u in g o rth o
gonality o f tem ple planning and its freedom from constraint. Bafna
2 0 0 0 : 41 refers to stellate plans with a nonorthogonal profile, yet
the angled buttresses o f these tem ples must be observed as right
a n g le d co rn ers o f tu rn e d sq u a re s22 (that is, as o rth o g o n a lity
unconstrained).
T h e g ro u n d plan o f the G argaj M ah ad ev a tem ple at Indor
combines what Kramrisch 1946: 62, using Vastuvidya, has identified
as sthiravastu and caravastu, marking the tem ples functions as both
cosm ogram and chronogram . On the walls o f the tem ple at Indor,
Siva and his family mark fixed cardinal directions; eight dikpalas,
22
P re v io u s sch o lars have often not o b se rv e d this. W illis 1997: 60, fo r e x am p le,
d e s c rib e s In d o r as h a v in g a stellate s h a p e w ith sq u are and a c u tc p r o j e c t i o n s
while his plans sh o w obtuse corners rather than right-angled ones.
263
g u a rd ia n s o f the quarters, stand on the intermediate rotating bhadras,

fa c in g sub-cardinal points (Illustration 6 B).23
T h a t architects from the seventh century in D aksina K osala to
1 1 th- to 13th-century K arnataka, M aharashtra, and 16th-century
R a j a s t h a n took the g re a t tro u b le to b u ild s u c h c o m p le x ly
constructed turned-square monuments in brick and stone (Illustration
7) m ust be the best evidence for some special symbolism associated
w ith the com position.2,1 It cannot be in doubt that the constructional
m e c h a n is m m ak in g possible such co m p o sitio n s w as the sim ple
g e o m e t r y o f the S u lb a -S u tr a s -not the grid itself, w h ic h is
c o n se q u e n t that had located sacred ground for so m any centuries.
P erh ap s that is w hat the vastupurusam andala hid. B a f n a s com m ents
t h a t bu ild in g s can both pro v id e a structure for an em b ed d ed
m an d ala, and also serve to hide it (Bafna 2001: 46) is valid, but I
w o u ld reverse his conclusion.25 It is the building that acts in place o f
th e grid, beco m in g the mandala. As K ram risch, citing the M ahab h a ra ta , had pointed out for the palaces o f the th ree w orlds the
te m p le mimics, [t]hey revolved, each on its own level; they were
p art o f a revolving universe (Kramrisch 1981: 414).
23 T h is g e o m e try , as w ith all p ra sa d a s, is interrupted by t h e p ra g g rlv a e n tra n c e to

th e interior sanctum .
24 B a fn a 200 0 ; 4 4 , h o w ev er, calls this p ro b le m a tic .
25 T h e grid acts in p la c e o f t h e b u ild in g , r a th e r th an se rv in g a s the b a sis o f i f '
(B a fn a 2 000: 44).
264
M IC H A E L W . M E IST E R
VASTUPURUSAM ANDALAS
265
2. V a s tu p u ru s a m a n d a la o f 81 squares, as d e sc rib e d in the B rh a t-S a m h ita
266
M IC H A E L W . M EIST E R
3. G ro u n d p lan s and constructing m andalas: A. B haratesvara temple, B h u b an esh w ar, O rissa; B. Siva
te m p le no. 2, M a h u a , M a d h y a Pradesh; C. M a h a d e v a tem ple, A m rol, M a d h y a Pradesh; D. N a k tim a ta
tem ple, Bhavanipur, Rajasthan
VASTUPURUSAM ANDALAS
4. BrhadTsvara tem ple, G a n g a ik 5 n d a c o ja p u ra m , T a m iln a d u
267
MICHAEL
W. MEISTER
5. Khajuraho, Madhya Pradesh: A. Visvanatha temple; B. Laksmana temple, socle and vedibandha mouldings;
C. Kandariya Mahadeva temple
269
6. A. Sabari tem ple, Kharod, M a d h y a P radesh; B. Gargaj, M a h a d e v a tem ple, Indor,

M a d h y a Prad esh
270
M IC H A E L W . M E IS T E R
7. C o m p a riso n o f the use o f constructing g e o m e try and o d d -n u m b e re d grids in plans b a s e d on

three and six turned squares
COLOUR PLATES
I . A yantra o f G uhyakali
2. T h e haliharcm a o f the vaisvadeva rite
3. A dom estic siva p a n ca ya ta n a
4. A rudrapithamahayantra
5. A sa rv a to b h a d ra
6. A ram aU ngatobhadra w ith 26 rd m am udras

and 28 lin g a s
1. A ca tu rm u d ra ra m a lin g a to b h a d ra with 4 ra m a m u d ra s and 8 I in g a s an d a

sa rva to b h a d ra in the centre
8. A g a n e sa b h a d ra w ith five icons o f G anesa
9. A ganesabhadra with 21 ic o n s o f G a n e s a
10. A su ry a h h a d ra w ith 12 icons o f th e sun
The construction o f a d va d a sa lin g a to b h a d ra w ith a sa rv a to b h a d ra in the ccntr
12. T h e invocation o f dcitic.s into arcca nuts placcd on a sa rva to b h a d ra
A vessel o f plenty placed on a sa rv a to b h a d ra
=wi*3iproq
(*)
14. T h e cakrabjaniaiulala according lo die Pad m a-S am h ila
srspt*q*n*s*CTC.
tip ^ M 'funrftgrftrqprait
15, T h e navapadm aD iaiuiala according to the Ja y a k h y a -S a m h ita
16. T h e s n m a n d a la o( the N etralanlra follow ing K scm arajas co m m e n ta ry
17, An alternative structure o f the s n m a n d a la o f the N etratan tra
18. T h e m an d ala o f the nine lotuses (n a va n a b h a m a n d a la ) according lo the S v a c ch an d atan tra and
K s e m a r a ja s c o m m en tary
19. A tentative reconstruction o f the trident m a n d a la o f the SiddhayogesvarTm ata (long recen sio n )
a c c o rd in g to the T an tralo k a
B IB LIO G R A PH Y AND A B B R E V IA T IO N S 1
Texts and Translations

A g h o ra S iv a c a ry ap a d d h a ti. A ghoraS iviicaryapaddhati (= K r iy a k r a m a d y o tik a ) , w ith
C o m m e n ta r y ( Pmbhii) b y N ir m a la m a n i. E d ited by R a m aS astrin and A m b a la v a n a jn a n a s a m b a iid h a p a r a ^ a k tis v a m in . C i d a m b a r a m , 1927 (in g ra n th a c h a r a c
ters).
A jitagam a. Edition criliquc par N.R . Bhatt. 3 volum es. Po n d ich ery : Institut Frangais,
1964-1991.
A th a rv a -V e d a
A th a rv a -V e d a Sarhhita. T ran slated w ith a Critical an d E x e g e tic a l C o m m e n ta r y
b y W .D , W hitney, revised an d bro u g h t nearer to co m p le tio n and ed ite d b y C.R.
L an m an . 2 volumes. C am bridge, M assachusetts; H arvard U niversity, 1905.
A n iru d d h a -S a m h ita . Srec A n iru d h a S am h ita, o n e o f D i v y a s a m h ita in P a n c h a ra tra.
E d ite d by A. Srccnivasa Iyengar. M ysore, 1956.
A h irb u d h n y a -S a m h itii. A h ir b u d h n y a - S a m h ita o f the P a n c a r a tr a g a m a . E d i t e d by
M .D . R a m a n u ja c h a ry a u n d e r the S u p erv isio n o f F.O . S c h ra d e r. R e v is e d by V.
K rish n am ach ary a. 2 volum es. Madras; T h e A d y a r L ib rary and R e se a rc h Centre,
1916, 1966 (second edition), 1986 (reprint).
AR. A n a n d a ra m a y an a
S n v a lm i k i m a h a m u n i k r t a ^ a t a k o t i r a m a c a r i t a n t a r g a t a m a n a n d a r a m a y a n a m ... R am a te ja p a n d e y e n a k rta y a jy o ts n a b h id h a y a b h a s a tik a y a tikitam . E d ite d b y Y.K..
DvivedT. Varanasi: P anditapustakalaya, 1977 (second edition).
TSanaSivagurudevapaddhati. T h e is a n a s iv a g u ru d e v a p a d d h a ti b y is a n a S iv a g u ru d e v a m isra. Edited by T. G a n a p a ti Sastri. 4 volu m es. T riv a n d ru m : T r iv a n d r u m U n i
versity Press, 1920-1925, Delhi: B haratiya V id y a P rakashan, 1988 (reprint).
ISvara-Sam hita. T^varasamhita p ra tiv a d ib h a y a n k ara n a n ta c a ty a is sam Sodhita. Kanci:
SudarSanamudraksaraSala, 1923.
R g - V e d a . D ie H y m n c n dcs R ig v e d a . H c r a u s g e g e b e n von Th. A u fre c h t. 2 parts.
W iesb ad en : Otto H arrassow itz, 1968 (fourth reprint).
R g v e d iy a b r a h m a k a r m a s a m u c c a y a . Atha rg v c d ly a b r a h m a k a r m a s a m u c c a y a h . E d ite d
by GaneS Sastri 6cndye. P une: S.P, B a rv c ParaSuram , 1979.
K aly an am an d irasto tra. See B h aktam arastotra.
K alivilasa-T antra. K alivilasa Tantra. Edited by Parvati C h a ra n a T arkatirtha. L on d o n :
L u z a c & Co., 1917.
K iranavrtti. B h a tta ra m a k a n th a v ira c ita kiranavrttih, B h a tta R a m a k a n t h a s C o m m e n
tary on the K iranatantra. V o lu m e I: C h a p te rs 1-6: critical e d itio n and a n n o ta te d
translation < b y > D. Goodall. P ondichery: Institut Frangais de P o n d ic h e ry , 1998.
1
1 regret that so m e b ib lio g ra p h ic a l entries p e rta in in g to B r u n n e r s c o n trib u tio n
re m a in incom plete. 1 w a s u n a b le to ad d o r c o n firm n a m e s o f s o m e e d ito rs a n d /o r
p ublishers q u o te d in her contribution since the b o o k s w e re not acc e ssib le to me. T h e
tran sliteratio n sy stem from S outh Indian la n g u a g e s in the r e f e r e n c e s to B r u n n e r s
articlc follow s the sy stem she has u sed in h er earlier publications.
272
BI BLI OGRAP HY AND ABBREVI ATIONS
K iranagam a. Kiranatantra. Edited by T.R. PancapageSaSivacarya and K.M. SubrahmanyaSastrT. Devakottai: S iv agam asiddhantaparipalanasarigha, 1932 (in g ra n t ha
charactcrs).
K u b jik a-U p an isad . T h e K ubjika U panisad. E dited with a translation, introduction,
notes and ap p cn d iccs by T. G o u d riaan and .I.A. S choterm an. Groningen: Egbert
Forsten, 1994,
K u la rn a v a -T a n tra . K u la rn a v a ta n tra . In tro d u ctio n A. A v a lo n (S ir J. W o o d ro ffe ),
R e a d in g s M.P. P andit, S a n sk rit T e x t T a r a n a t h a V id y a ra tn a . D elhi: M otilal
B anarsidass, 1965 (reprint).
G a n e s a ta p a n iy a -U p a n is a d . Die G a n c S a tap an iy a-U p an isad . Teil 1: T e x t und Obersctzung. Teil 2: A ntncrkungen. Inauguraldissertation vorgelcgt von U. B crgm ann
(unpublished doctoral dissertation, subm itted to M arburg University in 1965).
C a tu rv a rg a c in ta m a n i II. C h a tu rv a rg a C h in ta m a n i by H em ad ri. V o lu m e II: VrataK h an d a. 2 parts. Part I. E d ited by B h a ra ta c a n d ra S irom ani. P art II. E d ite d by
B h a ra ta c a n d ra S irom ani, YajneSvara B hattacaryya and K a m a k h y a n a th a T arkaratna. Calcutta: GaneSa Press, 1978-1879.
Ja y a k h y a -S a m h ita . Jay a k h y a sa m h ita . Critical Edition with an Introduction in S a n s
krit, Indices etc. by E. Krishnam acharya. Baroda: Oriental Institute, 1931.
T a n tra sa d b h a v a. A draft edition o f selected passages p repared by J. T o rz so k based
on tw o m anuscripts from the National Archives, K ath m an d u (5-1985 and 5-445).
T an trasara o f A b h in av ag u p ta. Edited with notes b y M u k u n d a R am Sastri, B om bay:
N irnaya S a g a r Press, 1918.
TA. T antraloka (by A bhinavagupta)
T h e T a n tra lo k a o f A b h in a v a G u p ta, with c o m m e n ta ry b y R a ja n a k a Jayaratha.
Edited by M u k u n d R am Shastri. 12 volumes. A llah ab ad /B o m b ay : V c n k a tc sh v a r
S team Press c t a l . , 1918-1938.
G noli, R. 1999. A b h in av ag u p ta: L uce dei tantra, T an tralo k a, a cura di R. Gnoli.
Milano: Adelphi.
[C o m p le te ly revised version o f the earlier translation o f the T a n tra lo k a w hich
w a s p u b lis h e d as: G noli, R.: L u c e dellc S acre Scritturc (T an tralo k a). T orino:
U nione tipograflco-editrice torinense, 1972.]
Silburn, L . f and A. P ad o u x 1998. A b h in av ag u p ta: La L u m ie rc sur les T antras,
chapitres I a 5 du Tantraloka. Traduits ct com m entes. Paris: D e Boccard.
D ev im a h a tm y a . Devi M a h a tm y a m (G lory o f the D ivine M other). 700 M antras on Sri
D u rg a. < S a n s k rit Text a n d > E nglish T ra n sla tio n by S w am i J a g a d isw a ra n a n d a.
Madras: Sri R am akrishna Math, no date (fifth im pression).
N aradlya-S am hita. NaradTya Sam hita. Edited by R.P. C h audhary. Tirupati: K endriya
Sanskrit V idyapcetha, 1971.
Nityasoda& ikarnava. ( V a m a k e S v a rata n tra n ta rg a ta h ) nity aso d aS ik arn av ah . Srib h ask a ra ra y o n n itas etu b an d h ak h y av y ak h y asah itah . Edited by K.V . A bh y am k ar. Pune:
A n an d asram a, 1973.
N ityotsava. N ity o tsav a o f U m a n a n d a n a th a [Supplem ent to ParaSuram a-kalpa-sutra],
E d ite d b y A . M a h a d e v a Sastri. R e v is e d and e n la rg e d by T r iv ik r a m a T irtha,
B aroda: Oriental Institute, 1977.
N is v a s a . N iS v asatattv asam h ita. M a n u s c rip t 1-227, p re se rv e d in the N ational A r
chives, K a th m a n d u (see G oudriaan 1981: 3 3 -3 5 ).
NT. N etratantra
N T j . T h e N ctra T antram , with c o m m e n ta ry (U dd y o ta) by K shem araja. Edited by
M ad h u su d an Kaul. 2 volum es. B om bay: T a tv a-V iv ech ak a Press, 1 9 2 6-1939.
N T 2 - N etratan tram with the c o m m e n ta ry U d d y o ta o f K scm araja. Edited by V.
Dw ivedi. Delhi: Parim al Publications, 1985.
BIBLI OGRAPHY A ND A B BR EV I AT I ON S
273
P a ra m a -S a m h ita . P a ra m a s a m h ita f o f the P aiicharatra], E d ite d and tr a n s la te d into

English with an intro d u ctio n by S. K ris h n a s w a m i A iy a n g a r. B a ro d a : O rie n ta l
Institute, 1940,
P aratrim S ik a and laghuvrtti. La P a ra trim S ik a la g h u v rtti d c A b h i n a v a g u p t a . T e x te
traduit ct annote par A. Padoux. Paris: D c B occard, 1975.
P a d m a -S a m h ita . P ad m a S am hita. C ritically ed ite d by S. P a d m a n a b h a n . 2 volum es.
M adras: Pancaratra Parisodhana Parisad, 1 9 7 4 -1 9 8 2 .
P a r a m e S v a r a - S a m h ita . Sri P a ra m e S v a ra S a m h ita S ri G o v i n d a c a r y a i h s a m s k r ta ,
a nckavidha darâdibhih sam yojita ca. Srirangam : K o d a n d a ra m a r S annidhi, 1953.
PuraScaryarnava. P uraS caryam ava o f His M ajesty Shri Pratap Sinh Sah D e v K in g o f
N e p a l. E d ited by M. Jha. D elhi: C h a u k h a m b a S a n s k r it P r a t i s h t h a n , 1985
(reprint).
P u rv a -K a ra n a g a m a . S rim a t-P u rv a k a ra n a g a m a m . M ad ras (Q intatripettai): ^ i v a f i a N a potayantragalai, 1922 ( ka liyu g a 5023) (in grantha characters).
P au sS . P auskara-Sam hita
P a u s S ] . Sree Poushkara Samhita. O n e o f th e T h re e G e m s in P an ch aratra, E d ited
by S a m p a th k u m a r a R a m a n u ja M uni. B a n g a lo re : A. S r in iv a s a A i y a n g a r a n d
M .C. T hirum alachariar, 1934.
PausS2- Pauskara Samhita. Critically edited <and translated> b y P.P. Apte. < P art
l.> Tirupati: R ashtriya Sanskrit V idyapcctha, 1991.
P raty a b h ijn a h rd a y a o f K sem araja. Edited by J.C. Chatterji. Srinagar: A rch aeo lo g ical
and Research D epartm ent, 191 1.
B rh at-S am h ita
V a ra h a n iih ira s Brhat Sam hita with English T ranslation, E x h a u stiv e N o te s and
L ite ra ry C o m m e n t s b y M. R a m a k ris h n a B hat. 2 parts. D elhi: M o tila l B a n a r sidass, 1981-1982.
T h e B rhat-Sanhita; or, C o m p le te S y ste m o f N atural A stro lo g y o f V arah a-m ih ira.
T ranslated from Sanskrit into E nglish by H. Kern. Journal o f the R oyal A siatic
S o c ic ty o f G reat Britain and Ireland, new series, 4. 1870: 4 3 0 - 4 7 9 ; 5. 1871:
4 5 - 9 0 , 2 3 1 - 2 8 8 ; 6. 1872: 3 6 - 9 1 , 2 7 9 - 3 3 8 ; 7. 1875: 8 1 -1 3 4 .
B M . B h a d ra m a rta n d a
A th a b r h a j j y o t i s a r n a v a n t a r g a t e sa sth e m i s r a s k a n d h e b h a d r a m a r t a n d a k h y a h
sa p ta d a io d h y a y a h prarab h y ate. B o m b a y : SriverikateSvar Press, 1933 (seco n d
edition; first printed at A u ran g ab ad in 1883 and at B o m b a y in 1902).
B r a h m a y a m a la . M an u scrip t 3 -3 7 0 , p reserv ed in the N atio n al A rc h iv e s, K a th m a n d u
(see G oudrtaan 1981: 4 2 - 4 4 ) .
B h a k ta m a r a s to tr a . C a m a k ti m a n g a lm a y a a d b h u ta - n a v a s m a r a n a m . S r iv a r d d h a m a n
b h a k t a m a r a , n a v a s m a r a n a m , k a ly a n a r n a n g a l a s to tr a m . E d ite d b y G h a s lla ljl
M a h a ra j/M a d a n la lji M a h a ra j/S .D . G o m a n th . K arnkaroli: Sri H in d K a n u n P r i n
ting Press, 1943/1944 (third edition).
B h a rg av a-T an tra. B harg av a T antra. E d ited by R.P. C h au d h ary . A llah ab ad : R ash triy a
Sanskrit Sansthan, 1981.
M a ta iig a p a r a m e S v a r a g a m a . M a ta h g a p a r a m e ^ v a r a g a m a , a v e c le c o m m e n t a i r e de
B h a tta R a m a k a n th a . E dition c ritiq u e p a r N .R . B hatt. 2 v o lu m e s . P o n d ic h e ry :
Institut Frangais d lndologie, 1 9 7 7 -1 9 8 2 .
M a n u -S m rti. M a n u s m rti W ith the S a n s k rit C o m m e n t a r y M an v a rth a m u k ta v alT o f
K u llu k a B h atta. E d ite d by J.L. Shastri. D elhi: M otilal B a n a rsid a ss, 1983, 1990
(reprint).
M a n t r a m a h o d a d h i . M a n t r a m a h o d a d h i h s a tik a h , B o m b a y : S r i v e n k a t e ^ v a r S te a m
Press, 1910, 1962 (re-edited), 1983 (reprint).
274
BI BLI OGR AP HY AND ABBREVIATI ONS
M ayam ata. M a y a m a ta , traitc sanskrit d architecturc. Edition critiquc, traduction ct

n o tes par B. D a g c n s . 2 v o lu m e s. P o n d ic h c ry ; institut F rangais d In dologie,
1 970-1976.
[English version: M ayam atam : treatise on housing, architecture and iconography.
S anskrit text edited and translated by B. Dagcns. 2 volum es. N e w Delhi: Indira
G a n d h i N atio n al C e n tre for the A rts and M otilal B a n a rs id a s s P u b lish e rs Pvt.
Ltd., 1994.]
M alinivijaya/M alinivijayottara
M a lin lv ija y a . Sri M a lin lv ija y o tta ra T a n tra m . < E d itc d > by M a d h u s u d a n ECaul.
B o m b a y : T a tv a-V iv ech ak a Press, 1922.
M a lin lv ija y o tta r a . E l e c tr o n ic tex t p r e p a r e d by S. V a s u d e v a , c o n ta in in g his
u n p u b lish e d critical edition o f chapters 1-4, 7 and 1 117.z
M r g c n d ra g a m a
M r g c n d ra g a m a (E dition o f kriyapada and caryapiida). M rg cn d rag am a ( K riyapada
e t C aryapada) avee le co m m e n ta irc dc B h atta-N aray an ak an fh a. Edition critique
p a r N . R . Bhatt. Po n d ich ery : Institut Frangais d lndologic, 1962.
M r g c n d r a g a m a ( T r a n s l a t i o n o f k r iy a p a d a and ca rya p iid a ). M r g c n d r a g a m a .
Section d cs Rites et Section du C o m p o rtc m c n t, avee la vrtti de B h atta n a ra y an a kantha. T rad u ctio n , Introduction ct N otes par H. B runner. Po n d ich ery : Institut
Frangais d lndologie, 1985.
Y antraeintam ani. Y antracintam anih o f D am odara. Critically edited by H.-G. Turstig.
Stuttgart: Steiner Vcrlag, 1988.
YH. Y o ginlhrdaya. See P ad o u x 1994.
R a u r a v a g a m a . E d itio n c ritiq u e p ar N.R. Bhatt. 3 v o lu m e s. P o n d ic h c ry : Institut
Frangais d lndologie, 1961-1988.
L ak sm i-T an tra
L a k s m i-T a n tra . A P a n caratra A g a m a . E dited w ith Sanskrit G lo ss and Introduc
tion by V. K rish n a m a c h a ry a. M adras: T h e A dynr L ibrary and R esearch C entre,
1959, 1975 (reprint).
L a k s m i T antra. A Pan caratra Text. T ran slatio n and N o te s by S. G u p ta. Leiden:
E.J. Brill, 1972.
L a lita s a h a s ra n a m a . L a lita s a h a s r a n a m a ( o f the s e c o n d part o f B r a h m a n d a p u r a n a .)
W ith the C o m m e n t a r y S a u b h a g y a - B h a s k a r a o f B h a s k a ra ra y a, E d ite d by W.L.
S'astri Pans'ilcar. B o m b a y : N irnaya Sagar Press, 1927 (third edition).
V isv a m itra -S a m h ita . ViSvamitra S am hita. Critically ed ited by U. S h a n k a ra Bhatta.
T irupati: K cn d riy a Sanskrit V idyapcetha, 1970.
V is n u -S a rn h ita . T h e V isn u S a m h ita . E d ite d by M .M . G a n a p a ti Sastri. W ith an
E la b o ra te Introduction by N.P. Unni. Delhi: N a g P ublishers, 1991 (rev ised and
en larg ed edition; first edition 1925).
V isvaksena-S anihita. V isvaksena Samhita. Critically edited by L.N . Bhatta. Tirupati:
K e n d riy a Sanskrit V idyapeetha, 1972.
V inaikha. T h e V lnaîkhatantra. A Saiva T antra o f the L eft Current. Edited with an
in troduction and a translation by T. Goudriaan. Delhi: Motilal B anarsidass, 1985.
V r a t o d y a p a n a k a u m u d l . A th a c itta p a v a n a k u l a b d h i s a m b h a v a - g h a r c - i t y u p a n a m a k a b a llalasu nsun u -S am karaviracita v ra to dy apanak aum u dip raram b h ah . B om bay:
Jn an ad arp an Press, no date.
S ab d ak alp ad ru m a. See Deva, Raja R adha Kanta 1961.
2
j. T o rz s o k w o u ld like to e x p r e s s h er th a n k s to S. V a s u d e v a fo r m a k in g this
electronic text available to her.
BI BLI OGRAPHY AND A B B R EV I AT I O NS
275
Saktan an d ataran g in T . B r a h m a n a n d a g i r i s S a k a t a n a n d t a r a n g i n i . T e x t w ith H in d i

Translation by R.K. Rai. Varanasi: Pracliya Prakashan, 1993.
S a n tisa ra . Atlia S a n tisa ra p ra ram b h a h < d in a k a r a b h a tta k r ta h > . M u m b a i: V .V . G o d bolc, 1861.
ST. Saradatilaka
Saradatilakam o f Sri Lakm anadcikcndra [s/'c/] w ith Padarthjidarsa C o m m e n ta r y
by Srim ad R aghava Bhatta. Edited by M. Jha Bkashi. V aran asi: C h a u k h a m b h a
Sanskrit Sansthan, 1986 (third edition).
Sritattvacintam ani. P u r n a n a n d a s Sritattvacintam ani. C ritically E d ited fro m O riginal
M an u scrip ts (C h ap ters IXV111) w ith an original c o m m e n ta r y b y B h u v a n m o h a n
S a n k h y a tirth a and (C h a p te rs X I X - X X V I ) w ith N o te s by C h i n t a m a n i B h a tta ch a ry a . In tro d u c tio n by P ra b o d h c h a n d r a B ag ch i. D e lh i M o tila l B a n a r s i d a s s ,
1994 (reprint).
S r lp r a S n a - S a m h ita . r ip r a n a S a m h ita . E d ite d by S. P a d m a n a b h a n w i t h t h e
F o rew ord o f V. Raghavan. Tirupati: K cndriya Sanskrit V id y a p e e th a , 1969.
S a n a tk u m a ra -S a m h ita . S a n a tk u m a ra -S a m h ita o f the P a n c a r a tr a g a m a . E d ite d by V.
K rish n am ach ary a. M adras: T h e A d y a r L ibrary and R esearch C entre, 1969.
Satvata-S am hita
S a tv ata-S am h ita. W ith C o m m e n ta r y b y A la sin g a B h atta. E d ite d by V .V . D w i vedi. Varanasi: S am p u rn a n a n d Sanskrit V ishv av id y alay a, 1982.
H ikita, H. 1990. S attv ata S am h ita: A n A n n o ta te d T ra n s la tio n , C h a p te r 17 (1).
Sdlo-shu ken k y u kiyfi 21: 190-137.
Hikita, H, 1991. S attvata S am h ita: A n A n n o ta te d T ra n s la tio n , C h a p t e r 1 0 - 1 1 .
A ich ig ak u in daig ak u zen ken k y u jo kiyo 18 -1 9 : 3 4 0 - 2 9 4 .
Hikita, H. 1992. Sattvata S am h ita: A n A n n o ta te d T r a n s la tio n , C h a p t e r 17 (2).
S5t6-shu kenkyu kiy5 23: 214168.
S a tv ata-S am h ita-B h asy a. See Satvata-Sam hita.
S a r d h a tr i a tik a lo tta ra g a m a . S a r d h a triS a tik a lo tta ra g a rn a a v e c le e o m m e n t a i r e de
B h a tta R a m a k a n t h a . E d itio n c ritiq u e p a r N .R . B h a tt. P o n d i c h c r y : I n s titu t
Franâis d lndologie, 1979.
SiddhayogeSvarim ata. See T o rz so k 1999a and T o rz s o k forthcom ing.
Sid dhantaSckhara. U b h ay av c d a n ti-v iS v a n ath a k rta h siddhantaS ekharah. E d ite d b y S.
Som ayajin. M ysore: T a n d a v a m u rti M udranalaya, 1971.
S i d d h a n ta s a ra v a li (w ith A n a n ta S a m b h u s c o m m e n ta ry ). S id d h a n ta s a r a v a li o f T r ilocan aS iv acary a w ith c o m m e n ta r y o f A n a n ta S iv a c a ry a. E d ite d by A .A . R a m a n a t h a n et al. 6 p a rts . G o v e r n m e n t O r ie n ta l M a n u s c r i p t s L i b r a r y B u l l e t i n
(M a d r a s ) 17.1: 2 9 - 6 8 ; 17.2: 1 -4 8 ; 18.2: 1 - 6 4 ; 19.1: 5 3 - 8 4 ; 19.2: 1 -4 8 ; 2 0.2:
4 9 - 7 1 . 1965-1972.
S u p r a b h e d a g a m a . (N o e d ito r a c c r e d ite d .) M a d r a s ((întatripettai): QMvanaNapotayantra<?alai, 1929 (ka liy u g a 5030) (in grantha characters).
S u ry a sid d b a n ta by V a rah am ih ira
S u ry a -S id d h a n ta : A T ext B o o k o f Hindu A stro n o m y . T r a n s la te d b y E. B u rg ess.
N e w H aven: A m erican O riental S ociety, 1860.
S au n d a ry a la h a ri. S a u n d a r y a - L a h a r i o f Sri S a m k a ra c a ry a . W ith C o m m e n t a r i e s Saub h a g y a v a r d h a n i o f K aiv a ly a ra m a, L a k s m id h a r a o f L a k s m l d h a r a c a r y a , A r u n a m o d in l o f K am cS v arasu rin , E n g lis h T ra n s la tio n and N o t e s b y R. A n a n ta k r s n a
Sastrl and K.R, Garu. P ra y o g a s illustrated with Y antras. M ad ras: G a n e s h & C o
1957.
276
BI BLIOGRAPHY AND A BBREVI ATIONS
SP. S o m aiam b h u p ad d liati

S P I , SP2, SP3, SP4. S o m aS am b h u p ad d h ati. T cxtc, trad u c tio n ct notes p a r H.
B r u n n c r - L a c h a u x . 4 v o lu m e s . P o n d ic h c r y : In stitu t F ra n g a is d l n d o l o g i c ,
1963-1998.
S p an d an irn ay a. The S p a n d a k a rik a s o f V asugupla. With the N irn ay a by K scm araja.
E d ited by M a d h u su d a n Kaul Shastri. Srinagar: K a s h m ir Pratap S team P ress,
1925.
SvT. S vacchandatantra
S v T ] . T h e S w a c c h a n d a -T a n tra , w ith c o m m e n ta r y o f K sh c m a ra ja . E d ited w ith
n o te s by M a d h u s u d a n K aul. 6 v o l u m e s . B o m b a y : N i r n a y a - S a g a r Press,
1 9 2 1 -1 9 3 5 .
'
SvT2- S v a c c h a n d a ta n tra . S v a c c h a n d a ta n tr a m w ith c o m m e n ta r y U d d y o t a by
K sem araja. Edited by V.V. Dwivcdi, Delhi: Parimal Publications, 1985.
S v a y a m b h u v a s u tra s a m g ra h a . E lectro n ic text prep ared by D. G o o d all, based on the
M y s o re ed itio n o f 1937 (edited by V cnkatasubrahm anyaS aslrl), with corrections
an d notes.1
Secondary Sources
A bbott, J. 1932. T h e K ey s o f Power: A S tudy o f Indian Ritual and Belief. L on d o n :
M eth u en & Co.
A c h a r y a , P.K. 1946. An E n c y c lo p e d ia o f H in d u A rc h ite c tu re. L o n d o n : O x fo rd
U n iv ersity Press.
A p te, P.P. 1973. M a n d a la r a d h a n a (D ia g ra m W o rsh ip in A g a m a s), Journal o f the
O riental Institute (B aroda) 22: 5 0 1 -5 2 4 .
Apte, P.P. 1975. M a n d a la -D ia g ram s (An A rtists A pproach). Pro ceed in g s o f the AllIndia Oriental C onference, Poona: 3 6 7 -3 7 4 .
A p te , P.P. 1986. V a s tu -P u ru s a -M a n d a la - A M ic r o - C o s m ic A rc h ite c tu r a l Plan.
S e m in a r on V a a stu sh a stra in P e rsp ectiv e, January 27, 28, 1986, o rg a n iz e d by:
P u n e C o n s t r u c t i o n E n g i n e e r i n g R e s e a rc h F o u n d a t i o n , P u n e -4. In stitu te o f
E n g in e e rin g (India). P u n e Local C entre, P u n e -5. V aastu V cd ic R e se a rc h F o u n
dation, M adras. M ah arash tra C haptcr, Pune: 1-35.
A p te, P.P. 1987. T h e S c h c m c o f M a n d a la -D ia g ra m s and V a s tu p u ru s a -M a n d a la in
the V aisn av a A g a m a : Structure, C o lours, R itual and A rchitectural A spects. In:
V a is n a v is m in In d ia n A rts and C u ltu re (C o lle c te d P a p e rs o f the U n iv e rs ity
G ra n ts C o m m is s io n N ational S e m in a r on Im pact o f V a is n a v is m in the Indian
A r ts ). E dited by R. Parim oo. N e w Delhi: B o oks & Books: 127-149.
A p te, P .< P .> 1999. T h e In tro d u c tio n to Ja y a k h y a S a m h ita by B c n o y to s h B h a tta c h ary y a and Indological C h ronology, In: T an tric B u d d h ism : C entennial T rib u te
to Dr. B e n o y to s h B h a tta c h a ry y a . E d ited b y N .N . B h a tta c h a r y y a , N e w D e lh i:
M anohar: 1722.
Apte, P.P. an d S.G. S u p e k a r 1983. V a s tu -P u ru s a -M a n d a la in the Pauskara-S ariihita
an d B r h a t - S a m h i t a - A C o m p a r a tiv e S tudy. B u lle tin o f the D e c c a n C o lle g e
R e s e a rc h Institute 43: 5 - 1 7 . [This article also ap p e a re d in: A g a m a and Silpa.
P r o c e e d i n g s o f th e S e m i n a r H eld in D e c e m b e r , 1981. E d i t e d b y K .K .A .
V e n k a ta c h a ri. B o m b a y : A n a n th a c h a r y a Indological R e s e a rc h Institute, 1984:
1 3 2 -1 4 8 .]
3
J. T o rz s o k w o u ld like to express her thanks to D. G oodall for m ak in g this text
available to her.
BI B LI OG RAP HY A ND A B B R E V I A T I O N S
277
A pte, R.N. 1926. S o m e P oints C o n n ected w ith the C o n s tru c tiv e G e o m e tr y o f V e d ic

Altars. Annals o f the B han d ark ar Institute 7: 1-16,
A p te , V.S. 1957. T h e Practical S a n s k rit-E n g lis h D ic tio n a ry . R e v is e d & E n la r g e d
Edition, Poona: P rasad P rak ash an , K yoto: R in scn B o o k C o m p a n y , 1992 (third
reprint).
^
A rraj, W J . 1988. T h e S v acch an d atan tram : History and S tructure o f Saiva S cripture,
(u n p u b lish ed doctoral d issertation, s u b m itte d to th e D e p a r tm e n t o f S outh A sia n
L an g u ag es and Civilizations, the U niversity o f C hicago).
A u e r, G. and N. G u ts c h o w 1974. B h a k ta p u r : G e sta lt, F u n k t i o n e n an d re lig io s e
S y m b o lik e in c r n e p a lis c h e n S tadt im v o r in d u s tr ic lle n E n t w i c k l u n g s s t a d i u m .
A u f g c z c ic h n c t und in tc rp rc tie rt von G. A u e r an d N. G u t s c h o w . D a r m s t a d t :
T ech n isc h c Universitat.
B a f n a , S. 2000. O n the Idea o f the M a n d a la as a G o v e r n i n g D e v ic e in I n d ia n
A rch itectu ral T ra d itio n . Journal o f the S o c ie ty o f A r c h ite c tu ra l H is to ria n s 59:
26-49.
B a g , A.K . 1971. T h e K n o w le d g e o f G e o m etrical F ig u res, In stru m e n ts and U nits in
the Sulbasutras. East and W est 21: 111-1 1 9 .
B a h u lk a r, S. 1979. C o n c e p t o f D h a rm o d a y a ( C h o s H b y u h ). R e p o rt o f the J a p a n e s e
A ssociation for T ibetan Studies 25: 13-16.
B a n e r ji, S.C. 1978: T a n tr a in B e n g a l: A S tu d y in its O rig in , D e v e l o p m e n t an d
Influence. Calcutta: Naya Prokash.
B a n g d e l, D. 1999. M a n ife s tin g the M a n d a la : A S tu d y o f th e C o r e I c o n o g r a p h ic
P r o g r a m o f N e w a r B u d d h is t M o n a s te r ie s in N e p a l. 3 v o lu m e s ( u n p u b l i s h e d
doctoral dissertation, su b m itte d to the D e p a rtm e n t o f H isto ry o f Art, O h io State
University).
Baurner, B. 1986. P an jara ct yantra: le d ia g ra m m e de l im age sacree. In: M a n tra s et
d i a g r a m m c s ritu c ls d ans Ih in d o u is m e . T a b l e R o n d e , P aris 2 1 - 2 2 j u i n 1984.
< E d itc d by A. P a d o u x .> Paris: E d itio n s du C e n tre n a tio n a l de la r e c h e r c h e
scientifique: 4 9 - 6 1 .
B e c k m a n , H. 1996, M an tras, Y an tras & F a b u lo u s G e m s . T h e H e a lin g secrets o f the
ancicnt Vedas. [Hastings:] Balaji P u b lish in g Co.
B e g le y , W .E. 1973. V i s n u s F l a m in g W h e e l: T h e I c o n o g r a p h y o f th e SudarsanaCakra. N e w York: N e w Y ork U niv ersity Press.
B e g u in , G. 1990. A rt e so tc riq u e de l H im a la y a : C a ta lo g u e d e la d o n a tio n L io n e l
Fournier. Paris: E d itio n s de la R e u n io n dcs n iu se e s nationaux.
B e rn ie r, R .M . 1978, W o o d e n W in d o w s o f N ep al A n Illustrated A n a ly sis. A r tib u s
A siae 39: 2 5 1 - 2 6 7 .
B ern ier, R .M . 1979. T h e N e p a le s e Pagoda: O rig in s and Style. N e w D elhi: S. C h a n d
& C o m p a n y Ltd.
B h a n d a rk a r, R,G . 1965. V a is n a v is m , S a iv is m and M in o r R e lig io u s S y ste m s. V a r a
nasi: Indological B o o k House.
B h a tta c h a r y y a , T. 1986. T h e C a n o n s o f Indian A rt, or, A S t u d y on V a s tu v id y a .
C a lcu tta: F ir m a K.L. M u k h o p a d h y a y P riv a te Ltd. (th ird r e v is e d an d e n l a r g e d
edition).
Bizot, F. 1981. N o te s sur les ya n tra b o u d d h iq u e s d indochine. In: T a n tric an d T a o is t
Stu d ies in H o n o u r o f R.A . Stein. E d ited b y M. S tric k m a n n . B ru x e lle s : In stitu t
beige des hautes e tu d es chinoises. V o lu m e 1: 155-1 9 1 .
B o d e w itz , H .W . 1979. T h e R aising o f the Central P illar (M a d e o f U d u m b a r a W o o d )
o f the Sadas Hut by the U dgatr, a c c o r d in g to J a i m i m y a B r a h m a n a 1.70-72. In:
L u d w ik S te rn b a c h F elicitatio n V o lu m e . Part 1. E d ite d b y J.P. S inha. L u c k n o w :
A khila B haratiya Sanskrit Parishad: 7 7 - 8 2 .
27 8
BI BLIOGRAPHY A N D ABBREVI ATI ONS
Bolton, N.J. and D .N.G. M acleod 1977. The G e o m e try o f the Sri-Y antra. Religion:
Journal o f Religion and Religions 7: 6 6 -8 5 .
B o n e r, A. 1975. E x tr a c ts from the S ilp a s a rim , In: S tu d ie s in Indian T e m p l e
A rchitecture. P a p e rs p resented at a S e m in a r held in V aranasi, 1967. E dited by
P ram o d C handra. N ew Delhi: A m erican Institute o f Indian Studies: 5 7 -7 9 .
B ourquin, A. 1884. B ra h m a k a rm a ou rites sacres dcs B ralunancs, traduit du Sanscrit
et annotc. A nnales du M u see G uim et 7. Paris: Ernest Lcroux,
B rooks, D.R. 1990. T he S c c rc t o f the T h ree C itie s: An Introduction to Hindu Sakta
Tantrism . C hicago: The University o f C hicago Press.
B rooks, D.R. 1992. A uspicious W isdom : T h e T exts and T raditions o f S iiv id y a Sakta
T a n tr is m in S outh India. A lb an y , N e w Y ork: Slate U n iv e rs ity o f N e w Y o rk
Press.
B ro w n , W .N . 1942. T h e C reation M yth o f the Rig Veda. Journal o f the A m e ric a n
O riental S ociety 62: 8 5 - 9 8 (reprinted in R o ch er 1978:
2 0 -3 3 ).
B ro w n , W .N . 1965. T h eo ries o f C reation in the Rig Veda. Journal o f th e A m e ric a n
Oriental Society 85: 2 3 - 3 4 (reprinted in Rochcr 1978:
4 0 -5 2 ).
B ru n n e r, H. 1975. Le sadhaka, p e rs o n n a g c o u b lic du
SivaTsme du Sud. Journal
A siatiq u e 263: 411443.
B ru n n e r, H. 1977. Im p o rtan ce dc la littcrature a g am iq u c p our 1ctudc dcs relig io n s
vivantes dc Ilndc. In: Indologica T aurinensia 111IV. Torino: 107-1 2 4 .
B r u n n e r , H. 1986. M a n d a la et y a n tra d ans le Sivai'sme a g a m iq u c : D e f in itio n ,
descrip tio n , usage rituel. In: M antras et d ia g ra m m e s ritucls dans 1hindouism e.
T a b le R onde, Paris 2 1 - 2 2 ju in 1984. < E dited by A. P a d o u x .> Paris: E d itio n s du
C entre national dc la recherche scientifiquc 11-31.
B ru n n e r, H. 1992. Jn a n a and Kriya: R elation b etw een T h e o ry and P ractice in the
S a iv a g a m a s. In: Ritual and Speculation in E arly T antrism . Studies in H o n o r o f
A n d re P ad o u x . Edited by T, G oudriaan. A lbany, N ew York: State U n iv ersity o f
N e w Y o rk Press: 1-59.
B i ih n e m a n n , G. 1983. B u d h a - K a u S ik a s R a m a ra k s a s to tra : A C o n trib u tio n to the
S tudy o f Sanskrit D evotional Poetry. Vienna: Institute for Indology.
B iih n e m a n n , G. 1987. B h a d ra m a n d a la s in the Ritual Practicc. W ie n e r Z citsch rift fur
die K u n d e Siidasiens 31: 4 3 - 7 3 .
B iih n e m a n n , G. 1988a. P u ja : A S tu d y in S m a rta R itual. V ie n n a : In stitu te fo r
Indology.
B iih n e m a n n , G. 1988b. T h e W o rsh ip o f M ahaganapati acco rd in g to the N ityotsava.
W ichtrach: Institut fiir Indologie.
B iih n e m a n n , G. 1992. On Purascarana: K ularnavatantra, C h a p te r 15. In: Ritual and
S p ecu latio n in E arly Tantrism . Studies in H o n o r o f A n d re Padoux. E d ited b y T.
G oudriaan. A lb an y , N e w York: State University o f N e w Y ork Press: 6 1 - 1 0 6 .
B iih n em an n , G. 2 0 0 0 - 2 0 0 1 . T h e Ic o n o g ra p h y o f H in d u T antric Deities, 2 volum es.
V o lu m e I: T h e P an th e o n o f the M an tram ah o d ad h i. V o lu m e II: T h e P an th e o n s o f
the P rap an casara and the Saradatilaka. G roningen: Egbert Forsten.
B iih n e m a n n , G. 2003. Smarta. In: T h e Hindu W orld, E d ite d b y S. M ittal and G.R .
Thu rsb y , L o ndon: R ou tled g c (forthcom ing).
B u lc k e , C. 1962. R a m - K a th a (utpatti aur vikas). Prayag: H in d i P arisad PrakaSan
(second edition).
B u n c e , F .W . 2 0 0 1 . T h e Y a n tr a s o f D e i t i e s an d t h e i r N u m c r o l o g i c a l F o u n
dations an Iconographic Consideration. N ew Delhi: D.K. Printw orld (P) Ltd.
B u n c e , F .W . 2 0 0 2 . T h e I c o n o g ra p h y o f A rch ite c tu ral Plans: A S tu d y o f the In
fluence o f B u d d h is m and H induism on Plans o f S outh and S o u th e a st Asia. N e w
Delhi: D.K. P rintw orld (P) Ltd.
BI BLI OGRAP HY A N D A B BR E V I A T I O N S
279
C airn s, G.E, 1992. Man as M ic ro c o sm in T a n tric H in d u ism . N e w D e lh i: M a n o h a r

Publications,
C a m m a n n , S. 1969. Islamic and Indian M agic Squares. 2 parts. H istory o f R e lig io n s
8 :1 8 1 -2 0 9 ,2 7 1 -2 9 9 .
C h a k ra v o rty Spivak, G. 1999, M o v in g Devi. In: D evi, T h e G re a t G o d d e s s : F e m a le
Div inity in South Asian Art. Edited b y V. Dehcjia. W ash in g to n , D.C.: A rth u r M.
Sackler Gallery: 181-200.
C h atto p a d h y a y a , D. 1973. L okayata: A Study in A n c ic n t Indian M a terialism . N e w
Delhi: P e o p le s Publishing H ouse (third edition).
C hatto p ad h y ay a, S. 1978. Reflections on the T antras. Delhi: M otilal B anarsidass.
C h a w d h ri, L.R. 1990. Practicals o f Yantras. N e w Delhi: S a g a r P u b lic a tio n s (se c o n d
revised and enlarged edition).
C h a w d h ri, L.R. 1992. S ecrets o f Y an tra, M a n tra and T a n tra . N e w D elhi: S terlin g
Publishers,
C h e n c t, F. 1986, D c l e ff ic ie n c c p s y c h a g o g i q u e d es m antras et d e s y a n tr a s . In:
M antras et dia g ra m m c s ritucls dans l hind o u ism e. T a b le R o n d e, Paris 2 1 - 2 2 ju in
1984. <Edited by A. P ad o u x .> Paris: E ditions du C entre national de la re c h e rch e
scientifique: 6 5 -7 8 .
C o b u rn , T.B. 1991. E nco u n terin g the Goddess: A T ra n sla tio n o f th e D e v i-M a h a tm y a
and A Study o f Its Interpretation. A lbany: State U niversity o f N e w Y ork Press,
C olas, G. 1986. La vision de la divinitc dans les d ia g ra m m e s selon le v is h n o u is m e
vaikhanasa. In: M antras et d ia g ra m m e s rituels d ans P h in d o u is m e . T a b le R o n d e,
P aris 2 1 - 2 2 j u i n 1984. < E d itc d b y A. P a d o u x > P a ris : E d itio n s du C e n tr e
national dc la recherche scientifique: 8 3 -9 5 .
C o m o , C. 1986. Les d ia g ra m m e s c o sm o g o n iq u c s du Sv acch an d atan tra: P e rs p e c tiv e s
p h ilo s o p h iq u e s . In: M a n tra s et d ia g r a m m e s r itu c ls d a n s Ih in d o u is m e . T a b le
R onde, Paris 2 1 - 2 2 ju in 1984. < E d ited by A. P a d o u x .> Paris: E d itio n s du C en tre
national de la reehcrchc scientifique: 9 9 - 1 1 3 .
C o o m a r a s w a m y , A .K . (1 9 2 7 ) 1985, H isto ry o f Indian a n d I n d o n e s ia n Art. N e w
York: D o v e r Publications.
C orrea, C. 1996, C harles Correa. London: T h a m e s & H udson.
D a b ra l, L. 1934. S iiyantra ka svarup. K aly an (G ita Press, G o ra k h p u r), Sakti aiika:
' 592-609.
D a k s h i n a r a n j a n S h astri 1940. A lta rs, D i a g r a m s , etc. in th e R itu a l o f A n c e s to r W orship. jo u r n a l o f the Indian S o ciety o f Oriental A rt 8: 166173.
D a k s h in a r a n ja n Shastri 1963: O rigin and D e v e l o p m e n t o f th e R itu a ls o f A n c e s to r
W o rsh ip in India. Calcutta: B o o k la n d Private L im ited.
D an iclo u , A. 1964. Hindu P olytheism . L ondon: R o u tle d g e & K e g a n Paul.
[S lig h tly d ifferent from; A. D a n ic lo u : L e p o l y t h e i s m e h in d o u . Paris: B u c h e t /
Chaste!, I960.]
D a n ic lo u , A. 1977. Lc te m p le hindou: C e n tre m a g iq u e du m o n d e , Paris: B u c h e t/
Chastcl.
[English translation: T h e Hindu T em p le: D eification o f E ro ticism . T ra n sla tio n b y
K. Hurry. R och ester, V erm ont: Inner T raditions, 2001.]
D a tta , B. 1932. T h e S c ie n c e o f the u lb a : A S tu d y in E a r l y H i n d u G e o m e tr y .
Calcutta: T h e U niversity o f Calcutta.
D e M allm ann, M ,-Th. 1986. In tro d u c tio n a 1ic o n o g ra p h ie du ta n trism e b o u d d h iq u e .
Paris: J. M aisonneuve,
D eep , D .K . 1993. P o p u la r Deities, E m b le m s & Im a g e s o f N e p a l. J a ip u r /N e w D elhi:
N ira la Publications.
280
BI BLI OGRAP HY AND ABBREVI ATI ONS
D e h e jia , V, 1986. Y oginI C u lt and T e m p le s : A T a n tric T ra d itio n . N e w Delhi:

National M useum .
D eva, R aja R adha Kanta 1961. S a b d a -K a lp a d ru m : An E n c y c lo p c d ic D ictio n ary o f
Sanskrit Words. 5 parts. Varanasi: C h o w k h a m b a Sanskrit Series Office.
D cvaraj V id y av acasp ati 1934. T an tra m e m yantra aur mantra. Kalyan (G ita Press,
G orakhpur), Sakti aiika: 3 8 7 -3 9 7 .
D ietrich, A. 1998. T an tric Healing in the K a th m a n d u V alley: A co m p a ra tiv e study
o f H in d u and B u d d h is t spiritual h e a lin g traditions in urb an N epalese socicty.
Delhi: B o o k Faith India.
D rege, J.-P. 1 9 9 9 -2 0 0 0 . Les p re m ie re s im p ressio n s des dharani dc M ahapratisara.
Cahiers d E x trcm c-A sic 11: 2 5 - 4 4 .
D y c z k o w sk i, M.S.G . 1988. T h e C anon o f the S aivagam a and the K ubjika T antras o f
the W estern K au la Tradition: Albany, N e w York: State U niversity o f N ew Y o rk
Press.
D y c z k o w s k i , M .S .G . 200 1 . T h e C u lt o f the G o d d e s s K u b jik a : A P r e lim in a r y
C o m p a r a tiv e T e x tu a l and A n th r o p o lo g ic a l S u rv e y o f a Secret N e w a r K aula
G oddess. Stuttgart: F ra n z Steiner Verlag.
Elder, G.R. 1978. T h e S am p u ta Tantra: Edition and T ranslation, C hapters 1-1V (u n
p u b lish e d doctoral dissertation, su b m itted to the F aculty o f Philosophy, C o l u m
bia University).
Eliade, M. 1969. Yoga: Im m ortality and Freedom . Princeton, N ew Jersey: Princeton
U n iv ersity Press (se c o n d edition with corrections and additional b ibliographical
notes).
E nglish, E. 2002. V ajrayogini: H e r Visualizations, Rituals, & Forms. A S tudy o f the
C ult o f V ajray o g in i in India. Somervi lle, M assachusetts: W isd o m Publications.
F o n seca, R. 1986. C o n stru c tiv e G e o m e try and the S ri-C akra D iagram . R eligion 16:
3 3 -4 9 .
G a b o rie a u , G. 1993. D es dieux dans to utes les directions: C o n c e p tio n in dienne de
l espace et classification des dieux. In: C lasser les dieux? D es pan th eo n s e n Asie
d u Sud. C la s s if y in g G o d s ? P a n th e o n s in S o u th Asia. E tu d e s r e u n ie s p a r V.
B o u i l l i e r e t G. Toffin. Paris: C ollection Purusartha: 2 3 - 4 2 .
G a e ffk e , P. 1987. B in d u M a n d a la s. In: T h e E n c y c lo p e d ia o f R elig io n . E d ito r in
C h ie f: M . E lia d e . V o lu m e 9. N e w Y ork: M a c M illa n P u b lis h in g C o m p a n y :
153-1 5 5 .
Gail, A.J. 19841988. Ternpel in Nepal. 2 volumes. Graz: A k a d e m isc h e D ruck- und
Verlagsanstalt.
Garzilli, E. 2000. F lo w e rs o f C o n sc io u sn e ss in T a n tric Texts: T h e Sacred Lotus, In:
P a n d a n u s (2000): Natural S y m b o lism in Indian L iteratures. E d ited by J. Vacck.
Prague: Signeta: 7 3 -1 0 2 .
G h o s h a l Sastri, S.N . 1983. E le m e n ts o f In d ian A e s th e tic s (in three V o lu m e s ).
V o lu m e II: T w o S tream s o f Indian Arts (in four parts). Part IV: P rim itiv e Arts,
C r a f t s a n d A l p a n a ( a c c o m p a n i e d w ith t h r e e s e p a r a t e p a rts ). V a r a n a s i:
C b a u k h a m b h a Orientalia.
G o d e , P.K . 1947. H isto ry o f th e ran g a v a lli (rangoH) art b e tw e e n c. A .D . 50 and
1900. A n n a l s o f the B h a n d a r k a r O rie n ta l R e s e a r c h In stitu te 28: 2 2 6 - 2 4 6 .
[ R e p r in te d in: S tu d ies in In d ian C u ltu ral H isto ry b y P.K. G ode. V o l u m e 3.
P oona: B h a n d a rk a r O riental R esearch Institute, 1969: 8 7 -1 0 2 .]
G o n d a , J. 1977. M e d ie v a l R e lig io u s L i te r a tu r e in S an sk rit. W i e s b a d e n : O tto
H arrassow itz.
G onda, J. 1980. V edic Ritual: T h e N o n -S o le m n Rites. Leiden: E.J. Brill.
BI BLIOGRAPHY A N D A B B R E V I A T I O N S
281
G o o d a ll, D. 2000, P r o b le m s o f N a m e and L in e a g e : R e l a tio n s h ip s b e t w e e n S o u th

Indian auth o rs o f th e S a iv a S id d h an ta. Jo u rn a l o f th e R o y a l A s ia tic S o ciety .
Scries 3. 10: 2 0 5 -2 1 6 .
G o u d ria a n , T. 1978. M ay a D ivine and H u m an : A stu d y o f m a g ic and its re lig io u s
foundation in Sanskrit texts, w ith p articu lar atten tio n to a f r a g m e n t on V i s n u s
m ay a preserved in Bali. Delhi: Motilal B anarsidass.
G o u d ria a n , T. and S. G u p ta 1981. Hindu T a n tric an d S ak ta L iterature. W iesb a d e n :
Otto Harrassowitz,
G ra n o ff, P. 2000. O th e r P e o p l e s R itu als: R itu a l E c le c tic is m in E a rly M e d ie v a l
Indian Religious [s/c/]. Journal o f Indian P h ilo so p h y 28: 3 9 9^424.
G ranoff, P. 2001. My Rituals and M y G ods: Ritual E x c lu s iv e n e s s in M e d ie v a l India.
Journal o f Indian Philosophy 29: 109-134.
G u p ta , A. 1984. A n a n d a - R a m a y a n a k a s a m sk rtik a d h y a y a n . Dill!: E a s t e r n B o o k
Linkers.
G u p ta , S. 1976. V is v a k s e n a the D iv in e Protector. W ie n e r Z e its c h rift fu r die K u n d e
Siidasicns 20: 7 5 - 8 9 .
G u p ta , S. 1988. T h e M a n d a la as an Im a g e o f M a n . In: In d ia n R itu a l a n d Its
E x e g e s is . O x f o r d U n iv e r s ity P a p e rs on In d ia , v o l u m e 2/2. E d i t e d b y R .F.
G om brich. Delhi: O x fo rd University Press: 3 2 - 4 1 .
G upta, S./D.J. H oens/T. G o u d riaan 1979. H indu T antrism . L eid en : E.J. Brill.
G u ts c h o w , N. 1982. S ta d tra u m und Ritual d e r n e w a r is c h e n Stadte im K a th m a n d u Tal: E inc a rch itek tu ran th ro p o lo g is ch e U ntersuchung, Stuttgart: W . K o h lh a m m e r.
G u ts c h o w , N. 1997. T h e N e p a le s e C aitya: 1500 Y e a r s o f B u d d h i s t V o tiv e A r c h i
tecture in the K a th m a n d u Valley. S tuttgart/London: E d itio n A xel M en g es.
G u ts c h o w , N. and B. K o lv e r 1975. O r d e r e d S p ace: C o n c e p ts an d F u n c tio n s in a
T o w n o f Nepal. W iesbaden: F ra n z Steiner G m b H .
G u ts c h o w , N ./B . K o lv e r /I .S h r e s th a c a r y a 1987. N e w a r T o w n s and B u ild in g s : A n
Illustrated D ic tio n a ry N e w a r i - E nglish. S ankt A u g u stin : V G H W is s e n s c h a fts verlag.
Havell [C handra] 1964.
H a v e ll, E .B .: T h e A rt H e r ita g e o f India. C o m p r i s i n g In d ian S c u lp tu r e a n d
P a in tin g a n d Id e a ls o f In d ian Art. R e v is e d E d itio n w i t h N o te s by P r a m o d
Chandra. B o m b ay : D.B. T a ra p o re v ala S o n g s & Co.
H eestcrm an, J.C. 1962. V ratya and Sacrifice. In d o-Iranian J o u rn a l 6: 1-37.
Hcilijigcrs-Seclen, D. 1990. T h e D o c trin e o f the S atcak ra a c c o r d in g to th e K u b jik a m ata. In: T h e S a n s k rit T r a d itio n and T a n tr is m . P a n e ls o f th e V l l t h W o r ld
S anskrit C o n fe re n c e , K ern Institute, Leiden: A u g u s t 2 3 - 2 9 , 1987. E d ite d b y T.
G oudriaan. Leiden: E.J. Brill: 5 6 -6 5 ,
H eilijgers-S cclcn, D. 1994. T h e S ystem o f F iv e C ak ras in K u b jika m a ta ta n tra 14-16.
G roningen: E g b e rt Forsten.
Hikita, H. 1990, 1991, 1992. See S atvata-Sam hita.
H o o y k a a s , C. 1980. D r a w in g s o f B a lin e se S orcery, p u b lis h e d w ith a C o m m e n ta r y
Leiden: E.J. Brill.
H uyler, S.P. 1999. M e e tin g God: E le m e n ts o f H in d u D e v o tio n . N e w H av e n : Y a le
U niversity Press.
Ja in , J. 1997. G a n g a D ev i: T r a d i t i o n an d E x p r e s s i o n in M i t h i l a P a i n t i n g s .
M id d leto w n , N e w Jersey: G ran th a C orporation.
Jh a v c ry , M.B. 1944. C o m p a r a tiv e and Critical S tudy o f M a n tra s a s tra (W ith Special
T re a tm e n t o f Jain M a n tra v a d a ), B e in g th e I n tro d u c tio n to Sri B h a ir a v a P a d m a vati Kalpa. A h m cd ab ad : Sarabhai M anilal N aw ab .
Johari, H. 1986. T ools for T antra. R och ester, V erm ont: D e stin y B ooks.
282
BI BLI OGRAPHY AND ABBREVI ATI ONS
Jones, C.R. 1981. Dhulicitra: Historical P er sp e ctiv e s on Art and Ritual. In: Kaladarsana: A m e rica n S tu dies in the Art o f India. Edited by J.G. W illia m s . N e w
Delh i: Oxford & IBM Publishin g Co.: 6 9 - 7 5 .
Joshi, H.R. 1981. Shatkon. R olam ba (Tri-m onthly research journal, Joshi R esearch
Institute, Lalitpur, N ep a l) 1: 2 - 4 .
Jung, C.G . 1950. U b e r M a n d a la s y m b o lik . In: G e s ta ltu n g e n d e s U n b e w u s s t c n .
Zurich: R ascher Verlag: 1 8 7 - 2 3 5 .
[E n g lish translation: C o n c er n in g Mandala S y m b o lis m . In: T h e A rc h e ty p es and
the C o lle c t iv e U n c o n s c io u s . T h e C o lle c te d Works o f C.G. Jung, v o lu m e 9, part
1. London: R outledgc & K egan Paul, 1953 (sec o n d edition): 3 55ff.]
Jung, C .G . 1964. D i e M a n d a la s y m b o li k . In: T r a u m s y m b o l c d cs In d iv id u a tio n s p rozesscs. Olten/Freiburg im Breisgau: W alter-Verlag: 10 3 ff.
K agal, C. (cd .) 1986. Vistara: T h e A rchitecture o f India. C a ta lo g u e o f the E x h i
bition. Bom bay: Festiv al o f India.
Kahrs, E. 1998. Indian s e m a n ti c an a ly sis. T h e n irv a c a n a 1 tradition, C am b ridge:
C am b ridge U niversity Press.
Kane, P.V . 1 9 6 8 - 1 9 7 7 . H istory o f Dharma&astra (A n cicn t and M ediaeval R e li g io u s
and C iv il L a w in India). 5 v o lu m e s . P o o na : Bhandarkar Oriental R es e a rc h
Institute (seco n d edition).
K e n o y e r, J.M. 1998. A n c ien t C ities o f the Indus V a lle y C iv iliza tio n . Karachi: O x
ford U n iv ers ity Press.
Khan, H. 1987. C h a r les Correa. Singap ore: C o n c ep t M e d ia Pte Ltd ( c o m p l e t e l y
revised edition o f the first edition 1984).
Khanna, M. 1979. Yantra: T h e Tantric S y m b o l o f C o s m ic Unity. London: T h a m e s
and Hudson.
Khanna, M. 1986. T h e C o ncep t and Liturgy o f the Sricakra b a sed on S iv a n a n d a s
T r i lo g y ( u n p u b l i s h e d d octoral d isse rta tio n , su b m itte d to W o l f s o n C o l l e g e ,
O xford U niversity),
Khanna, M. 1987. Yantra. In: T h e E n c y c lo p e d ia o f R eli g io n . Editor in Chief: M.
Eliade. V o l u m e 15. N e w York: M a cM illa n P ublishin g C om pany: 5 0 0 - 5 0 2 .
Khurrana, P. 2 0 0 0 . H ea lin g through Yantra. N e w Delhi: Crest Publishers.
K o lv e r, B. 1992a. S o m e E x a m p le s o f S yncretism in N ep al. In: A s p e c t s o f N e p a l e s e
Traditions. Edited by B. Kolver. Stuttgart: Franz Steiner Verlag: 2 0 9 - 2 2 2 .
K o lv e r , B . 1 9 9 2 b . D a s S y m b o l cva n i Stu dien zur I n d o lo g i e und Iranistik 16:
101-107.
K olver, B. 1996. Constructin g P agodas according to Traditional N e p a l e s e D ra w in g s.
Berlin: A k a d e m ie Verlag.
K o lv e r , U. and I. S h rcsth a ca ry a 1994. A D ic t io n a r y o f C o n tem p o ra ry N e w a r i.
N ew a ri - English. Bonn: V G H W issen sch aftsverlag.
K o ss a k , S .M . and J.C. S in g er 1998. Sacred V is io n s : Early P ain tin gs from Central
Tibet. N e w York: T h e Metropolitan M u seum o f Art.
Kramrisch, S. 1946. T h e Hin du T e m p le . 2 v o lu m e s . Calcutta: U n iv ers ity o f Calcutta
Press.
Kramrisch, S. 1981. T h e P resen ce o f Siva. Princeton: Princeton U niversity Press.
Kreijger, H.E, 1999. K athm andu V a lle y Painting: T h e Jucker C o lle ctio n . London:
Serindia Publications.
K u la ic h e v , A .P . 1984. Sriyantra and its M athem atical Properties. Indian journal o f
History o f S c ie n c e 19: 2 7 9 - 2 9 2 .
K u la ic h e v , A .P . and D .M . R a m c n d ic 1989. $riyantra T h e A n c ic n t Instrument to
Control the P s y c h o p h y s i o l o g i c a l State o f Man. Indian Journal o f H is to ry o f
S c ie n c e 24: 1 3 7 - 1 4 9 .
BI BLI OGR AP HY A N D A B B R E V I A T I O N S
28 3
Kulkarni, R.P. 1979. V astupadam andala. Journal o f the Oriental Institute (B a ro d a )

2 8:1 0 7 -1 3 8 .
"
L Hcrnault, F, 1987. D arasuram: cp ig r a p h ica l stud y, e t u d e a r c h i t e c t u r a l , etu d e
ic o n o g r a p h iq u e . W ith the co lla b o r a tio n o f P.R . S r in iv a s a n a n d J. D u m a r g a y .
T o m e 1 - t e x t c . Paris: P ublications dc l E c o le Frangaise d E x tre m e Orient.
L cid y , D.P. and R .A .F . Thurm an 1997. Mandala: T h e A rc h ite ctu r e o f E n l i g h t e n
ment, N e w York: A sia S o cic ty G alleries and Tibet H o u se.
L e s s in g , F.D. and A. W a y m a n 1978. Introduction to the B u dd h ist Tantric S y s te m s .
Transla ted from M K H A S G R U B R J E S R g y u d s d e s p y ih i rnam par g z a g pa
rgyas p a rb rjo d W ith Original T e x t and A nn otation . D elh i: M otilal B a n a rs id a ss
(second edition).
L e v y , R.I. 1990. M e s o c o s m : H in d uism and the O rganization o f a Traditional N e w a r
City in N ep a l. B erkeley: T h e U niversity o f California Press.
Libcrsat, M. 1988. Architects and Architecture in Kerala. In M eister 1988: 9 6 - 1 0 2 .
Liebert, G. 1976, I c o n o g r a p h ic D ic t i o n a r y o f the In dian R e li g io n s : H in d u i s m B u dd h ism - ja inism . Leiden: E.J. Brill.
Licnhard, S. 1997. Martial Art and Poetics: S o m e M ore O b ser v a tio n s on Citrakavya.
In: Lex ct Litterac: S tu d ie s in H o n o u r o f P r o f e s s o r O sc a r B o tto . E d ited b y S.
Lienhard and I. P iovan o. Torino: Edizioni d e l l Orso: 3 4 3 - 3 5 9 .
M a c d o n a ld , A .W . and A. Vergati Stahl 1979. N e w a r Art: N e p a l e s e Art during the
Malla Period. Warminster: A ris & Phillips Ltd.
M atsubara, M. 1994. Pancaratra S a m h it a s & Early V a is n a v a T h e o l o g y . W ith a
Translation and Critical N o t e s from Chapters on T h e o l o g y in the A h irb u d h n y a
Samhita. Delhi: Motilal Banarsidass.
M a y r h o fe r, M. 1 9 8 6 - 2 0 0 1 . E t y m o l o g i s c h e s W o r te r b u c h d es A lt in d o a r i s c h e n . 3
v o lu m e s . H eidelberg: U niversitatsverlag C. Winter.
M c G c e , M. 1987. F ea stin g and fasting: T h e vrata tradition and its s i g n i f ic a n c e for
H indu w o m e n (u n p u b lish ed d octoral d issertation , s u b m itte d to the F a c u lt y o f
Harvard D iv in ity S c h o o l , Harvard U niversity).
M c is e z a h l , R .O . 1 9 7 4 . S m aS a n a v id h i d e s Luyi: T e x tk ritik n a ch der t ib e t i s c h e n
V e r s i o n d e s K o m m e n t a r s L u y ip a d a b h i s a m a y a v r tt i S a m b a r o d a y a n a m a v o n
Tathiigatavajra. Zentralasialische Studien 8: 9 178.
M eister, M .W . 1978. C onstruction and C o n cep tio n : M a n d a p ik a S h r in e s o f Central
India. East and W est. N e w S eries 26: 4 0 9 - 4 1 8 .
M eister, M .W . 1979. M an d ala and Practice in N ag a ra A rc hitecture in N o rth India.
Journal o f the A m e rica n Oriental S o c i e t y 99: 2 0 4 - 2 1 9 .
M e ister , M .W . 1 981a. Darra and the Early G upta T r a d itio n . In: C h h a v i- 2 . R ai
K rishnadasa F elicita tio n V o l u m e . Edited b y A . K rishna. Banaras: Bharat Kala
Bhavan: 1 9 2 - 2 0 5 .
M eister, M .W . 1981b. M undeSvari: A m b i g u i t y and C ertainty in the A n a l y s i s o f a
T e m p le Plan. In: KaladarSana: A m erica n S tu d ie s in the Art o f India. E d ited b y
J.G. W illia m s. N e w D elhi: O xford & IBH P ub lis h in g Co.: 7 7 - 8 9 .
Meister, M .W . 1982. A n a l y s is o f T e m p l e Plans: Indor. Artib us A s i a e 43: 3 0 2 - 3 2 0 .
M eister, M .W . 1983a. G e o m e tr y and M easu re in Indian T e m p l e Plans: R ecta n g u la r
T e m p le s . Artib us A s ia e 44: 2 6 6 - 2 9 6 .
M eister, M .W . 1983b. T h e UdayeSvara T e m p l e Plan. In: Srinidhih: P e r s p e c tiv e s in
Indian A r c h a e o lo g y , Art and Culture. Edited by K .V , R a m a n et al. Madras: N e w
Era Publications: 8 5 - 9 3 .
M eister, M .W . 1984. S i v a s Forts in Central India: T e m p le s in D a k s i n a K o s a l a and
T h eir D aem onic Plans. In: D is c o u r s e s on Siv a. P r o c e e d in g s o f a S y m p o s i u m on
284
B IBLI OGRAPHY A N D ABBREVI ATI ONS
the N a tu re o f R e l i g i o u s Im a g ery . E d ite d b y M .W . M e ister , P h ila d e lp h ia :

U niversity o f P en nsy lv a nia Press: 1 1 9 -1 4 2 .
M eister, M .W . 1985. M ea su rem en t and Proportion in Hindu T e m p le A rchitecture.
Interdisciplinary S c ie n c e R e v i e w s 10: 2 4 8 - 2 5 8 .
Meister, M .W . 1986. On the D e v e lo p m e n t o f a M o r p h o lo g y for a S y m b o lic A r c h i
tecture: India. R es A nthro po lo g y and A esth etics 12: 3 3 - 5 0 .
M eister, M .W . (ed.) 1988. M a k in g T h in g s in South Asia: T h e R o le o f Artist and
Craftsman. Philadelphia; Departm ent o f South A sia R egional Studies.
Meister, M .W . 1989. R eading M o n u m e n ts and S eein g Texts. In: Shastric Traditions
in Indian Arts. Edited by A .L . D allapiccola. 2 v olu m es. Stuttgart: Steiner Verlag:
v o lu m e 1: 1 6 7 - 1 7 3 ; v o lu m e 2: 9 4 - 1 0 8 & plates 8 1 - 8 8 .
M eister, M .W . 1990. D e - and R e-co n stru cting the Indian T e m p le . Art Journal 49:
395-400.
Meister, M .W . 1997. R ea ss e s sin g the Text. In: C ontemporary Architecture and City
Form: T h e South A sia n Paradigm. Edited by F. A m e e n . B o m b a y : Marg P u b li
cations: 8 8 - 1 0 0 .
M e is ter , M .W . 1 9 9 9 . R e v i e w o f L .S . Thakur: T h e A rchitectura l H e r it a g e o f
H im a c h a l Pradesh: O rigin and D e v e l o p m e n t o f T e m p le S ty le s . T h e In dian
Historical R e v ie w 26: 1 8 7 - 1 9 1 .
M eister, M . W ./ M . A . D h a k y /K . D e v a ( c d s .) 1988. Encylopffidia o f Indian T e m p le
A rc h ite ctu r e. V o l u m e 2/1: North India: F o u n d a tio n s o f North Indian S ty le .
Princeton: Princeton University Press.
M e n o n , C .P .S . 1932. Early A st r o n o m y and C o s m o lo g y : A R e c o n stru ctio n o f the
Earliest C o s m ic S ystem . London: G e o r g e A llen & U n w in Ltd.
M ichael, T. 1986. Le Sri-cakra dans la Saundarya-Lahari. In: Mantras et d iagram m es
rituels dans l h in d o u ism e . T a b le R o n d e , Paris 2 1 - 2 2 juin 1984. < E d ited b y A.
P a d o u x .> Paris: E d it io n s du Centre n a tio n a l de la rc ch e r c h e s c ie n tifiq u c :
127-137.
M ichaels, A. 1996. G o d d e ss o f the Secret: GuhycSvari in N ep a l and Her Festival. In:
W ild G o d d e s s e s in India and N ep al. P roceedings o f an International S y m p o s iu m
B erne and Zurich, N o v e m b e r 1994. Edited by A, M ich a cls / C. V o g e ls a n g e r / A.
W ilk e. Bern: Peter Lang: 3 0 3 - 3 4 2 .
Misra, J. 1959. A Yantra o f the Tantrikas. Journal o f the Bihar and Orissa R esearch
S o c ie ty 45: 4 8 2 4 8 8 .
Mitra, H. 1958. Ganapati. Visva-Bharati A nn als 8: 1120.
M o o k er jee, A . 1971. Tantra Asana: A W a y to S elf-R e a liz a tio n . B a sel / Paris / N e w
York: R a v i Kumar.
M ori, M. 1 9 9 7 . T h e V ajravali o f A bhayakaragupta: A Critical Stu dy, Sanskrit
Edition o f S e le c t e d Chapters and C o m p le te Tibetan V er sio n (un pu blish ed d o c
toral d isserta tio n , su bm itted to the S c h o o l o f Oriental and A frica n S tu d ie s ,
U niversity o f London).
M osteller, J.F. 1988. T e x ts and Craftsmen at: Work. In Meister 1988: 2 4 - 3 3 .
M o s te lle r, J.F. 1991. T h e M e a su re o f Form: A N e w A pp ro a ch for the S tu d y o f
Indian Sculpture. N e w Delhi: A b h in a v Publications.
M u s, P. ( 1 9 3 5 ) 1998. Barabudur: S k e tc h o f a H is to r y o f B u d d h is m b a s e d on
A r c h a e o lo g ic a l C ritic ism o f the T e x ts . Translated from the French b y A .W .
M acdonald . N e w D elhi: Sterling Publishers Private Lim ited.
[Translation of: M u s, P.: Barabudur: E s q u is s c d une h istoirc du b ou d d h ism e
f o n d c e sur la critique a rc h e o lo g iq u e des textcs. Hanoi: Impr. d E xtrem c-O rient,
1935. 2 v o lu m e s ( p u b lish e d in o n e v o lu m e , N e w York; A m o Press, 1978 and
Paris: Arm a Artis, 1990)].
BI BL IO G RA PH Y AN D A BB R E V I A T I O N S
285
Nath, R. 1 9 7 5 - 1 9 7 6 . D c p ic tio n o f a Tantric S y m b o l in M ughal Architecture. Journal

o f the Indian S o cie ty o f Oriental Art. N e w S eries 7: 7 3 - 8 5 .
Padtnaja, V.T. 1985. Yantra W orship in the Sakta Centres o f Gujarat. Journal o f the
Oriental Institute (Baroda) 34: 1 7 3 - 1 7 9 .
Padoux, A. 1974, Introduction: Mantra et yantra-mandala. In: G. T u cc i: T h e o r ie et
Pratique du Mandala. Traduit de Iitalien par H.J. M a x w e ll. Paris: Fayard: 1 - 8 .
Padoux, A. 1977. Un tcrm c techn iqu e du mantrasastra : vidarbha. Journal A sia tiq u e
265:345-349.
P a do u x , A. 19 8 6 (ed.). Mantras et d ia g r a m m e s ritu els dans l h in d o u i s m e . T a b le
Ronde, Paris 2 1 - 2 2 juin 1984. Paris: E ditions du Centre national de la rech erch e
scientifique.
Padoux, A. 19861992. Mantric Practices and the Nature o f Mantric U tterances. The
Journal o f Oriental Research 5 6 - 6 2 : 6 5 - 7 6 .
Padou x, A. 1987a. Cakras. In: T h e E n c y c lo p e d ia o f R e lig io n . Editor in C hief: M.
Eliadc. V o lu m e 3. N e w York: M a cM illa n P ub lishin g C om p any: 4 - 5 .
Padoux, A. 1987b. Tantrism. In: T h e E n c y c lo p e d ia o f R e lig io n . Editor in Chief: M.
Eliadc. V o lu m e 14, N e w York: M a c M illa n P u b lis h in g C o m p a n y : 2 7 2 - 2 7 4 .
Padou x, A. 1 9 9 0 a . V a c. T h e C o n c ep tio n o f the W ord in S e le c t e d H in d u Tantras.
Translated b y J. G onthier. A lb a n y , N e w York: State U n iv e r s ity o f N e w Y o rk
Press.
Padou x, A. 1990b . T h e B o d y in Tantric Ritual; T h e C a s e o f the Mudras. In; T h e
Sanskrit Tradition and Tantrism. P anels o f the V l l t h W o rld Sanskrit C o n fer en ce ,
Kern Institute, Leiden: A u g u s t 2 3 - 2 9 , 1987. Edited b y T. G oudriaan. Leid en ;
E.J. Brill: 6 6 - 7 5 .
Padou x, A. 1994. L e cocur dc la Y o g in l. Y o g in lh r d a y a , a v e e le co m m e n ta ir e D ip ik a
d Amrtananda. Paris: D e Boccard.
Pal, P. 19741978. T h e Arts o f N ep a l. 2 parts. Leiden: E.J. Brill.
Pal, P, 1992. H im a la y a n Mandalas in the Z im m er m a n C o llectio n . O rientations 2 3 /2 :
3 6 45.
Pal, P. 1993. Indian Pain ting; A C a ta lo g u e o f the L o s A n g e l e s C o u n ty M u s e u m o f
Art C ollection , L o s A n g e le s : L o s A n g e l e s C o un ty M u se u m o f Art.
Pal, P. 1997. A C o lle c t in g O d y s s e y : Indian, H im a la y a n , and S o u th e a st A s ia n Art
fro m the J a m e s and M a r ily n n A l s d o r f C o lle c t io n , N e w Y o rk ; T h a m e s and
H u dson Ltd.
P anofsk y, E. 1955. T h e History o f the T h eo r y o f H u m a n Proportions as a R e f le c t io n
o f the History o f S ty le s. In: M e a n in g in the V isu a l Arts: Papers in and on Art
History, Garden City, N e w York: D o u b le d a y & C o m p a n y , Inc.: 5 5 - 1 0 7 .
Pichard, P. 1995. Tanjavur BrhadlSvara: A n Architectural Study, N e w D elh i: Indira
Gandhi National Centre for the Arts.
Pott, P.H, ( 1 9 4 6 ) 1966. Y o g a and Yantra: T h eir Interrelation and their S ig n i fic a n c e
for Indian A r c h a e o lo g y . Translated from the D utch b y R. N e e d h a m . T h e Hague:
Martinus N ijhoff.
[Translation of: Pott, P.H,: Y o g a en yantra in h unne b e t e e k e n is v o o r de In dis ch e
archeologie. Leid en: E.J. Brill, 1946.]
Pramar, V .S . 1 9 8 5 . S o m e E v i d e n c e on the W o o d e n O r ig in s o f the V a s t u p u r u sa
mandala. Artib us A sia e 46: 3 0 5 - 3 1 1 .
Pranavananda, S w a m i < 1 9 7 7 > . A Treatise on Sricakra. Y e n u g u la Mahal: Sri S w a m i
Pranavananda Trust: no year, a p pro x im a tely 1977.
Pratyagatmananda Sarasw ati, S w a m i. T h e Yantram. Calcutta: B e n l y Press' n o y ear
1960?
286
BI BLI OGRAP HY A N D ABBREVI ATI ONS
Pratyagatm an and a Sarasw ati, S w a m i and J. W o o d r o f f e 1963. Sadhana for S e l f

R ealization (Mantras, Yantras & Tantras), Madras: Gancsh & Co.
Pruscha, C. 1994. D a s O bscrvatoriu m dcs Jai Singh: D a s Yantra v o n Jaipur, In: D a s
B a u w e r k und die Stadt. T h e B u ild in g and the T o w n . A ufsiitzc fur / E s s a y s for
E.F. Sekler. H c ra u sg cg e b e n von W. B o h m . Wien: B ohlau Verlag: 2 6 5 - 2 6 8 .
R abe, M .D . 2 0 0 0 . Secret Yantras and Erotic D is p la y for Hindu T e m p le s . In: Tantra
in Practice. Edited by D .G . W hite. Princcton, N e w Jersey: Princeton U n iv ers ity
Press: 4 3 4 446,
R a g h a v a n , V . 1956. Yantras or M e ch a n ica l C o ntriv a nces in A ncicn t India. B a n g a
lore: T h e Indian Institute o f Culture (sec o n d edition, revised and enlarged).
R ag ha v a n , V. ( f ) 1998. Sanskrit R am ayanas Other Than V a l m i k is - T h e Adbhuta,
A d h y a tm a , and A n a n d a R a m a y a n a s . C hennai: Dr. V . R a g h a v a n C en tr e for
P erform ing Arts.
R a m b elli, F. 1991. R e-in scr ib in g Mandala : S c m io tic O perations on a W ord and its
Object. S tudies in Central and East A sia n R e lig io n s 4: 1 - 1 2 4 .
R a ngachari, D .B .K , 1930. T h e Sri V a is h n a v a Brahm an s. B u lletin o f the Madras
G o v e rn m e n t M u seum . N e w Series. General S ection 2, 2. Delh i: Gian P ub lishin g
H o u s e , 1 9 8 6 (first reprint).
R ao, C .S. 1998. Sriyantra - A Study o f Spherical and Pla ne Forms. Indian Journal o f
History o f S c ie n c e 33: 2 0 3 - 2 2 7 .
Rao, D .M . 1995. Vastuilpaastra. Bangalore: S .B .S . Publishers.
Rao, S.K .R. 1988. T h e Yantras. D elhi: Sri Satguru Publications.
R a o , S .K .R . 1 9 8 8 - 1 9 9 0 . M a n d a la s in T e m p l e W o rs hip . 2 v o l u m e s . B a n g a lo r e :
Kalpatharu Research A ca d em y ,
R a o , S .K .R . 1 9 8 8 - 1 9 9 2 . P ratim a-K osha. E n c y c lo p a e d ia o f Indian Ic o nograph y. 6
v o lu m e s . B angalore: Kalpatharu Research A ca d cm y .
R a o , S.K .R . 1990. T h e Tantrik Practices in S ri-V id y a (with ri Sarada-ChatuatT).
Bangalore: Kalpatharu Research A ca d em y .
Rao, T .A .G . 1 9 1 4 - 1 9 1 6 . E le m e n ts o f Hin du Iconography, 2 v o lu m e s . Madras: Law
Printing H o u se.
R a stelli, M. 1999. P h i l o s o p h i s c h - t h e o l o g is c h e G ru n d a n sch a u u n g en dcr J ayak hyasamhita. M il cincr D a rstellun g des taglichen Rituals. Wien: V erlag dcr Ostcrreic h is c h e n A k a d e m ie der W issen sch aften .
R a s t e lli, M. 2 0 0 0 a , D i e f iin f Z e iten (panca kala s) in den a llc s ten PancaratraSamhitas. W iener Zeitschrift fur d ie Kunde Siidasiens 44: 1 0 1 - 1 3 4 .
R a s t e lli , M. 2 0 0 0 b . T h e R e l i g i o u s P ra c tice o f the S a d h a k a A c c o r d i n g to the
Jayakhyasam hita. Indo-Iranian Journal 43: 3 1 9 - 3 9 5 .
R astelli, M. 2 0 0 2 . T h e a sana A cc o r d in g to the ParameSvarasamhita or A M e th o d o f
W riting a Sam hita. In: S tu d ie s in H in d uism III: Pancaratra and V iS istadvaitavedanta. Edited b y G. Q b er h a m m e r and M. R astelli. W ien: V erlag dcr O stcrreichisch en A k a d e m ie der W issen sch aften : 9 - 3 2 .
R astelli, M. 2 0 0 3 . D er T e m p el als M ythisierung der T ranszendenz. In: M y th isie ru n g
der T r a n s z e n d e n z als E n t w u r f ihrer Erfahrung. A r b e it s d o k u m e n t a t io n e i n c s
S y m p o s i o n s . H e r a u s g e g e b e n v on G. O b er h a m m e r und M. S ch m iick er, W ien:
V erla g der O sterreichischen A k a d e m ie dcr W issenschaften: 3 1 3 - 3 4 8 .
R a y , R .A . 1973. M andala S y m b o l i s m in Tantric B u d d h is m (u n p u b lish ed doctoral
d isse rta tio n , s u b m itte d to the F a c u lty o f the D iv i n i t y S c h o o l , U n iv e r s ity o f
C h ica g o ).
R eitz, F. 1998. PancayaJa/ja-K omplexe in N ordin dien. Entstehung, E n tw i c k lu n g und
reg io n a le B e so n d e rh eitcn einer in disch cn Architckturform (unpu blish ed doctoral
B IBLI OGRAPHY A N D A B BR E V I A T I O N S
287
d is se rta tio n , s u b m it te d to the In stitut fur I n d is c h e P h i l o l o g i e und K u n s tgescliiclilc, Frcie Universitat, Berlin).
R en o u , L. 1998. T h e V c d ic H o u s e . Translated and ed ited b y M .W . M eister. R es
A nthro po lo g y and A cs th ctics 34: 1 4 1 - 1 6 1 .
R c n o u , L. and J. F illio z a t 1 9 4 7 - 1 9 5 3 . L In d e c l a s s iq u e : M a n u e l d e s e t u d e s
indiennes. 2 v o lu m e s . Paris: Payot.
Rhie, M .M . and R .A .F . Thurman 1999: W o rld s o f Transform ation: T ib eta n Art o f
W is d o m and C o m p a ssio n . N e w York: Tibet House.
R h o d es, N .G ./K , G a b r is c h /f C .V .P . d ella R o c c h e tt a 1989. T h e C o in a g e o f N e p a l
from the earliest time until 1911. London: R oyal N u m is m a t ic S o cie ty .
Rocher, R. (ed.) 1978. India and In dology. S e le c te d A rticles b y W. N o rm a n B ro w n.
Delhi: Motilal Banarsidass.
R o ? u, A. 1986a. Mantra et yantra dans la in e d e c in e ct 1a lc h im ie in d ien n e s. Journal
Asiatique 274: 2 0 3 - 2 6 8 .
R o ? u , A. 1986b, M a n tra et ya n tr a dans la m e d e c in e et P a l c h i m i e in d ie n n e s. In:
Mantras ct d ia g ra m m c s ritucls dans 1h ind ouism e. T a b le R o n d e, Paris 2 1 - 2 2 juin
1984. < Edited b y A. P a d o u x .> Paris: E d itio n s du Centre national de la recherch e
scicntifiquc: 1 1 7 - 1 2 1 .
Roû, A, 1989. L es carrcs m a g iq u e s in dien s et l h istoire des id e e s en A s ie . Z eitschrift der D cu ts ch en M o r g en la n d isch en G e sellsch a ft 139: 1 2 0 - 1 5 8 .
R o y , A .K . 1977, T h e D rea m and the Plan. M arg 4: 2 5 - 2 6 .
R o y , M. 1984, T h e C o n c e p t o f Y a n tra in the Sam arangana-Sutradhara o f Bh oja.
Indian Journal o f History o f S c ie n c c 19: 1 1 8 - 1 2 4 .
Sabarathinam, S.P. 1995. A g a m i c Treatm ent o f Mahabhutas in R elation to M an d alas
and Arts. In: Prakrti: T h e Integral V is io n . V o l u m e 3: T h e A g a m i c Tradition and
the Arts, Edited b y B . Baum er. N e w D elh i: D .K . Printworld (P) Ltd.: 4 7 - 6 6 .
S a n d er so n , A. 1986, M andala and A g a m i c Id entity in the Trika o f K a sh m ir. In:
Mantras et d ia g ra m m e s ritucls dans l h ind ouism e. T a b le R o n d e , Paris 2 1 22 juin
1984. < E d itcd b y A. P a d o u x .> Paris: E ditions du C entre national de la rech erche
scientifique: 1 6 9 - 2 0 7 .
S a nderson, A. 1988. S a iv is m and the Tantric Traditions. In: T h e W o r l d s R e lig io n s .
T h e R e li g io n s o f A s ia . Edited b y F. Hardy. L o n d o n : R o u t le d g e , 1 9 8 8 , 1 9 9 0
(reprint): 1 2 8 - 1 7 2 .
S a n d er so n , A . 1 9 9 0 . T h e V is u a liz a t io n o f th e D e i t i e s o f the Trika. In: L im a g e
d iv in e . C u lte et m e d ita t io n dans l h in d o u i s m e . E d it e d b y A . P a d o u x . Paris:
E d ition s du Centre national de la recherch e scientifique: 3 1 - 8 8 .
Sanderson, A. 2 0 0 1 . History T h ro u g h Textual C riticism in the S tu dy o f S a iv ism , the
Paiicaratra and the Bu dd h ist Y ogim tantras. In: L es S o u rces et le T e m p s . S o u r ces
and T im e. A C o llo q u iu m , P o ndicherry 11-13 January 1997. Edited b y F. Grimal.
Pondicherry: Institut Frangais d e Pondichery: 1 - 4 7 .
Schneider, U. 1988. Tantra - Endpunkt ein e s strukturierten A b la u fs ? S a e c u lu m 39:
96-104.
Shankaranarayanan, S. 1970. Sri Chakra. Pondicherry: D ip ti Pub lications.
Sharf, R .H . 2 0 0 1 . V is u a liz a t io n and M a n d a la in S h in g o n B u d d h is m . In: L i v i n g
Im ages: Jap anese B u d d h is t Ic o n s in C o n tex t. E d ited b y R .H . S h a r f and E.H .
Sharf. Stanford, California: Stanford U n iv ers ity Press: 1 5 1 - 1 9 7 , 2 3 9 - 2 5 1 .
Sharma, B.R . 1994. A B r i e f N o t e on Yantras and M an d alas. In: Pand it N .R . Bhatt
F elicita tio n V o l u m e . E d ited by P.-S . F illio z a t /S .P . N a r a n g /C .P . Bhatta. D e lh i:
Motilal Banarsidass: 4 1 9 - 4 2 4 .
Shubhakaran, K.T. 1992. M y s tic a l F o rm ula e. Part II, Y antras. N e w D elh i: S a g a r
Publications.
2 88
B IB L IO G R A P H Y A N D A B B R E V IA T IO N S
S h u k la , D . N . 1967. R o y a l Arts: Yantras and Citras. L u ck n o w : V a s t u v a n m a y a Prakasna-Sala 1967 (Student Edition),

Sin g h , S.K . 2 0 0 0 (cd.). Mandala Festival. Kathmandu: N epal Lithographing Co. (P)
Ltd.
Sinha, A.J. 2 0 0 0 . Im agining Architccts: Creativity in the R e li g io u s M o n u m e n ts o f
India. Newark: University o f D ela w a re Press.
Sircar, D .C . 1968. S tudies in Indian Coins. Delhi: Motilal Banarsidass.
S m ith , B.K . 1987. E x o r c is in g the Transcendent: S trategics for D e f in in g H in d u ism
and R elig io n. History o f R elig io n s 27: 3 2 - 5 5 .
S m ith , V .A . 1 9 7 2 . C o in s o f A n c ic n t India: C a ta lo g u e o f the C o in s in the Indian
M u se u m Calcutta. V o lu m e I. Delhi: Indological B ook H o u se (reprint).
S n e llg r o v e , D .L . 1959. T h e Hevajra Tantra: A Critical Study. 2 parts. London: O x
ford U niversity Press.
Srin ivasan, P.R. 1987. S e e L Hernault, F. 1987.
Staa), F. 1984. A gni. T h e V c d ic Ritual o f the Fire Altar. 2 v o lu m e s. D elh i: M otilal
Banarsidass.
Stadtner, D. 1998 (ed.). A India, portico do norte. < ca ta lo g u c o f an cx h ib itio n > 3 de
outu bro 1998 - 10 d c x a n cir o 1999. S a n tia g o d e C o m p o s te la : A u d ito rio de
G alicia.
Stoddard, H. 1999. D y n a m ic Structures in B u dd h ist Mandalas. Apradaksina [s/c/J
and M y s tic Heat in the M other Tantra S ectio n o f the A nuttarayoga Tantras.
A rtib us A s ia e 58: 169213.
S y c d , R. 1990. D i e Flora A ltin d ie n s in Litcratur und Kunst (un p u b lish ed doctoral
d isserta tio n , s u bm itted to the P h i lo s o p h is c h c Fakultat, L u d w ig - M a x im i lia n s Universitat, Munchen).
T ak ash im a, J. 1992. D iksa in the Tantraloka. T h e M em o irs o f the Institute o f O rien
tal Culture (The U niv ers ity o f T o k y o ) 1 19: 4 5 - 8 4 .
T a y lo r, K. 2 0 0 1 . Sir John W o o d r o f fe , Tantra and B e n g a l: A n Indian S o u l in a
European B o d y ? R ich m ond , Surrey: Curzon Press.
Thakur, L.S. 1990, A pp lic a tio n o f Vastupurasamandala [ s /c / ] in the Indian T e m p le
A rchitecture: A n A n a lysis o f the Nagara Tem ple Plans o f H imachal Pradesh.
A rtib us A s ia e 50: 2 6 3 284.
Thakur, L.S. 1996. T h e A rchitectural Heritage o f Him achal Pradesh: O rigin and
D e v e lo p m e n t o f T e m p le Styles. N e w D elh i: Munshiram Manoharlal Publishers.
Thurston, E. 1912, O m e n s and S uperstitions o f Southern India. London: T. F ish er
U n w in .
T illo tso n , G .H .R . 1987. The Rajput Palaces: T h e D e v e lo p m e n t o f an Architectural
Style, 1 4 5 0 - 1 7 5 0 . N e w Haven/L ondon: Y a le University Press.
T o f f m , G. 1999. Spatial C ategories o f the N e w a r s o f Kathm andu V alley: T h e Inside
- O u ts id e O p p o sitio n . In: H im a la y a n Space: Cultural H o r iz o n s and P ractices.
Edited by B. B i c k c l and M, Gacnzlc. Zurich: V o lk crk u nd em useum : 3 3 - 7 0 .
T o g a n o o , S.M . 1971. T h e S y m b o l- S y s t c m o f S h in g o n B u dd h ism (un pu blish ed d o c
toral d isse rta tio n , su bm itted to the F a c u lty o f C la re m o n t G raduate S c h o o l,
C larem ont University).
T o r z s o k , J. 1999a. T h e D o ctrin e o f M a g ic F e m a le S p irits: A Critical Edition o f
S e le c t e d C hap ters o f the Siddhayogcgvarimata(tantra) with A n n o ta t e d T r a n s
lation and A n a l y s i s (u n p u b lis h e d d octoral d isse rta tio n , su bm itted to O x fo rd
U niversity).
T o rzso k , J. 1 9 9 9 b . W h y throw a flo w er ? S o m e p ro b lem s co n ce rn in g the S a m a y a
initiation (u n pu blish ed paper presented at the Tantric sem in ar held in All S o u ls
C o lle g e , Oxford, Hilary term 1999).
BI BLI OG RAP HY A N D A B B R E V I A T I O N S
289
T orzsok , J. forthcom ing. T h e S idd h a y o g e sv a rim a ta . A critical edition and annotated

translation.
Trib e, A .H .F . 1994. T h e N a m e s o f W is d o m : A C ritical E d it io n and A n n o ta t e d
Translation o f Chapters 1 - 5 o f V ila s a v a j r a s C o m m en ta ry on the N a m a s a m g lt i,
With Introduction and T e x tu a l N o t e s ( u n p u b lish e d d octoral d isse rta tio n , su b
mitted to the Faculty o f Oriental Studies, Oxford U niversity).
Tsuda, S. 1990. T h e C ult o f S m a sa n a , the R e a litie s o f Tantrism . In: T h e Sanskrit
Tradition and Tantrism. P a n els o f the V llt h W orld Sanskrit C o n f e r e n c e , Kern
Institute, Leiden: A u g u s t 2 3 - 2 9 , 1987. Ed ited b y T. G oudriaan. L e id en : E.J.
Brill: 9 6 - 1 0 8 .
Tucci, G. 1949. Tibetan Painted Scrolls. 2 v o lu m e s. R om a: La Libreria D e l l o Stato.
T u c c i, G. ( 1 9 4 9 ) 1961. T h e T h e o r y and P ra c tice o f th e M a n d a la . W ith special
rcfercnce to the m o d e m p sy c h o lo g y o f the subconscious. Translated from the

Italian b y A . H . Brodrick. London: Rider & C om pany.
[Translation of: T u cc i, G.: Teoria e pratica del mandala, co n particolare riguardo
alia moderna p sic o lo g ia dc profondo. R om a: A strolab io, 1949.]
Untracht, O. 1997. Traditional Jew elry o f India. London: T h a m e s and H u d so n Ltd.
Urban, H.B. 1999. T h e E x trem e Orient: T h e C onstruction o f T a n tr ism as a C a te
g o ry in the Orientalist Imagination, R e lig io n 29: 1 2 3 - 1 4 6 .
V a tsy a y a n , K. 1983. T h e Square and the C ircle o f the Indian Arts. N e w D elh i: R oli
B o o k s International.
Verardi, G. 1994. H om a and Other Fire R itu a ls in Gandhara. N a p o li: Istituto u niversitario orientale.
Vergati, A . 1982. A S k e tc h - B o o k o f N e w a r Iconography. Printed as an appendix in:
L o k cs h Chandra: B u dd h ist Ico no g ra ph y in N e p a le s e S k e t c h - B o o k s . N e w D elh i:
Jayyed Press, 1984.
V ergati, A . 1986. Q u e lq u e s rem arques sur l u s a g e du m a nd a la et du yantra dans la
v a l l c e d e K a t h m a n d u , N e p a l . In: M a n tra s et d i a g r a m m e s r itu e ls d a n s
l h in d o u is m e . T a b le R o n d e , Paris 2 1 - 2 2 ju in 1 9 8 4 . < E d ite d b y A . P a d o u x , >
Paris: E ditions du Centre national d e la recherche scientifique: 3 7 - 4 5 .
V o lw a h s e n , A. 1969. L iv in g Architecture: Indian. N e w York: G rosset & D unlap.
V o lw a h s e n , A. 2 0 0 1 . C o s m ic A rchitecture in India: T h e A s t r o n o m ic a l M o n u m e n t s
o f Maharaja Jai Singh II. Munich: Prestel.
W a y m a n, A. 1973. T h e B u d d h is t Tantras: L ig h t on In do -T ib eta n E s o te ric is m . N e w
York: S a m u el W eiser.
W a y m a n, A. 1 9 7 4 . T h e Ritual in Tantric B u d d h is m o f the D i s c i p l e s E ntrance into
the Mandala. Studia M issio n a lia 23: 4 1 - 5 7 .
W a y m a n , A. 1999. T h e Manda and the -la o f the T e rm Mandala. In: T a n tric
Buddhism : C entennial Tribute to Dr. B e n o y t o s h B hattacharyya. E d ited b y N .N .
Bhattacharyya. N e w Delh i: Manohar: 2 3 - 3 0 .
W eg n er , G .M . 1992. In v o ca tio n o f N a sa h d y a h . In: A s p e c t s o f N e p a le s e Traditions.
Edited by B . K olvcr. Stuttgart: Franz S teiner Verlag; 1 2 5 - 1 3 4 .
W h ea tley , P. 1971. T h e P iv o t o f the F our Quarters: a P relim in a ry E n qu iry into the
Origins and Character o f the A n c ie n t C h in e s e City. Edin burgh: Edinburgh U n i
versity Press.
W hite, D .G . 1998. Transform ations in the Art o f L o v e : K a m a ka la Practices in Hin du
Tantric and Kaula Traditions. H isto r y o f R e lig io n s 38: 1 7 2 - 1 9 8 .
W ilk e , A. 2 0 0 3 . M e n ta l Jou rn ey, C o s m i c G e o g r a p h y , and In te rm ed ia r y S p a ce:
Sricakra and Sricakrapuja. In: S in g h , R .B . (ed.): S a c r e d G e o g r a p h y o f H in d u
G o d d e ss e s . Festschrift to D a v id K in sle y (fo rth co m in g ).
290
B IB L IO G R A P H Y A N D A B B R E V IA T IO N S
W illis, M .D . 1997. T e m p le s o f Gopaksctra: A R eg io n a l History o f Architecture and

Sculpture in Central India A D 6 0 0 - 9 0 0 . London: British M useum Press.
W itzel, M. 1992. M eaningful Ritual: V e d ic , M e d iev a l, and Contem porary C o n c ep ts
in the N e p a l e s e A gnihotra Ritual. In: Ritual, State and H istory in South Asia:
E s s a y s in H onour o f J.C. H cesterm an. Edited b y A .W . van den H ock / D .H .A .
K o l f f / M.S. Oort. Leiden: Brill: 7 7 4 - 8 2 7 .
W o o d ro ffe, J. 1914. Principles o f Tantra: The Tantra-Tattva o f Srlyukta Siva Candra
V idyarnava Bhattacarya [s/c/] Mahodaya. 2 v o lu m e s. London: L uzac & Co.
W o o d r o f fe , J. 1956. Introduction to Tantra Sastra. Madras: G a n esh & C o. (third
edition).
Z anen, M. 1986. T h e G o d d e s s V a jr a y o g in i and the k in g d o m o f Sankhu (N e p a l).
Purusartha 10: 1 2 5 - 1 6 6 .
Z im m er , H. ( 1 9 2 6 ) 1984. Artistic Form and Y o g a in the Sacred Im ages o f India.
T ranslated from the G erm an and ed ited by G. C ha p pie and J.B. L a w s o n in
collaboration with J.M. M cK n ig ht. Princeton, N e w Jersey: Princeton U niversity
Press,
[Translation of: Zim m er, H.: Kunstform und Y o g a ini indischcn Kultbild. Berlin:
Frankfurter Verlags-Anstalt, 1926.]
IN D EX
A bhinavagupta, xiii, 10, 155, 189,

191, 192, 193, 194, 1 9 5 , 2 0 5 , 2 2 5 ,
2 2 7 , 2 3 0 , 2 3 1 , 2 7 2 , 273
air-line, 2 1 6
ajM , 5 1 , 52, 2 4 2
akadamacakra, 5 1
akarsanayantra , ix, 3 3 , 35
akathahacakra, 51
alchcm ist, 4 8
alc h c m y , 1 , 4 4
alpana, 1, 2 8 0
altar, xiv , 26, 48, 164, 2 5 2 , 2 5 3 , 277,
2 7 9 , 28 8
amalasaraka , 223
A m b a Mata Yantra, 2
Am b aji, 2
Amrtananda, x iv , 2 4 0 , 2 4 3 , 2 4 6 , 28 5
amulet, 3 3 , 3 5 , 119, 143, 146, 147,
162,
202
a nahata, 51, 52
Anandaramayana (A R ), 74, 7 7 , 78,
271
'
anantavijaya(mandala), 164
anckakajagarbhamandala , 131
angle, 17, 2 9 , 3 9
antyesti, 2 3 0
apex, 4 2
appeasem en t, 3 5 , 183, 2 0 7
architecture, 2, 6, 7, 8, 11, 2 1 , 2 8 , 60,
75, 1 6 9 , 2 5 2 , 2 5 3 , 2 5 5 , 2 5 6 , 2 5 7 ,
2 7 4 , 2 7 6 , 2 7 8 , 2 8 1, 2 8 2 , 2 8 3 , 2 8 4 ,
285,
288, 289, 29 0
areca nut, xv, xvi, 6 1 , 81, 135
ArthaSastra, 13
asana(yantra), 35, 140, 167, 2 8 6
a sra, 17, 3 9
astadalamandala, xi, 83
a stakona, 47
astalingalatapadmavithiprakara, 82
astalirigatobhadra , xi, 83
astamudraramalihgatobhadra, xii, 85
astamudraramatobhadra, xii, 85
astara, 4 7
astavimSatilingatobhadra , xii, 84
astoftarasahasralingatobhadra, 85
astottarasahasraramalihgatobhadra,
85
astottarasatalingatobhadra, xii, 84
astoftaraiataramalingatobhadra, xii,
85
astotiarasataramatobhadra, 85, 86
astra, 127
avarana, 4 0 , 61, 169, 174, 2 0 3
avrti, 40
A y u r -V e d a , 1
Bafna, 6, 8, 15, 2 5 2 , 2 5 5 , 2 5 6 , 2 5 7 ,
258,
259, 260, 261, 262, 263, 277
baliharanacakra, x, 8, 50, 5 7 , 58, 5 9

baliharanamandala, x, 5 0 , 5 8
balimandala, 160
beesiyantra, 2
belt, 4 0
B e n g a l , 3, 2 7 7 , 2 8 8
bhadra, x v , x v i, 25, 73, 74, 7 5 , 76,
77, 78, 80, 81, 83, 84, 164, 2 5 9 ,
26 1 ,2 6 3
bhadrakamandala, 2 4 , 79, 124
bhadranmndala, x, x v i, 4, 8, 2 4 , 2 5 ,
48,
5 7 ,6 8 , 7 3 , 7 4 , 76, 8 2 , 2 7 8
Bhadramartanda ( B M ) , x, x v i, 7 3 , 7 4 ,
7 5 , 7 6 , 11, 7 8 , 7 9 , 8 0 , 8 1 , 8 2 , 83,
84, 2 7 3
bhadravyasa s y s t e m , x iv
bhaga, 2 1 5 , 2 1 6 , 2 1 7 , 2 2 2 , 22 3
Bhairava, 6 4 , 6 5 , 6 6 , 6 8 , 180, 183,
190, 192, 1 9 9 , 2 0 0 , 2 0 2 , 2 0 3 , 2 0 4 ,
2 0 5 ,2 0 6 , 209, 229, 2 3 1 ,2 3 5 ,2 4 1 ,
245,
281
Bhairava-T antras, 3 0
Bhaktamarastotra, 3 7 , 2 7 1 , 2 7 3
Bhaskararaya, 1 7 , 4 0 , 4 2
bhavana, 16, 182
bhugrha, 4 7 , 6 2 , 6 4
bhukrama(sricakra), 31
bhuprastara(srlcakra), 30, 31
bhuprstha(sricakra), 3 0 , 31
292
INDEX
bhupura, 4 7 , 6 2 , 6 4 , 6 6
bhutatfuddhi, 2 0 , 128, 161
bhuvana, 16, 182
Bihar, 1, 2 8 4
bija, ix, 2 9 , 3 6 , 3 7 , 63, 131, 140, 141,
150, 151, 163, 173, 1 8 3 , 2 0 6 , 2 3 5 ,
243
bindu, 39, 4 1 , 4 4 , 4 5 , 2 3 5 , 2 4 0 , 2 4 1 ,
2 4 2 ,2 4 4
birch-bark, ix, 2 9 , 34, 143, 144, 147,
162,2 0 2
blsonyantra, 2
black, x, 2 5 , 2 6 , 6 6 , 76, 79, 80, 123,
141, 163, 167, 171, 1 7 3 , 2 0 6 , 2 0 7 ,
2 0 8 ,2 1 8 ,
223
b lin dfold, 9, 133, 134, 135, 136, 184,
187,
189, 2 2 8
b lo o d , ix, 2 9 , 34, 4 9 , 163, 2 0 5 , 2 0 6 ,
209
body, 1 , 3 , 9 , 10, 1 1 , 2 0 , 2 4 , 2 6 , 3 3 ,
3 4 , 3 5 , 3 7 , 5 0 , 5 1 , 6 1 , 6 2 , 128,
129, 139, 140, 145, 146, 147, 161,
174, 184, 186, 1 9 3 , 2 0 5 , 2 2 7 , 2 3 1 ,
2 3 3 ,2 3 9 ,240, 2 4 2 ,2 4 4 ,2 4 5 ,2 5 1 ,
2 5 3 ,2 5 6 ,
2 8 5 , 28 8
b o w -sh a p e d , 20 8
Brahma, 51, 59, 60, 62, 6 4 , 81, 128,
141,
1 4 8 ,2 1 0 ,2 1 1
brahmasthana, x iv , 2 5 9 , 2 6 0 , 261
Brahm in ical, 199, 2 0 0 , 2 0 3 , 2 1 0
brhadgauritilaka, xi, 83
Brhat-Sam hita, xiv, 11, 2 7 , 75, 2 5 3 ,
' 27 3
Brunner, x iii, 6 , 8 , 9, 10, 12, 13, 15,

19, 20, 2 1 , 25, 2 9 , 30, 32, 33, 39,
4 8 ,4 9 ,
5 0 , 5 1 , 7 5 , 7 6 , 79, 82, 122,
143, 153, 154, 158, 166, 180, 181,
182, 183, 184, 185, 186, 188, 190,
1 9 2 ,2 1 1 ,2 1 2 ,2 1 3 ,2 1 5 ,2 3 9 ,2 4 0 ,
2 7 1 ,2 7 4 ,
278
Buddha, 2 0 9 , 2 1 0 , 2 1 1 , 2 1 3 , 2 1 4
Buddhist, 1 , 5 , 7 , 9, 13, 14, 16, 18,
22, 2 3 , 3 2 , 3 6 , 4 1 , 4 4 , 4 5 , 4 9 , 197,
1 9 8 ,2 0 0 ,2 1 0 ,2 1 3 ,2 7 7 ,2 8 0 ,2 8 1 ,
2 8 3 ,2 8 7 ,
288, 289
buildin g, 2 , 6 , 8 , 1 5 , 3 2 , 4 7 , 121, 191,
1 9 5 ,2 1 0 ,2 5 2 , 2 5 3 ,2 5 4 ,2 5 5 ,2 5 6 ,
2 5 7 ,2 5 8 , 2 6 3 ,2 8 1 ,2 8 6
cakra, ix, x iv , 1, 2, 4 , 5, 6, 7, 8, 10,

11, 15, 16, 17, 18, 1 9 , 2 4 , 2 7 , 3 0 ,
3 1 ,3 4 ,3 5 ,3 9 ,4 0 , 4 2 ,4 3 ,4 5 ,4 9 ,
50, 5 1 - 5 2 , 58, 6 5 , 119, 124, 127,
136, 1 4 5 , 1 5 3 , 1 5 6 , 1 6 1 , 1 6 2 , 163,
180, 181, 182, 1 8 3 , 2 2 7 , 2 3 1 , 2 3 9 ,
2 4 0 ,2 4 1 ,2 4 2 ,2 4 3 , 244, 245, 246,
2 4 7 .2 8 0 .2 8 1 .2 8 4 .2 8 8
cakras in the body, 5 1 - 5 2 ;
m ea n in g s o f cakra, 4 9 - 5 2 ,
163-164, 180-185
S e e also baliharanacakra; Sricakra
cakrabja, 82, 125, 137, 164
cakrabjamandala, x vi, 9, 2 4 , 7 9 , 124,
1 27,131
cakramandala, 132, 135, 136, 141,
142
cakrambujamandala, 181
cakriiravindamandala, xii, 7 5 , 86
cakravatiira, 2 3 9 , 2 4 0
caravastu, 261
cardinal directions, 2 1 , 37, 4 7 , 59, 60,
62,
7 8 ,2 1 5 ,2 6 1
caturasra, 2 1 5
caturdaSalingatobhadra, xii, 86
caturlingatobhadra, xi, 83
catunnudraramaliiigatobhadra, xii,
x v i, 77, 85
caturvimSatilirigatobhadra, xii, 84
catvarimSallihgatobhadra, x ii, 8 6
cayana altar, 2 6
central point, 2, 41, 4 4
centrc, ix, xi, xvi, 2, 5, 14, 17, 2 2 , 24,
2 5 ,3 5 ,3 6 ,3 7 ,3 9 ,4 0 ,4 2 , 4 3 ,4 5 ,
4 8 , 5 1 , 5 9 , 60, 6 1 , 6 2 , 64, 77, 78,
80, 81, 122, 123, 124, 127, 137,
13 8,140 , 1 4 4 ,1 4 5 ,1 5 1 , 153 ,1 5 7 ,
1 5 9 , 1 6 4 , 167, 174, 181, 182, 199,
200,202, 204, 209, 216, 217, 220,
2 2 2 , 2 2 3 , 2 2 6 , 2 3 3 , 2 4 1 , 242, 244,
2 4 5 ,2 4 6 ,2 4 7 , 253, 270, 273, 274,
2 7 5 ,2 7 6 ,2 7 7 ,2 7 8 ,2 8 3 ,2 8 4 ,2 8 5 ,
2 8 6 .2 8 8
Chandralckha, 2
chatra(yantra), 35
chhatar(yantra), 35
ch ronogram , 261
chvasa, 23
circlc, 2, 10, 13, 16, 2 4 , 3 3 , 3 7 , 3 9 ,
4 4 , 4 7 , 4 8 , 4 9 , 50, 5 2 , 5 9 , 6 1 , 6 2 ,
1 1 9 , 1 2 2 , 1 2 4 , 140, 156, 162, 163,
IN D E X
167, 169, 170, 172, 174, 180, 181,
183, 193, 2 0 3 , 2 0 4 , 2 1 0 , 2 1 1 , 2 1 5 ,
2 1 6 .2 1 7 , 2 2 2 ,2 2 7 ,2 4 0 , 2 5 1 ,2 5 9 ,
288
circlc o f f la m e s , 4 9
circum ambulation, 24, 25, 80
cilrabandha, 2
co e rc iv e diagram, 2 9 , 163
coins, 21, 22, 4 4 , 4 5 , 4 6 , 4 7 , 4 8 , 2 8 7
colour, xi, xii, x v , x vi, x v ii, 6, 10, 19,
2 3 , 2 4 , 2 6 , 2 9 , 39, 4 6 , 4 8 , 58, 60,
6 1 , 6 2 , 6 5 , 7 6 , 7 7 , 7 8 , 79, 8 0 , 8 1 ,
82, 85, 86, 120, 122, 123, 124,
157, 166, 167, 168, 170, 171, 172,
173, 174, 175, 182, 183, 187, 189,
1 9 5 ,20 0, 2 0 5 ,2 0 6 , 2 0 7 ,2 1 5 ,2 1 6 ,
2 1 7 .2 1 8 , 220, 221, 223, 232, 242,
24 4 ,2 7 5
colour mandala, 187, 188, 189
coloured pow d er, 1, 14, 1 9 , 2 2 , 24,
6 1 , 80, 156, 157, 159, 161, 170,
172,
175, 181, 185, 1 8 7 , 2 0 8 , 2 1 3 ,
218, 226, 228, 256
conccntric, 14, 2 1 , 2 4 , 37, 3 9 , 4 7 , 59,
6 1 , 6 3 , 124, 160, 195, 1 9 9 , 2 0 0 ,
2 1 3 ,2 1 5 ,
240
co n ch shell, 7 9 , 124, 125, 1 2 7 , 2 1 8
c o n s c io u s n e s s , 4 1 , 196, 2 3 2 , 2 3 3 ,
234,
2 4 1 ,2 4 4 , 2 4 5 ,2 7 9
consecration, 2 6 , 2 9 , 130, 150, 2 3 2
copperplate, 3, 162
corner, 17, 2 5 , 2 6 , 2 9 , 39, 4 8 , 60, 62,
64,
7 9 , 8 1 , 124, 139, 150, 160,
164, 167, 1 6 9 , 2 1 6 , 2 1 8 , 2 2 0 , 2 2 1 ,
2 58,2 6 1
Correa, 2, 2 5 5 , 2 7 8 , 281
corridor, 2 4 , 2 5
c o s m ic s y m b o lis m , 1, 175, 2 1 2
c o s m o g r a m , 261
c o s m o lo g ic a l, 2 6 , 25 5
c o w - d u n g m andala, 156
creation, 2 1 , 2 3 , 3 4 , 4 1 , 52, 140, 145,
198, 2 4 6 , 2 7 7
crecper, 2 5 , 8 0
crem ation grounds, 34, 4 9 , 2 0 2 , 2 0 5 ,
208
crescent m o o n , 2 5 , 80
dala, 2 5 , 39, 124, 167

dalagra, 2 5 , 4 0 , 167, 2 1 6
dalasamdhi, 2 1 6
29 3
B a n ie lo u , 6, 4 0 , 4 1 , 4 2 , 4 6 , 4 7 , 2 7 8
darsanafyanlra), 35
d esig n , 1 , 2 , 3, 4, 7, 8, 10, 14, 15, 16,
17, 1 9 , 2 0 , 2 1 , 2 2 , 2 3 , 2 4 , 2 5 , 2 6 ,
3 0 ,3 1 ,3 5 ,3 6 ,3 9 , 4 0 ,4 5 ,4 7 ,4 8 ,
49,
5 7 , 6 1 , 6 5 , 6 8 , 7 5 , 183, 199,
2 5 5 ,2 5 9
d c s ig n - c o m p le x , 125
d esire-oriented rite, 3 2 , 3 6 , 39, 63,
162
devatayantra , 3 2
D e v i, 52, 6 0 , 6 3 , 6 4 , 155, 166, 2 7 1 ,
278
D e v im a h a tm y a , 6 3 , 64, 166, 271
dharakayantra, 9, 150
dharanayantra, 3 3 , 3 5
dharani, 3 6 , 2 7
dharmodaya, 41
dharmodaya, 4 1 , 2 7 6
dhyanayantra, 3 2
diagram , x, xi, x iv , 3, 4, 11, 15, 16,
2 3 ,2 7 ,2 8 , 3 3 ,4 3 ,5 0 ,5 1 , 5 9 ,6 6 ,
73, 7 6 , 79, 122, 143, 144, 148,
154, 160, 166, 181, 187, 1 8 8 , 2 0 2 ,
205, 206, 207, 213, 225, 226, 230,
232,233, 239, 240, 245, 246, 247,
2 5 0 ,2 5 2 ,2 5 3 ,2 5 4 , 2 5 5 ,2 7 5 ,2 7 8 ,
279
S e e a lso distributive diagram
dikpala, x v , 6 2 , 6 4 , 127, 261
diksa, 1 , 9 , 10, 1 4 , 2 9 , 123, 126, 130,

131, 132, 133, 136, 137, 158, 165,
166, 171, 174, 185, 186, 190, 2 0 1 ,
205,
208, 225, 227, 287
direction, xiii, x v , xv ii, 2 1 , 3 2 , 3 3 , 3 5 ,
4 1 ,4 2 , 4 3 , 4 4 ,4 7 ,4 8 ,5 9 , 6 0 ,6 1 ,
62,
6 3 , 6 4 , 6 7 , 7 5 , 160, 170, 176,
202, 2 0 4 ,2 1 5 ,2 1 6 , 2 5 1 ,2 5 9 , 2 7 9
directional guardian, 3 8 , 6 2 , 63, 64,
65,
67, 68
distrib utive diagram, 19, 5 1 , 161
door, 1 9 , 2 5 , 2 6 , 2 9 , 3 6 , 124, 127,
135, 139, 140, 157, 166, 169, 182,
2 0 1 ,2 1 7 ,2 1 8 ,2 2 0 ,
2 2 1 ,2 2 2 ,2 2 3 ,
2 2 9 ,2 5 4 ,2 5 8
dot, xi, 4 4 , 2 0 3 , 2 4 0 , 241
d o w n w a rd -o r ien ted triangle, 41
d o w n w a rd -p o in tin g triangle, x, 4 1 ,
42, 4 3 , 4 5 , 64
D urga, 6 0 , 63, 6 4 , 6 5
DurgasaptaSat!, 63
2 94
IN D E X
durgasaptasatimahayantra, 63
dvadadalingatobhadra, xi, xvi, 83, 84
dvadasamudraramalingatobhadra, xii,
85
dvara, 2 5 , 4 7 , 124
dvarasobha, 169
earth, 30, 3 1 , 4 4 , 47, 59, 62, 121,
136,
141, 149, 150, 151, 170, 195,
2 1 3 ,2 4 1 ,2 5 7
earth citadel, 47
earth city, 4 7 , 62
earth house, 4 7 , 62
east, x, 1, 38, 4 4 , 59, 62, 64, 120,
127, 167, 2 5 2 , 2 5 3 , 2 5 6 , 2 7 7 , 283,
286
East A sia, 1, 2 8 6
eight crem ation grounds, 49
eight-petalled lotus, x, 2 1 , 2 2 , 44, 46,
4 7 , 6 3 , 6 4 , 6 5 , 144, 151, 157, 159,
1 6 7 ,2 1 5
eight-p oin ted star, 22
ekalingatobhadra, xi, 83
ekamudraramatobhadra, xii, 85
ckavimsottarasatalihgatobhadra, xii,
85'
elem en t, ix, 4, 14, 19, 2 0 , 21, 2 3 , 25,
2 6 ,3 0 , 4 0 ,4 1 ,4 2 ,4 3 ,4 4 , 4 7 ,4 8 ,
5 2 , 5 7 , 6 3 , 122, 128, 139, 140,
151,
161, 164, 166, 169, 171, 173,
185, 186, 197, 2 0 1 , 2 0 3 , 2 1 1 , 2 1 3 ,
2 1 7 , 2 2 0 , 2 2 3 , 2 2 6 , 2 2 9 , 2 3 0 , 24 5
enclosure, 2 4 , 25, 2 6 , 4 0 , 4 8 , 4 9 , 79,
8 0 , 1 2 4 , 1 3 3 , 1 3 9 , 169
en erg y centre (cakra), 51
enm ity, 35
equilateral triangle, 43
eradication, 35
esoteric, 1, 180, 190, 191, 192, 193,
194,
208
e t y m o lo g y , 13, 28, 29
evam , 4 4 , 140, 142, 2 0 1 , 2 4 5 , 2 4 6 ,
282
ex p a n sio n , 10, 189, 2 0 1 , 2 0 3 , 2 3 5 ,
240, 242, 260
fem a le organ, 43
filaments, 25, 61, 123, 124, 1 4 1 ,2 1 1
fire, 14, 2 0 , 2 6 , 4 2 , 4 3 , 4 4 , 4 8 , 5 8 ,
120, 125, 126, 127, 128, 129, 130,
131, 132, 140, 150, 151, 156, 159,
161, 164, 185, 187, 1 9 0 , 2 1 0 , 2 1 1 ,

2 8 8 ,2 8 9
fire triangle, 44
five, x, xvi, 1 9 , 2 2 , 2 3 , 2 6 , 4 2 , 4 6 , 5 1 ,
52, 57, 6 0 , 6 5 , 66, 67, 7 6 , 7 7 , 78,
80, 121, 122, 124, 132, 139, 141,
1 4 5 , 1 5 7 , 1 6 0 , 161, 171, 173, 196,
1 9 7 ,2 0 3 ,2 0 4 ,2 1 3 ,2 1 8 ,2 2 5 ,2 4 5 ,
281
five deities, x, 57, 6 0 , 203
five-pointed star, 4 6
floor design, 1
flower, 9, 34, 81, 120, 121, 123, 126,
1 2 7 , 1 2 9 , 1 3 2 , 133, 1 3 4 , 1 3 5 , 136,
137, 160, 161, 174, 187, 188, 189,
2 2 8 ,2 2 9 ,2 3 2 ,2 3 3 ,2 4 5 ,2 8 8
folk art, 1
four, ix, x, 6, 10, 14, 19, 22, 2 5 , 37,
39, 4 7 , 4 8 , 5 1 , 6 0 , 6 1 , 6 2 , 6 4 , 65,
67, 7 5 , 7 6 , 7 7 , 7 8 , 7 9 , 8 1 , 8 2 , 124,
1 2 6 , 1 3 1 , 1 3 2 , 1 5 7 , 160, 167, 168,
169, 171, 194, 1 9 8 , 2 0 1 , 2 0 2 , 2 0 4 ,
2 0 5 , 2 1 1 , 2 1 5 , 2 1 6 , 2 1 7 , 2 2 0 , 222,
2 2 3 , 2 2 5 ,2 2 6 , 2 3 1 , 234, 240, 244,
2 4 5 ,2 5 1 ,2 5 2 ,2 5 5 ,2 5 7 ,2 5 9 ,2 8 0 ,
289
four places (catuhsthana), 14, 126
funeral rite, 2 3 0
Ganapati, xii, 3 2 , 6 0 , 6 5 , 75, 78, 86,
2 7 1 ,2 7 4 ,
284
ganapatibbadra, xii, 78, 86
gandhamandala, 14, 186, 1 8 9 , 2 2 6 ,
228
'
gandika, 223
Ganeaa, xvi, 2 3 , 25, 4 1 , 4 2 , 4 4 , 60,
7 5 ,7 8 ,8 0 ,
204, 2 0 7 ,2 2 9 , 272
ganchbhadra, x v i, 78
G a n ciapurvatapaniya-U panisad, 23,
4 1 ,4 2 ,4 4
"
gate, x, 26, 3 7 , 4 2 , 4 7 , 4 8 , 6 2 , 6 4 , 65,
67
S e e also T -sh a pcd gate;
protruding gate
gatra, 25, 167, 168
gatraka, 2 1 6
gauritilaka, xi, 76, 82, 83
geom etry, xv, 2 5 3 , 2 6 2 , 263
girdle, 3 9
g o d d e ss, xiii, 7, 18, 37, 45, 46, 51,
52, 60, 6 3 , 6 4 , 6 6 , 6 7 , 7 8 , 83, 155,
IN D E X
157,
196,
209,
235,
246,
290
159,
202,
211,
239,
247,
173,
204,
214,
240,
257,
180,
205,
223,
241,
279,
190,
206,
228,
242,
280,
193,
207,
229,
244,
284,
195,
208,
232,
245,
289,
graha, 61, 74, 126, 186, 192, 2 0 7 , 2 7 6

grahadevatamandala, x
granthi, 4 0 , 140, 2 3 4 , 2 4 2
grid, x iv , xv, 2 0 , 2 4 , 2 5 , 6 8 , 7 3 , 79,
1 2 2 ,2 1 5 ,2 2 0 ,2 5 1 ,2 5 2 , 2 5 6 ,2 5 8 ,
259,
2 6 0 , 2 6 1 , 263
ground plan, x iv , 2 5 6 , 2 5 9 , 2 6 2
guardian stone, 23
guardians o f the directions, xv
G uhyakali, xv, 4 6 , 4 8 , 4 9
Gujarat, 2 , 74
Hanumat, 35
harihararamalingatobhadra, 85
hariharatimka/hariharadvadasaliiigatobhadra, xi, 84
haribaratmakadvadasalihgatobhadra ,
xi, 84
heart, 2 2 , 24, 38, 52, 127, 128, 132,
140, 2 3 2 , 2 4 4
h ea v en ly b o d y , x, 61, 6 2 , 81
hexagram , x, 17, 21, 2 4 , 3 7 , 3 9 , 40,
4 2 ,4 3 ,4 4 ,4 5 ,4 6 ,5 1 ,6 4 ,
6 5 , 150
hierarchy, 14, 126, 189, 195, 196,
197, 199, 2 0 0 , 2 0 1 , 2 1 3 , 2 6 1
Hindu, 1, 3 , 4 , 5 , 7 , 8, 11, 15, 16, 18,
2 3 ,2 7 , 4 0 ,4 5 ,4 9 , 1 5 4 ,2 5 1 ,2 5 2 ,
256,
275, 276, 278, 27 9, 280, 281,
282, 283, 284, 285, 286, 289
H in d uism , 11, 16, 161, 2 7 8 , 2 7 9 , 2 8 3 ,
286, 288
hub, 2 4 , 5 0 , 124, 128, 145, 180, 181
hym n, 36, 37, 2 5 2
2 95
im m o b ile , 119
im m o b iliza tio n , 35
in d u s i v i s m , xiii, 179, 180, 2 0 0 , 2 1 0 ,
2 1 2 ,2 1 4
initiation, 1 , 9 , 10, 1 4 , 2 9 , 123, 126,
130, 131, 132, 133, 137, 138, 156,
174, 179,-183, 185, 186, 188, 189,
190, 194, 1 9 5 , 2 0 0 , 2 0 1 , 2 0 2 , 2 0 3 ,
2 0 5 ,2 0 6 , 2 0 8 ,2 0 9 ,2 1 3 ,2 2 5 ,2 2 7 ,
2 2 8 ,2 2 9 ,2 3 0 ,2 3 1 ,2 3 2 ,2 8 8
initiation m andala, 179, 185, 189,
190, 1 9 4 / 1 9 5 , 2 0 1 , 2 0 2 , 2 0 3 , 2 0 5 ,
2 0 6 ,2 0 8 ,2 0 9 ,2 1 3
installation, 1, 145, 150, 192, 193,
234
, ,
interm edia te directions, 2 1 , 4 4 , 59,
60 61 68,216
intersectin g squares, 4 7
in tersectio n, 4 0 , 4 4 , 2 1 6
inverted triangle, 4 1 , 4 2
lanaiv agurudevapaddhati, 82, 155,
159, 164, 165, 171, 179, 1 8 2 , 2 1 7 ,
218.2 7 1
Islamic, 3, 4 5 , 4 6 , 2 7 9
is o s c e l e s triangle, 4 2
ista(y antra), 35
jagadyoni, 41
Jain, 1, 37, 1 9 8 , 2 8 1
Jayakhya-Sam hita, x vi, 6, 2 7 , 4 1 , 76,
119, 121, 122, 123, 125, 128, 129,
130, 132, 133, 136, 137, 138, 143,
144,
146, 147, 148, 1 5 1 , 2 7 2
Jayaratha, 1 3 , 4 1 , 4 2 , 182, 192, 196,
197, 2 2 2 , 2 2 3 , 2 2 5 , 2 2 7 , 2 3 0 , 2 3 1 ,
2 3 2 .2 7 2
Jung, C .G ., 5, 4 8 , 2 8 2
Jupiter, 62
kacehapaprstha(sncakra), 31
icon, xiii, x v , x vi, 5, 7, 10, 19, 20, 40,
4 6 , 6 0 , 6 1 , 7 3 , 78, 179, 180, 184,
191,
192, 193, 194, 2 0 0 , 2 3 2 , 2 8 7
im age, x iv, x v , 5, 6, 7, 18, 2 0 , 2 5 , 29,
4 5 , 4 9 , 7 8 , 8 0 , 149, 158, 159, 161,
164, 168, 169, 173, 174, 175, 176,
179, 181, 182, 183, 184, 191, 192,
193, 194, 195, 2 0 0 , 2 0 2 , 2 0 6 , 2 1 3 ,
214, 2 3 1 , 2 4 1 ,2 5 4 , 2 5 5 ,2 5 7 ,2 6 0 ,
277,
2 7 9 ,2 8 1 ,2 8 7 ,2 9 0
im age-m and ala, 1 9 , 2 0 , 159, 162, 164
Kalacakratantra, 16
K alpacintam ani, 35
Kalyanamandirastotra, 38, 271
kamakala, 4 2 , 2 4 2
kam ya, 30, 3 3 , 157, 158, 162, 172
kantha, 2 6 , 6 6 , 2 1 7 , 2 2 0
kapola , 2 6 , 2 1 7 , 2 2 0
karnavyasa s y s t e m , x iv
karnika, 2 1 , 2 5 , 80, 123, 1 2 4 , 127,
1 4 2 ,1 6 7 ,1 6 8 ,2 1 5
2 96
INDEX
K a s h m ir , 6 , 2 9 , 1 5 5 , 2 2 6 , 2 4 0 , 2 7 6 ,
287
Kaula, 1 8 0 , 2 2 8 , 2 3 1 , 2 3 5 , 2 4 7 , 2 8 0 ,
289
Kautilya, 13
kesara, 2 5 , 124, 167
Ketu, 62
kbandendu, 2 5 , 80
kohbar mandala, 1
kolam , 1
kona, 17, 2 5 , 2 9 , 40, 4 3 , 4 4 , 64, 169,
216
kostha, 2 4 , 79, 167, 2 1 5 , 2 2 0 , 22 2
Kramrisch, xv, 6, 16, 2 8 , 160, 2 5 2 ,
253, 254, 256, 258, 259, 260, 262,
2 6 3 ,2 8 2
Ksemaraja, x v i i , 3 0 , 162, 184, 194,
197, 198, 199, 2 0 0 , 2 0 2 , 2 1 0 , 2 1 1 ,
2 1 2 ,2 1 3 ,2 1 5 ,2 1 6 ,2 1 7 ,2 1 8 ,2 2 0 ,
2 2 1 ,2 4 0 ,
272, 2 7 3 ,2 7 6
ksetrapalamandala, 82
Kubjika, 2 3 9 , 2 4 7 , 2 7 2 , 2 8 0
Kubjikamatatantra, 2 3 9 , 281
Kularnava-Tantra, 13, 2 8 , 33, 34,
1 6 3,272
kunnaprstha(yantra/sncakra), 30, 31
laghugauritilaka, xi, 83
laghusarvatobhadra, 166
laghusodasaramamudraramalingatobhadra, xii, 85 .
laghusrhkhala, 2 5 , 80
laghuvalli, 25, 80
latakaraiihga(mandala), 164
latalingatobhadra, xi, 84
latalihgodbhava, 8 2 , 84, 164
left, x v , 3 , 4 1 , 4 3 , 124, 127, 129, 133,
136, 165, 174, 2 0 0 , 2 0 5 , 2 1 6 , 2 1 7 ,
220,
2 2 1 ,2 2 2 , 2 2 3 ,2 3 4 , 274
line, 13, 14, 1 8 , 2 4 , 2 5 , 2 6 , 2 8 , 2 9 ,
3 1 , 3 9 , 4 2 , 4 7 , 6 5 , 6 7 , 76, 7 8 , 7 9 ,
8 0 , 82, 122, 123, 141, 145, 161,
168, 170, 172, 182, 184, 195, 199,
2 1 2 ,2 1 5 ,2 1 6 ,2 1 8 ,2 2 2 ,
2 2 6 ,2 4 4 ,
2 5 3 , 254, 2 5 6 , 261
S e e a lso nested lines
linear, 16, 17, 2 9 , 3 1 , 148, 162
linga, x i, x vi, 4, 24, 73, 75, 76, 77,
79, 80, 81, 82, 83, 84, 86, 155,
158, 159, 164, 165, 168, 173, 175,
191, 192, 193, 1 9 4 , 2 1 2 , 2 3 1 , 2 4 4 ,

24 5
Ungasvastikabhadra, xi, 84
lingatobhadra , 4, 24, 73, 75, 7 6 , 79,
80, 8 1 , 8 2 , 8 3 , 8 6
liquidation, 34
lotus, ix, x, xiii, xvi, xv ii, 2, 8, 14, 2 0 ,
2 1 ,2 2 , 2 3 ,2 4 , 2 5 ,3 7 , 39, 4 3 , 4 4 ,
4 5 ,4 6 , 4 7 ,4 8 , 5 0 ,5 1 ,5 2 ,6 1 ,6 2 ,
6 3 ,6 4 ,6 5 ,6 6 , 6 7 ,6 8 ,7 6 ,7 8 ,7 9 ,
8 0 ,1 2 0 ,1 2 3 ,1 2 4 ,1 2 5 ,1 2 7 ,1 2 8 ,
136, 139, 140, 141, 144, 148, 150,
157, 159, 161, 167, 168, 170, 173,
180, 181, 182, 184, 195, 199, 2 0 0 ,
201, 203, 204, 205, 206, 207, 209,
2 1 0 , 2 1 5 ,2 1 6 , 2 1 8 , 220, 221, 222,
223, 225, 226, 228, 229, 232, 234,
2 4 0 ,2 8 0
S ee also eight-petalled lotus; nine
lotuses; pericarp; six-petalled
lotus; va/ra-lotus
lotus petal, ix, 2 , 21, 22, 23, 39, 45,
4 8 , 50, 6 6 , 7 9 , 2 2 1 , 2 3 2 , 2 4 0
lotus stone, 22
lotus-w heel, 181
M adhya Pradesh, x iv , x v , 2 5 9 , 2 6 2
madhyasutra, 2 16
M agen D a v id , 4 4
m agic, 1 , 3 , 2 8 , 3 0 , 3 1 , 3 2 , 3 3 , 3 4 , 3 5 ,
36, 3 9 , 4 4 , 130, 162, 163, 171,
1 7 2 ,2 0 6 ,2 5 4 , 2 5 5 ,2 5 6 , 279, 281,
288
S e e a lso six rites o f m a g ic
mahamandala, 130, 166, 167
Maharastra, xv, xvi, 7, 8, 2 0 , 2 2 , 2 3 ,
5 7 ,5 8 ,6 3 ,7 3 ,7 5 ,7 7 ,8 1 ,8 2
Malinivijayottaratantra, xiii
mandah/mamdah, 23
mandala, ix, x, xii, xiii, xiv, x vi, xvii,
\ \ 2, 3, 4, 5, 6, 7, 8 , 9 , 10, 11, 13,
14, 15, 16, 1 8 , 1 9 , 2 0 , 2 1 , 2 2 , 23,
2 4 ,2 5 , 26, 2 7 ,2 8 , 2 9 , 3 0 , 3 1 , 3 2 ,
3 3 ,3 5 ,3 6 ,3 9 ,4 4 , 4 5 ,4 7 ,4 8 ,4 9 ,
5 0 , 5 1 , 5 7 , 5 8 , 5 9 , 6 0 , 6 1 , 6 2 , 63,
7 3 ,7 4 ,7 5 ,7 6 ,7 7 , 7 8 ,7 9 , 8 0 ,8 1 ,
82, 83, 8 4 , 119, 120, 121, 122,
1 2 3 , 1 2 4 , 1 2 5 , 126, 127, 128, 129,
1 3 0 , 1 3 1 , 1 3 2 , 133, 134, 135, 136,
137, 138, 139, 140, 141, 142, 143,
146, 149, 150, 151, 153, 155, 156,
297
IN D E X
157, 158, 159, 160, 161, 162, 163,
164, 165, 166, 167, 168, 169, 170,
172, 173, 174, 175, 176, 179, 180,
181, 182, 183, 184, 185, 186, 187,
188, 189, 190, 191, 193, 194, 195,
196, 199, 200, 2 0 1 , 2 0 2 , 2 0 3 , 2 0 4 ,
2 0 5 ,2 0 6 , 207, 208, 2 0 9 , 2 1 1 ,2 1 2 ,
2 1 3 ,2 1 4 ,2 1 5 ,2 1 6 ,2 1 7 ,2 1 8 ,2 2 0 ,
2 2 1 , 2 2 2 , 2 2 3 , 2 2 5 , 2 2 6 , 2 2 7 , 228,
2 2 9 ,2 3 0 , 231, 232, 233, 239, 240,
2 4 3 , 2 5 1 ,2 5 2 , 2 5 4 , 2 5 5 , 2 5 6 , 258,
259, 260, 2 6 1 ,2 6 2 , 2 6 3 ,2 7 6 , 277,
278, 280, 281, 283, 286, 287, 288,
289
construction o f mandalas,
119-125, 1 7 0 -1 7 3 ,'2 1 5 -2 2 4 ;
e t y m o lo g y o f m andala, 1;
initiation mandalas, 1 3 0 - 1 3 8 ,
165-170, 185-194, 228-232;
m andala in the b o d y , 2 3 3 2 3 5 ;
m andalas b e s t o w in g supernatural
p ow ers, 2 0 1 - 2 0 9 ; origin and date
o f mandalas, 2 6 - 2 7 ; structural
e lem en ts o f m andalas, 2 0 - 2 6 ;
s y m b o lic m ea n in g o f m andalas,
1 3 9 - 1 4 2 , 174176; term mandala,
1 - 1 9 , 1 8 0 - 1 8 5 ; types o f
mandalas, 1 9 - 2 0 , 1 5 6 - 1 6 1 ; use o f
mandalas, 1 2 5 - 1 3 8 ; w orsh ip o f
Siva on the m andala, 1 7 3 - 1 7 4
mandala hand, 138
mandala o f the eight great B o d h i
sattvas, 2 1 , 23
mandala pattern, 1, 15, 2 1 , 82
mandalaka, 184, 1 8 6 , 2 1 0
mandalapuja, 173
Mandalasamgraha, 74
mandala(yantra), 35
mandapa , xv, 122, 143, 149, 164
manipura, 51, 52
mantra, 1, 4, 5, 6, 9, 13, 18, 2 6 , 29,
3 3 , 3 5 , 3 6 , 3 7 , 3 8 , 3 9 , 40, 5 0 ,5 1 ,
6 3 , 6 5 , 83, 119, 121, 125, 127,
1 2 8 , 1 2 9 , 1 3 0 , 1 3 1 , 133, 134, 135,
136, 137, 138, 139, 142, 144, 145,
147, 148, 150, 159, 162, 166, 168,
173, 180, 181, 183, 187, 189, 192,
193, 202, 2 0 5 , 2 1 2 , 2 2 6 , 2 2 7 , 2 2 8 ,
229, 230, 231, 232, 233, 234, 235,
2 4 0 , 241, 2 4 2 , 2 4 3 , 2 4 5 , 2 7 2 , 2 7 7 ,
2 7 8 ,2 7 9 , 2 8 0 ,2 8 4 ,2 8 5 , 2 8 6 ,2 8 7 ,
289
manuscript, 3, 4 , 14, 119, 166, 180,
191,
192, 193, 1 9 9 , 2 0 6 , 2 0 8 , 2 1 3 ,
2 7 2 ,2 7 3 ,2 7 5
marman, 3 9 , 2 5 4
Mars, 62
matrkamandala, 82
mayacakra, 82
meditation, 8, 18, 3 2 , 161, 173, 2 4 0 ,
2 4 3 ,2 4 7 ,
248,2 5 1
Mercury, 62
memkrama(sricakra), 31
meruprastara(sncakra), 3 0 , 31
meruprstha(sricakra), 3 0 , 31
m ic r o c o s m , 17, 2 7 9
miniature chain, 2 5 , 80
miniature creeper, 2 5 , 80
m isra cakra m andala, 131
Mithila, 1, 4 8
m o b ile, 2 9 , 119, 142, 143, 149, 162,
164, 175
m o o n , 1 9 , 2 1 , 2 4 , 4 4 , 6 2 , 8 0 , 120,
127,
140, 141, 144, 1 6 1 , 2 1 0
muggulu, 1
muktimandala, 201
muladhara, 51, 5 2 , 2 4 4
naksatracakra, 51
n a m e -g iv in g , 136, 187, 188, 189,
1 9 0 ,2 0 1 ,2 0 6 ,2 1 3
N arasimha, 4 6 , 2 1 0 , 211
navagrahamandala, x, 2 , 8, 57, 61,
63,
65, 67
navakona, 4 9
navamudraramatobhadra, x ii, 85
navanabhamandala, x v ii, 2 3 , 2 6 , 164,
200
navanabhapadmasvastikamandala, 84
navapadmamandala, x vi, 9, 2 3 , 76,
'
1 2 4 ,1 3 6
'
navayoni, 241
N ep a l, 1 , 2 , 4 , 7, 15, 18, 2 2 , 4 2 , 4 4 ,
4 5 , 4 6 , 4 7 , 4 8 , 4 9 , 7 6 , 8 1 , 8 2 , 84,
2 7 3 .2 7 7 .2 7 9 , 280, 281, 282, 283,
284, 285, 287, 288, 290
N e p a le s e , 3, 4 , 2 2 , 4 4 , 4 5 , 4 6 , 4 8 ,
1 9 2 .2 7 7 .2 8 0 , 281, 282, 283, 289,
290
nested lines, 2 4 , 2 5 , 2 6 , 3 9 , 4 7 , 66,
68, 80
29 8
IN D E X
Netratantra (N T ), xiii, xvii, 155, 162,

163, 180, 181, 192, 193, 2 0 5 , 209,
210, 211, 212, 213, 214, 215, 217,
2 1 8 ,2 2 0 ,
2 2 1 ,2 2 3 ,2 7 2
N evar, 11, 1 5 , 2 3 , 4 2 , 4 9
nine, x v i, x v ii, 23, 3 0 , 31, 3 5 , 4 3 , 4 9 ,
50, 51, 6 1 , 136, 140, 151, 157,
160,
194, 199, 2 0 2 , 2 0 3 , 2 2 0 , 2 2 2 ,
240, 2 4 1 ,2 4 2 , 246, 247
nine corners, 49
n ine lotuses, xvi, xvii, 23, 136, 199,
202,
203, 220
north, x, 1, 2 5 , 4 5 , 4 6 , 59, 60, 62, 65,
8 0 , 120, 127, 140, 1 5 4 , 2 1 6 , 2 4 8 ,
252,
253, 256, 260, 283, 284
north-east, 5 9 , 6 0 , 62, 6 5 , 68, 120,
1 2 1 , 2 2 9 , 25 3
north-west, 6 0 , 6 2 , 6 5 , 68
number, x, 11, 17, 19, 22, 24, 33, 36,
3 8 , 4 0 , 4 2 , 4 6 , 4 8 , 5 1 , 52, 58, 59,
6 5 , 7 6 , 7 8 , 79, 80, 155, 162, 165,
166, 169, 172, 179, 180, 185, 186,
189, 190, 194, 195, 196, 197, 199,
2 0 3 , 2 0 8 ,2 1 0 , 2 1 3 , 2 1 5 , 2 1 7 , 220,
2 2 5 , 2 4 2 , 2 4 5 , 253
o cc a s io n a l rite, 1
o cta g o n , x v , 4 7 , 261
o ffse t, 25, 2 6 , 80
Orissa, x iv, 2 5 7 , 2 5 8 , 2 8 4
outer square, 2 5 , 42, 4 7 , 2 2 2 , 2 4 1 ,
242, 246
pada, 2 4 , 52, 79, 136, 1 6 0 , 1 6 7

pada, 2 5 , 82, 127, 167
Padm a-Sam hita, x v i, 79, 119, 120,
121,
122, 124, 125, 126, 130, 135,
136, 137, 138, 143, 144, 146, 273
padmodaramandala, 131
pancabhadra, xii, 75, 86
pan ca brahmamanda la, 82
Pancaratra, xii, 6, 7, 9, 14, 18, 2 3 , 24,
2 5 , 2 6 , 2 7 , 29, 4 8 , 7 4 , 76, 82, 119,
1 2 4 , 1 3 9 , 1 4 3 , 146, 149, 155, 199,
2 5 4 ,2 7 4 ,
283, 286
pancavimsatilingatobhadra, xii, 8 4
pancavimsatimudraramalingatobhadra, 85
pancayatana, x, 8, 60
parsvasutra, 2 1 6
parvatikanta, 82, 164, 165
paScimakannayan tra, 4 9
patala(yantra), 30, 2 0 2
patra, 17, 25, 167
Pauskara-Samhita (PausS), 6, 16, 2 5 ,
2 7 , 79, 83, 119, 120,' 121, 122,
124, 125, 129, 130, 131, 132, 137,
138,
139, 1 6 6 , 2 5 7 , 2 7 3
p avilion , 2 5 , 122, 128, 143, 149, 150,
160, 164, 169, 175
pentacle, 46
pentagram, 4 6 , 4 9 , 62
pericarp, ix, 21, 2 5 , 65, 6 6 , 8 0 , 123,
124, 127, 141, 144, 167, 1 8 0 , 2 0 0 ,
2 0 2 ,2 0 5 ,2 0 9 ,2 1 1 ,2 1 5 ,2 1 6
pikhalakhu, 23
pitha, 16, 24, 25, 73, 80, 128, 139,
140, 167, 168, 176, 1 8 2 , 2 1 6 , 2 2 3 ,
24 4 ,2 4 5
placc for Brahman, xiv
point (bindu), x vi, 2, 21, 3 9 , 4 0 , 4 1 ,
4 4 , 4 5 , 6 4 , 7 9 , 127, 156, 157, 158,
161, 163, 166, 169, 170, 173, 176,
179, 180, 184, 185, 186, 188, 189,
1 9 0 , 1 9 1 , 192, 195, 1 9 8 , 2 0 0 , 2 1 2 ,
2 1 3 ,2 1 6 , 2 2 1 ,2 2 2 , 223, 226, 242,
2 5 5 ,2 5 8 ,2 5 9 ,2 6 3
points o f intersection, 4 0 , 4 4
Pott, 4, 6, 1 8 , 4 9 , 285
po w d er mandala, 1 9 , 2 0 , 186
p ower, 9, 10, 11, 14, 1 9 , 2 1 , 2 6 , 3 0 ,
3 3 ,4 1 ,4 2 ,
4 5 , 4 6 , 5 1 , 6 6 , 7 8 , 130,
1 3 5 ,1 3 7 ,1 3 8 ,1 4 0 ,1 4 5 , 146,14 7,
148, 150, 160, 171, 173, 175, 176,
1 8 3 ,1 8 5 ,1 9 2 ,1 9 5 ,1 9 8 ,2 0 1 ,2 0 2 ,
203, 204, 205, 206, 207, 208, 211,
213, 227, 228, 229, 230, 233, 239,
240, 2 4 1 ,2 4 2 , 2 43, 244, 245, 246,
2 4 7 ,2 7 6
pranapratistha, 18, 38, 81
pranapratisthamantra, 38
Prapancasara, vii, 35, 2 7 8
pratistha, 130, 145, 149, 150, 157,
1 5 8 ,1 6 0 ,1 6 5 ,1 7 3 ,1 9 1 ,1 9 2 ,1 9 3
protection, 3 1 , 3 3 , 3 6 , 130, 135, 147
protruding gate, 62, 64, 66
puja, ix, 33, 3 5 , 4 3 , 58, 60, 81, 173,
1 7 4 ,2 1 7 ,2 2 7 ,2 3 3 ,2 7 8
pujadharayantra, 3 2
pujanayantra, 3 1
pujayantra, x, 32
pura, 15, 119, 182
299
IN D E X
purakara(mandala), 164
quadratura circuit, 48
Raghavabhatla, xiii, 2 8 , 38, 166, 173,
215
"
Rahu, 62
raja rama, 2 5 , 7 7 , 80
rajas, 2 5 , 4 8 , 6 6 , 8 0 , 122, 171, 197
Rajasthan, xiv , xvi, 4, 7, 4 5 , 7 8 , 2 6 3
rajomandala, 1 9 , 2 0 , 157, 158, 174,
175,' 186
raksa, 31, 32, 3 3 , 36, 135, 145, 149,
202
raksayantra, 3 1 , 3 2 , 33
Rama, 2 5 , 3 7 , 6 7 , 7 5 , 7 7 , 8 0 , 8 1
rama raja, 77
ramalihga, xvi, 77
ramalingatobhadra, xvi, 77
ramamudra, x v i, 7 7 , 8 0
Ramaraksastotra, ix, 3 7 , 27 8
ramaraksayantra, 37, 38
ramatobhadra, 25, 73, 76, 77, 80, 81
nlrigoli, 1, 2 8 0
Rao, 5 , 6 , 1 6 , 3 0 , 3 1 , 3 2 , 7 9 , 8 4 , 2 5 6 ,
286
radicakra, 51
red, ix, xi, 2 5 , 2 6 , 4 8 , 6 0 , 6 2 , 6 6 , 80,
123, 129, 141, 167, 171, 173, 2 0 5 ,
206, 208, 209, 218, 221, 223, 233,
235
R g -V e d in , xv
R g v ed ly a b ra h m a k a rm a sa m u c ca y a , x,
' x v , 58, 6 2 , 6 3 , 6 4 , 6 5 , 271
rice flour, 1, 171, 2 0 8
right, x v , 3 2 , 4 1 , 4 3 , 5 9 , 124, 127,
132,
135, 136, 137, 146, 156, 187,
205, 216, 22 3 ,2 6 1
ritual, x, x iii, x v i, 3, 6, 8, 9, 10, 11,
13, 14, 18, 1 9 , 2 0 , 2 2 , 2 3 , 2 4 , 2 6 ,
30, 3 2 , 3 3 , 3 4 , 3 6 , 3 9 , 4 3 , 4 4 , 5 0 ,
5 1 ,5 7 , 5 8 ,5 9 , 6 0 , 6 1 ,6 3 , 6 5 ,7 3 ,
7 5 , 7 7 , 81, 82, 121, 125, 126, 128,
129, 130, 132, 134, 136, 149, 153,
154, 156, 157, 158, 160, 161, 162,
163, 164, 165, 168, 169, 170, 171,
172, 174, 175, 176, 183, 185, 186,
187, 188, 189, 190, 191, 192, 2 0 6 ,
209 , 2 1 1 ,2 1 2 , 213, 214, 225, 226,
2 27, 228, 229, 230, 231, 232, 233,
2 34, 239, 247, 252, 253, 254, 255,
257,
262, 276, 278, 279, 280, 281,
282, 285, 2 8 6 , 2 8 8 , 2 8 9 , 2 9 0
Rudra, x v , 2 5 , 5 1 , 6 2 , 6 4 , 6 5 , 66, 68,
8 0 ,2 1 0 ,2 1 1
R udra/Siva, x v , 68
rudraplthamahayantra, x, x v , 57, 63,
64,
65
rudrayaga, x v
sacred space, 26, 2 5 7 , 2 5 8 , 2 6 0
sahasradalapadma, 52
sahasraracakra, 52
Saiv a, xiii, 7, 9, 10, 14, 1 9 , 2 0 , 23,
26, 2 7 , 3 9 , 4 5 , 5 0 , 5 7 ,7 6 , 8 2 ,1 5 5 ,
156, 160, 168, 174, 176, 179, 180,
1 8 5 ,1 8 6 ,1 8 7 ,1 8 9 ,1 9 0 ,1 9 1 ,1 9 3 ,
195, 196, 197, 198, 199, 2 0 2 , 2 0 9 ,
2 1 0 ,2 1 1 ,2 1 3 ,2 1 4 , 2 2 5 ,2 2 9 ,2 3 3 ,
f 2 4 0 ,2 4 4 ,2 4 5 ,2 7 4 ,2 7 7 ,2 8 1
Saiva Siddhanta, 9, 154, 2 7 8 , 2 8 0 ,
281
S a iv a g a m a , 9, 154, 178, 2 8 0
S a iv is m , xiii, 10, 19, 153, 155, 179,
196, 2 4 1 , 2 7 7 , 2 8 7
Sakta, 7, 2 3 , 39, 4 8 , 57, 6 3 , 192, 2 7 8 ,
2 8 1 ,2 8 5
Sakti, 4 0 , 4 2 , 4 3 , 4 5 , 6 4 , 6 5 , 6 7 , 77,
145,
150, 1 6 3 , 2 0 2 , 2 1 0 , 2 3 4 , 2 4 1 ,
242, 279, 280
saktimandala, 161, 2 0 4
samayamandala, 185, 2 0 8
samdhi, 39
&amkara, 2 , 3, 37
S am k hya, 4 7 , 6 6 , 8 0 , 1 9 7 , 198, 199
samskilra, 38, 61
sanctum , x iv , x v , 2 5 8 , 2 5 9 , 2 6 0 , 2 6 1 ,
26 3
S and erson, xiii, 6, 2 4 , 2 7 , 3 0 , 50,
134, 162, 166, 179, 180, 181, 183,
184, 185, 1 8 9 , 190, 191, 193, 195,
2 0 4 , 2 1 1 , 2 1 3 , 2 2 3 , 2 2 4 , 225, 230,
2 3 1 ,2 3 3 ,2 4 0 ,2 8 7
saptadaM ingatobhadra, xi, xii, 84
Saradatilaka (^T ), xiii, 10, 2 1 , 2 8 , 34,
3 8 , 4 2 , 4 4 , 4 7 , 4 8 , 5 1 , 5 2 , 7 6 , 155,
160, 161, 165, 166, 168, 169, 170,
1 7 2 ,2 1 5 ,2 7 5 ,2 7 8
sarira(yantra), 3 3 , 3 5
sarvatobhadra, xi, xiii, x v , x vi, 10,
2 5 ,7 5 ,7 6 , 7 7 ,7 9 ,8 0 ,8 1 ,8 2 ,8 3 ,
8 4 , 124, 164, 165, 166, 1 6 7 , 2 0 5
300
IN D E X
6atarnudraramatobhadra, 85
su ttva , 25, 47, 48, 65, 80, 139, 197
Saturn, 62
S atvata-S am hita, 23, 26, 27, 123,
128,
1 3 i, 133, 136, 137, 138, 139,
144, 147, 275, 281
S a u b h ag y ab h ask ara, 17
saudarsanayantra, 9, 29, 145, 148,
149, 150
Saundaryalahari, ix, 3, 37, 38, 275
scent m andala, 186
Seal o f Solo m o n , 44
seat-m andala, 19, 20, 33, 39, 156
seed syllable, 36, 37, 38, 44, 63, 139,
140, 141, 150, 183
sem i-circular, 13, 1 5 7 ,2 0 7
serpent, 66, 67, 68, 120
S h an k aran aray an an , 17, 30, 41, 42,
287
Shield o f D avid, 44
Shingon B u d d h ism , 8, 287, 288
Siddhanta, xiii, 10, 19, 154, 155, 157,
162, 166, 175, 188, 189, 192, 193,
194,
195, 196,.199
SiddhayogeSvarim ata, xvii, 180, 182,
1 8 3 , 1 8 5 , 1 8 6 , 1 8 7 , 1 9 1 , 192, 194,
201, 205, 208, 209, 212, 214, 222,
2 2 3 .2 7 5 .2 8 8 .2 8 9
siddhi, 10, 14, 30, 66, 185, 201, 202,
2 0 3 ,2 0 5 , 2 0 6 ,2 0 8 ,2 1 2 ,2 1 3
siddhim andala, 201, 205, 208, 209
sim hasana, 167, 168
Siva, xiv, xv, 10, 24, 25, 29, 42, 43,
4 5 , 4 6 , 51, 60, 64, 65, 6 6 , 7 5 , 76,
7 7 , 8 0 , 82, 128, 153, 156, 157,
158, 159, 161, 162, 164, 165, 167,
168, 169, 170, 172, 173, 174, 184,
187, 1 9 9 , 2 0 3 , 2 0 4 , 2 0 9 , 2 1 0 , 2 1 3 ,
226, 229, 234, 241, 242, 243, 259,
2 6 2 .2 8 2 .2 8 3 .2 9 0
sivalinga, 60, 77, 244
six cakras, 52
six rites o f m agic, 34
six -petalled lotus, 22, 44
six-pointed star, 44
sketchbook, 76, 84
skull, 48, 49, 191, 193, 199, 205,
235, 244
Sm arta, x, xvi, 7, 20, 24, 48, 57, 58,
6 0 ,6 1 ,6 3 ,
82, 278
snake, 34, 62, 66
io b h a , 25, 139
3 o b h i, 25, 26, 139, 169, 2 17, 220,
221
south, x, 1 , 2 , 3 , 7 , 8 , 2 1 , 32, 35, 36,
4 8 , 6 2 , 65, 75, 82, 120, 127, 140,
154, 155, 165, 186, 240, 2 4 7 , 2 5 1 ,
2 5 3 ,2 5 6 , 2 6 0 , 2 7 1 , 2 7 7 , 2 7 8 , 279,
2 8 0 ,2 8 1 ,2 8 4 ,
290
South Asia, x, 1, 2, 3, 7, 8, 21, 32, 36,
48, 8 2 , 2 5 1 , 2 7 7 , 2 7 9 , 2 8 0 , 284,
290
south-east, 16, 60, 62, 65, 67, 229
south-w est, 60, 62, 65, 67
specialization, 10, 201
spokes, 2 4 , 5 0 , 124, 128, 145, 150,
165,
181, 182, 198, 202, 206
square, x, xi, xiv, xv, 2, 13, 19, 20,
24, 2 5 , 2 6 , 33, 37, 39, 42, 4 4 , 4 6 ,
4 7 , 4 8 , 5 1 , 5 8 , 6 0 , 6 2 , 6 4 , 65, 67,
68, 7 3 , 7 9 , 8 0 , 8 1 , 8 2 , 122, 124,
150, 157, 159, 160, 166, 167, 168,
170, 182, 202, 2 0 7 , 2 1 5 , 2 1 6 , 2 2 0 ,
222, 226, 240, 241, 242, 246, 252,
253, 255, 256, 258, 259, 260, 262,
289
See also outer square
square grid, 20, 24, 25, 68, 73, 79,
258
squaring o f the circlc, 48
Srlcakra, xiv, 2, 4, 5, 6, 7, 8, 10, 27,
3 0 , 3 1 , 3 4 , 40, 42, 4 3 , 4 5 , 5 0 , 162,
2 3 9 , 2 4 0 , 241, 242, 244, 245, 246,
247,248,
2 8 2 ,2 8 9
srlcakra according to the
Y oginihrdaya, 2 3 9 - 2 4 8
S ee also bhukram a(sricakra),
bhuprslha(sncakra )\
kacchapaprslha(sncakra)]
kailasapraslara(srlcakra j;
kurm aprastanifsricakra )\
ku n naprstha(srlcakra X
m cru kra m a (irica kra );
m eruprastarafsricakra)\
m crupr$tlw (M cakra);
sriy antra
sricakrapuja, 239, 247, 289
grimandala, xiii, xvii, 26, 45, 76, 205,
215
Srlvidya, 50, 181, 239, 242, 245, 247,
278
IN D E X
sriyantra, 2, 27, 30, 34, 40, 50, 279,
282,
286
See also sricakra
srhkhala, 24, 80
stam en, 167
star, 22, 43, 46, 141, 160, 167, 174,
1 8 1 ,2 0 2 , 2 0 5 , 2 1 6 , 2 1 8 , 2 2 1 , 2 2 9 ,
242
Sec also five-pointed star; sixpointed star
Star o f S olom on, 46
sthandila, 158, 159, 161, 164, 193,
228
sthapanayantra, 20, 32
sthiraviistu, 262
stone, 9, 14, 22, 23, 31, 34, 42, 60,
77, 123, 143, 148, 1 7 1 ,1 7 2 , 191,
233,
254, 255, 257, 258, 259, 261,
263
stotra, 36, 37
stu ti, 36
subhadra(m andala), 164
subjugation, 35, 203, 207
substitution, 10, 201, 204, 209, 212,
213
sudarsana w heel, 45
sudarianacakra, 45
SudarSanacakrapurusa, xii, 46
sula, 226
sulabjanm ndala, 166, 225
Sulba-Sutra, xiv, 27, 252, 253, 262,
263
sun, xvi, 19, 25, 60, 62, 78, 80, 125,
128,
140, 1 6 1 ,2 1 1
tfunya, 241
supernatural pow ers, 10, 14, 30, 66,
147,
183, 185, 1 9 2 , 2 0 1 , 2 0 2 , 2 0 4 ,
205, 206, 207, 208
Surya, 59, 60, 75, 128, 159, 2 1 0
suryabhadra, xii, xvi, 25, 78, 80
S vacchanda, 155, 200, 201, 202, 203,
211
Svacch an d ab h airav a, 24, 183, 193,
200,201
S v acch an d atan tra (SvT), xvii, 6, 10,

155, 157, 162, 163, 168, 170, 174,
180, 181, 182, 183, 184, 186, 187,
188, 194, 195, 196, 197, 198, 199,
200, 201, 203, 204, 2 0 9 , 2 1 0 , 2 1 1 ,
2 1 2 ,2 1 3 ,2 1 5 ,2 1 7 ,
2 2 0 , 2 7 6 , 279
svudhistlm na, 5 1 , 5 2
301
sva stika , 41, 44, 77, 79, 125, 164,
165,
1 7 3 ,2 0 3 ,2 2 5 ,2 3 2
sva stika b ja d va ya , 82, 164
sva stika m a n d a la , xii, 75, 79, 86, 125,
233
svastikasarvatobhadra, 82, 164
s y m b o lism , 1, 5, 21, 26, 42, 43, 46,
4 7 , 4 8 , 6 1 , 8 0 , 160, 166, 172, 174,
175, 1 7 6 , 2 1 2 , 2 5 3 , 2 6 2 , 2 8 0 , 2 8 2 ,
286
T -s h a p e d gate, 42, 47
ta bular device, 50, 51
tam as, 25, 47, 66, 80, 197
T a m iln a d u , xiv
Tantra, 40, 52, 184, 188, 189, 193,
196, 199, 2 0 0 , 2 0 4 , 2 1 2 , 2 2 5 , 2 7 1 ,
2 7 3 ,2 7 4 ,2 7 7 ,2 7 9 ,2 8 0 ,2 8 1 ,2 8 4 ,
2 8 6 ,2 8 7 ,2 8 8 ,2 9 0
T an tra lo k a (TA), xiii, xvii, 10, 13, 14,
4 1 , 4 2 , 134, 135, 155, 166, 175,
1 8 2 ,1 8 6 ,1 8 9 ,1 9 1 ,1 9 2 ,1 9 3 ,1 9 6 ,
197, 2 1 4 , 2 1 7 , 2 2 2 , 2 2 3 , 2 2 5 , 2 2 7 ,
2 2 8 ,2 2 9 ,2 3 0 ,2 3 1 , 2 3 3 ,2 3 9 ,2 7 2 ,
288
Tantric, ix, 1, 3, 5, 9, 11, 13, 14, 20,
26, 2 7 , 3 2 , 3 3 , 3 7 , 3 9 , 4 1 , 4 3 , 4 5 ,
4 8 , 5 1 , 5 7 , 5 8 , 6 3 , 8 2 , 133, 154,
155, 157, 180, 1 9 0 , 2 1 0 , 2 1 3 , 2 2 7 ,
2 2 8 ,2 3 4 ,2 5 7 ,2 7 6 ,2 7 7 , 2 7 8 ,2 7 9 ,
2 8 0 ,2 8 1 ,2 8 2 ,2 8 3 ,2 8 5 ,2 8 6 ,2 8 7 ,
2 8 8 .2 8 9
T antrism , 11, 43, 153, 155, 162, 256,
2 7 8 .2 8 1 .2 8 5 .2 8 9
tem ple, xii, xiv, xv, 1, 5, 11, 15, 16,
18, 2 5 , 2 9 , 3 1 , 3 5 , 4 0 , 4 2 , 4 5 , 4 6 ,
47, 60, 75, 80, 82, 120, 121, 122,
130, 142, 148, 149, 150, 159, 164,
169, 174, 175, 176, 1 9 2 , 2 1 3 , 2 5 1 ,
2 5 2 ,2 5 4 ,2 5 5 ,2 5 6 ,2 5 7 ,2 5 8 ,2 5 9 ,
260, 26 1 , 262, 263, 278, 27 9 , 28 2 ,
283, 284, 286, 288
three, xi, xiii, xv, 9, 10, 15, 16, 17,
19, 2 2 , 2 3 , 2 4 , 2 5 , 3 0 , 3 1 , 3 3 , 3 8 ,
39, 4 0 , 4 1 , 4 2 , 4 3 , 4 7 , 4 8 , 5 8 , 6 4 ,
65, 66, 6 7 , 6 8 , 80, 120, 121, 124,
126, 132, 133, 139, 140, 154, 155,
156, 157, 161, 171, 173, 194, 195,
196, 1 9 7 , 2 0 1 , 2 0 3 , 2 0 4 , 2 0 5 , 2 1 1 ,
2 1 6 , 2 1 7 , 2 2 0 , 2 2 2 , 2 2 3 , 2 2 4 , 225,
226, 228, 229, 232, 234, 240, 241,
302
IN D E X
2 4 2 , 244, 245, 247, 252, 259, 260,

263, 280
three constituents (g u n a ), 47, 66, 80
three gunas, 42, 173, 197
three-dim ensional, 14, 1 6 , 2 4 , 3 0 , 3 1 ,
195, 251
throne, xiii, 24, 25, 80, 81, 140, 161,
167,
168, 174, 189, 1 9 5 ,2 1 1 , 234
T ibetan, 1 , 3 , 7 , 16, 4 4 , 4 9 , 7 6 , 7 9 ,
156,
1 7 6 , 2 1 3 , 2 7 7 , 2 8 4 , 2 8 7 , 289
to w n -p lan n in g , 1, 1 5 ,7 5
trayodagalingasam udbhavam andala,
xii, 86
trayodasam udraram atobhadra, xii, 85
triangle, x, 2, 13, 17, 20, 24, 31, 33,
37 ,3 9 , 4 0 , 4 1 ,4 2 ,4 3 ,4 4 ,5 0 ,5 1 ,
6 2 ,6 4 ,
6 5 , 2 4 0 , 2 4 1 , 244
See also d o w nw ard-oriented
triangle; dow n w ard -p o in tin g
triangle; equilateral triangle; fire
triangle; in verted triangle;
isosceles triangle; upw ardoriented triangle; upw ard-pointing
triangle; w ater triangle
triangular, 1 3 , 4 1 , 4 2 , 4 3 , 2 0 7 , 2 4 4
trident, xiii, xvii, 14, 24, 39, 48, 65,
67, 169, 179, 1 9 5 , 2 0 0 , 2 0 5 , 2 1 3 ,
2 22,
223, 224, 22 5 , 226, 228, 229,
2 3 2 ,2 3 4
Trika, 6, 14, 24, 29, 134, 155, 157,
162,
179, 180, 189, 190, 191, 195,
205, 213, 225, 226, 228, 229, 230,
231, 234, 287
T rik a sa d b h a v a tan tra , xiii
trikona, 17, 41, 64, 207
tripuram andala, 82
Tripurasundari, 6, 37, 45, 239, 240,
245
trisulabjam andala, xiii, 10, 2 4 , 2 2 6 ,
227,
228, 233
tritrisulabjam andala, xiii, 10, 24, 226,
229, 2 3 0
''
tryasra, 41
T ucci, 5, 16, 2 8 5 , 2 8 9
turm eric, 29, 123, 208
tw o-d im en sio n al, 16, 24, 30, 31
um akanta(m andala), 164
universe, 9, 2 4 , '2 6 , 2 8 , 139, 140, 141,
142, 196, 2 0 0 , 2 0 2 , 2 0 9 , 2 1 0 , 2 1 3 ,
234, 242, 244, 245, 246, 252, 253,

2 5 5 ,2 6 3
upakantha, 26, 217
upasobha, 25, 139
upaiobha, 25, 26, 139, 169, 217, 220,
221
upw ard-oriented triangle, 41
upw ard-pointing triangle, 42, 43, 46
utsava, 123, 130, 158
vahanasuraksayantra, 35
V aisnava, 9, 57, 76, 7 7 , 2 1 0 , 2 1 1 ,
2 1 2 ,2 7 6 ,2 8 3
vaisvadcva, xv, 57, 58
vajra, 22, 44, 48, 65, 67, 169
wi/ra-lotus, 2 2 , 4 4
vajra dim tum an da la, 23
Vajravarahi, 45, 49
V ajrayogini, 44, 45, 280, 290
vainya, 40
va Hi, 25, 80
vapi, 25, 79
V araham ihira, 11, 27, 253, 254, 255,
256, 2 5 8 , 2 5 9 , 2 6 2 , 2 7 3 , 2 7 5
vastu, 59, 254, 258
See also caravastu, sthiravastu
vastum andala, xiv, 11, 19, 160, 161,
256, 259, 260
vastupurusam andala, xiv, 6, 7, 11, 16,
27, 2 5 3 , 2 5 4 , 2 5 5 , 2 5 6 , 257, 258,
2 6 0 ,2 6 1 ,2 6 2 ,2 6 3 ,2 8 5
vedi, xiv, xv, 164, 170, 259
vedlbandha, xiv, xv, 259
Vedic, 26, 27, 4 8 , 5 7 , 60, 62, 133,
157, 1 9 0 , 2 0 1 , 2 1 0 , 2 5 1 , 2 5 2 , 2 5 3 ,
2 7 6 ,2 7 7 ,2 8 0 ,2 8 7 ,2 8 8 ,2 9 0
Venus, 62
vegm an, 16, 182
viglm am arda, xii
V im alaprabha, 16
vim âtilingatobhadra, xii, 86
Visnu, 14, 24, 26, 29, 42, 46, 51, 57,
60, 62, 64, 76, 77, 81, 121, 122,
123, 124, 127, 129, 130, 131, 132,
134, 136, 137, 138, 139, 140', 141,
142, 148, 150, 2 0 9 , 2 1 0 , 2 1 1 , 2 7 4 ,
2 7 7 ,2 8 1
V isn u -S am h ita, 14, 24, 26, 121, 122,
123,
124, 131, 136, 137, 138, 139,
1 4 0 , 1 4 1 , 1 4 2 , 274
visoyantra, 2
3 03
IN D E X
visualization, 8, 128, 132, 140, 146,
174, 180, 183, 185, 190, 1 9 1 ,2 0 6 ,
2 1 0 ,2 1 3 ,2 1 4 ,2 8 7
visuddhi, 5 1 , 5 2
vithi, 24, 25, 80, 128, 139, 140, 169,
1 8 2 . 2 1 7 . 2 1 8 , 220, 2 2 1 , 2 2 3
vivekam andala, 125
vrata, 8, 1A, 74, 7 6 , 7 8 , 8 1 , 8 3 , 2 8 3
vratodyapana, xvi, 73
vulva, 13, 157
vulva-shaped, 208
vyom arekha, 216
vyuha, 75, 125
wall, xiv, 1, 2, 4, 169, 256, 258, 259,
260,
261, 262
water, 2 1 , 4 1 , 4 4 , 4 9 , 81, 120, 121,
122,
128, 129, 130, 138, 147, 158,
1 6 4 ,2 1 0 ,2 5 5
water triangle, 44
well, 3, 9, 11, 14, 1 8 , 2 1 , 2 5 , 2 7 , 3 2 ,
36, 40, 41, 46, 49, 5 1 , 5 7 , 63, 76,
79, 80, 120, 121, 128, 154, 156,
157, 160, 161, 164, 166, 171, 179,
183, 186, 189, 190, 191, 193, 194,
195, 198, 2 0 1 , 2 0 5 , 2 0 7 , 2 0 8 ,2 0 9 ,
210, 211, 212, 214, 218, 2 2 3 ,2 3 1 ,
235, 239, 240, 2 5 1 , 2 5 3 , 2 5 5 , 2 5 6 ,
258,
259
west, xiii, xvii, 4, 62, 65, 66, 120,
252, 253, 256, 277, 283
wheel, 24, 46, 49, 50, 51, 52, 65, 79,
124, 128, 145, 151, 163, 180, 181,
182, 183, 1 9 8 ,2 0 2 , 206, 2 2 7 , 2 3 2 ,
277
white, x, 6, 25, 26, 48, 62, 66, 76, 79,
8 0 , 1 2 2 , 1 2 3 , 1 2 9 , 1 3 6 , 139, 141,
167, 170, 171, 1 7 3 , 2 0 5 , 2 0 8 , 2 0 9 ,
2 1 6 .2 1 8 ,
2 2 1 ,2 2 3 ,2 3 5
w o m b , 41, 124
w o m b o f the w orld, 41
W o o d ro ffe , 4, 17, 34, 272, 28 6 , 288,
290
w orship, xiii, 1 , 2 , 5 , 6, 9, 10, 14, 17,
18, 19, 2 0 , 2 3 , 2 8 , 2 9 , 3 0 , 3 1 , 3 2 ,
3 3 ,3 4 ,3 5 ,3 6 ,3 9 ,4 3 ,4 5 ,4 9 ,5 7 ,
58, 6 0 , 6 1 , 6 3 , 7 8 , 8 1 , 8 2 , 119,
120, 124, 125, 126, 127, 128, 129,
130, 132, 133, 134, 137, 138, 141,
144, 146, 148, 149, 153, 156, 157,
164, 173, 180, 181, 185, 186, 187,
1 9 0 , 1 9 1 , 1 9 2 , 1 9 3 , 1 9 4 , 1 9 5 ,1 9 9 ,
2 0 1 ,2 0 2 ,2 0 3 ,2 0 4 ,2 0 6 ,2 0 7 ,2 0 8 ,
2 0 9 ,2 1 0 ,2 1 1 ,2 1 2 ,2 1 3 ,2 2 6 ,2 2 7 ,
2 2 8 ,2 2 9 ,2 3 2 ,2 3 3 ,2 3 9 , 2 5 1 ,2 5 4 ,
2 5 7 ,2 7 6 ,2 7 8 ,2 7 9 ,2 8 5 ,2 8 6
ya g a , 1 6 , 6 3 , 1 1 9 , 125, 1 3 1 , 1 3 2 , 1 3 8 ,
173,
194, 204, 2 3 1 , 2 3 2
y a g a d ik sa , 132
yagahasta, 138
yantra, ix, x, xii, xiii, xv, xvi, 1, 2, 3,
4 , 5 , 6 , 7, 8, 9, 10, 11, 14, 15, 16,
18, 1 9 , 2 0 , 2 1 , 2 2 , 2 3 , 2 4 , 2 5 , 2 6 ,
27, 2 8 , 2 9 , 3 0 , 3 1 , 3 2 , 3 3 , 3 4 , 3 5 ,
3 6 ,3 7 ,3 8 ,3 9 ,4 0 ,4 1 ,4 2 ,4 3 ,4 4 ,
4 5 ,4 6 ,4 7 ,4 8 ,4 9 ,5 0 ,5 7 ,5 8 ,6 1 ,
63, 64, 65, 66, 6 7 , 6 8 , 1 1 9 , 1 4 2 ,
143, 144, 145, 1 4 6 ,1 4 7 , 148, 149,
1 5 0 ,1 5 3 , 1 5 5 , 1 5 6 , 1 6 1 , 1 6 2 , 1 6 3 ,
164, 172, 181, 1 8 3 , 2 0 2 , 2 0 3 , 2 4 0 ,
2 5 8 , 2 7 5 , 2 7 7 , 2 7 8 , 2 7 9 , 2 8 0 , 282,
284, 285, 286, 2 8 7 , 2 8 8 , 2 8 9
co n stitu en t parts o f yantras,
3 9 - 4 9 ; e t y m o lo g y o f yantra, 28;
general rem ark s on yantras,
2 8 - 3 0 , 1 4 2 -1 5 1 , 1 6 1 -1 6 3 ; types
o f yantras, 3 1 - 3 9 ; y a n tra
structures, 6 3 - 6 8
yantra for attraction, ix, 32
yantra for subjugation, ix
Y an tracin tam an i, ix, 35, 274
ya n tra g a ya tri, 38
yantras e m p lo y e d in optional desireoriented rites, 32, 34, 39, 63
yantras e m p lo y e d in reg u lar w orship,
33
y a n tra s for establishing a fo undation,
3 3 ,3 6 ,3 9
yellow , ix, xi, 26, 4 8 , 62, 80, 123,
1 2 9 ,1 3 9 ,1 4 1 ,1 4 3 ,
1 4 9 ,1 6 7 ,1 7 1 ,
1 7 3 ,2 0 8 ,2 1 6 ,2 1 8 ,2 2 1 ,2 2 3
Y ogini, xiv, 10, 13, 41, 65, 155, 180,
183, 206, 209, 22 7 , 239, 274, 280,
2 8 5 ,2 8 7
Y o g in ih rd a y a (Y H ), xiv, 10, 41, 239,
2 4 0 ,2 4 1 ,2 4 2 ,2 4 4 ,
2 4 6 ,2 7 4 ,2 8 5
y o n i, 1 3 , 4 1 , 241
Zimmer, 4, 6, 17, 290
BRILLS IN D O L O G IC A L LIBRARY
IS S N 0 9 2 5 -2 9 1 6
1. S harm a, A. (ed.). Essays on the Mahabharata. 1991.

ISBN 90 04 09211 0
2. A guilar i M atas, E. Rgvedic Society. 1991. ISB N 90 04 09352 4
3. Fuss, M . Buddhavacana and Dei Verbum. A Phenom enological and
Theological C om parison o f Scriptural Inspiration in the S addharm apundarika S utra and in the C hristian T radition. 1991.
ISBN 90 04 08991 8
4. Sil, N .P Ramakma Paramahamsa. A Psychological Profile. 1991.
ISBN 90 04 09478 4
5. N ijenhuis, E. te (ed. & tr.). Sangitasiromani. A M edieval H an d b o o k
o f Indian M usic. 1992. ISB N 90 04 09498 9
6. H arris, I.C . The Continuity o f Madhyamaka and Togcicdra in Indian M ahayana Buddhism. 1991. ISB N 90 04 09448 2
7. G etlun, R .M .L . The Buddhist Path to Awakening. A Study of the
Bodhi-Pakkhiya D ham m a. 1992. ISB N 90 04 09442 3
8 . G om m ans,
The Rise o f the Indo-Afghan Empire, c. 1710-1780.
1995. ISB N 90 04- 10109 8
9. T akahashi, T . Tamil Love Poetry and Poetics. 1996.
ISB N 90 04 10042 3
10. C hatterjee, K . Merchants, Politics and Society in Early Modern India.
Bihar: 1733-1820. 1996. ISB N 90 04 10303 1
11. B entor, Y. Consecration o f Images and Stupas in Indo-Tibetan Tantric
Buddhism. 1996. ISB N 90 04 10541 7
12. G uenther, H . The Teachings o f Padmasambhava. 1996.
ISBN 90 04 10542 5
13. H o u b en , J.E .M , (ed.). Ideology and Status o f Sanskrit. C ontributions to
the H istory o f the Sanskrit Language. 1996.
ISBN 90 04 10613 8
14. D onkin, R.A. Dragons Brain Pefume. A n H istorical G eography of
C am phor. 1999. ISBN 90 04 10983 8
15. M cLeod, J . Sovereignty, Power, Control. Politics in the States of
W estern India, 1916-1947. 1999. ISBN 90 04 11343 6
16. H o u b en , J.E .M . & Van Kooij, K .R . (eds.). Violence Denied. 1999
ISB N 90 04 11344 4
17.Palsetia,J.S. The Parsis o f India. P reservation o f Id en tity in B om bay
City. 2001. ISBN 90 04 121145
18.
B iihnem ann, G. et al. Mandalas and tantras m the Hindu Traditions
2003. ISBN 90 04 12902 2
'
19. G om m ans, J. & Prakash, O. (eds.). Circumambulations in South Asian
History. Essays in H o n o u r o f D irk H.A. KolfF. 2003
ISBN 90 04 13155 8
'

Gudrun Bühnemann - Mandalas and Yantras in The Hindu Traditions

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gudrun Bühnemann - Mandalas and Yantras in The Hindu Traditions

Uploaded by

Copyright:

Available Formats

BRILLS

MANDALAS AND YANTRAS

T his book is printed on acid-free paper.

Copyright 2 0 0 3 by Koninklijke B rill NV, I A d en , T he .Netherlands

Notes on Contributors ...................................................................

In tro d u c tio n .......................................................................

Some O b s e rv a tio n s ..........................

Mandalas and Yantras in Smarta R it u a l...........................................

The Use o f Mandalas and Yantras in the Pancaratra Tradition ...

Vastupurusamandalas: Planning in the Image of Man ................

Bibliography and Abbreviations.........................................................

H E L E N E B R U N N E R [-LACHAUX] was a m em b er o f the Centre

M I C H A E L W . MEISTER holds the W. Norman Brown Chair o f South

is Dirccteur de recherche honoraire at the Centre

NOTES ON CONTRI BUTORS

N ew York: State University o f N ew York Press, 1990); L e coeur de

(Indology) in 1998 from the University o f Vienna. She is presently a

Illustrations to M andala, Yantra and Cakra: S o m e O bservations b y

Illustrations to M andalas and Yantras in Smarta R itu a l b y Gudrun

Part II, Bhadramandalas

LIST OF I LLUSTRATI ONS

here. Since the original numbers o f the illustrations as printed in the

29. Saptadasaliiigatobhadra, type 2

Illustration to The Use o f M andalas and Yantras in the Pahcaratra

LIST OF I LLUS TRATI ONS

Illustration to M andala and Yantra in the Siddhanta S c h o o l o f

Illustration to Icons o f Inclusivism : M andalas in S o m e E arly Saiva

Illustrations to M andalas in A b h in a va g u p ta s T antraloka b y A n d re

Illustrations to The Sricakra according to the First Chapter o f the

Illustrations to Vastupurusam andalas: Planning in the Im age o f

LIST OF I LLUSTRATI ONS

5. Khajuraho, M adhya Pradesh:

6 . A ramalihgatobhadra with 26 ramamudras and 28 Jihgas; pain

LI ST OF I LLUS TRATI ONS

16. The M m andala o f the N etratantra follow ing K s e m a ra ja s

Gudrun Buhnem ann

yantras, like those o f mandalas, have had widespread influence. In

ledges the influence o f the Saundaryalahari attributed to Sam kara on

often repetitive. One problem is the authenticity o f mandala and

large sections, dealing w ith mandalas and yantras, w hich influenced

Several authors have studied individual mandalas and yantras.

The Scope o f this B o o k

o f yantras in Indian medical systems, astrology or folk traditions, or

M arion R astellis essay focuses on the use o f m andalas and y a n

This chapter offers a description o f the d e s c e n t (avatara) o f the

cies in spellings o f words from Indian languages used by the diffe

Gudrun B uhnem ann

1 F o r a d is c u s s io n o f the u n c e rta in e ty m o lo g y o f th e w ord m a n d a la , see M ayrh o fer 1 9 8 6 - 2 0 0 1 , v o lu m e 2: 294, A re lig io u s e ty m o lo g y o f the w o r d a p p e a rs in

m angalatvac ca dakinya yoginiganasarnsrayat /

coloured powders, precious stones, fruits and leaves, and fragrant

M ANDALA, YANTRA AND CAKRA

o f the m andala its m ost elevated part, w hich could be taken to

s See, fo r e x am p le, G u ts c h o w /K o lv e r 1975, w h e re the a u th o rs describe the layout

GUDRUN 1301-IN EM ANN

synonym s o f m andala found in the literature are yHgn,H) blmvana/

angles and petal-like parts; that w hich consists o f angles alone is

appears to be based on a statement by Pott (1946) 1966: 71, w ho

determine how the terms have been em ployed by different authors

M AN D ALA, Y ANTRA AND CAKRA

near thresholds, on roads or in public places. T hese stones, w hich

M AN D AL A, YANTRA AND CAKRA

tobhadra include (see Table, p. 87): a w ell (vapi), an offset design

tirimandala and the navanabliamandala reconslructed in Torzsok (sec