SMC 80 Coomaraswamy 1928 6 1-43

SMITHSONIAN MISCELLANEOUS COLLECTIONS
VOLUME 80. NUMBER 6
YAKSAS
(With 23 Plates)
BY
ANANDA K. GOOMARASWAMY
Keeper of Indian, Persian, and Muhammadan Art
Museum of Fine Arts, Boston
(Publication 2926)
CITY OF WASHINGTON
PUBLISHED BY THE SMITHSONIAN INSTITUTION
MAY 8, 1928
ZU Boxi) (âfttmore (preee
BALTIMORE, MD., C. S. A.
YAKSAS
By ANANDA K. COOMARASWAMY
(With Twenty-three Pi.ates)
I. INTRODUCTION
In centuries preceding the Christian era, when the fusion of races
in India had already far advanced, the religion of India passed
through its underwent the most profound changes.
greatest crises and
Vedic ritual, indeed, has survived in part up to the present day but ;
the religious outlook of medieval and modern India is so profoundly

different from that of the Vedic period, as known to us from the
extant literature, that we cannot apply to both a common designation ;
medieval and modern Hinduism is one thing, Vedic Brahmanism

another. The change is twofold, at once inward and spiritual, and
outward and formal.
No doubt we are sufificiently aware of the spiritual revolution indi-
cated in the Upanisads and Buddhism, whereby the emphasis was
shifted from the outer world to the inner life, salvation became the
highest goal, and knowledge the means of attainment. But while this
philosophic development and spiritual coming of age have gradually
perfumed (to use a characteristically Indian phrase) the whole of
Indian civilization, there are here a background and ultimate signifi-
cance given to the social order, rather than the means of its actual
integration ; the philosophy of the Upanisads, the psychology of
Buddhism, indeed, were originally means only for those who had left
behind them the life of a householder, and thus in their immediate
application anti-social. But few in any generation are ripe for the
attainment of spiritual emancipation, and were it otherwise the social
order could not survive. The immediate purpose of Indian civiliza-
tion is not Nirvana or Moksa, but Dharma; not a desertion of the
household life, but the fulfillment of function. And here, in Karma-
yoga, the spiritual support is found, not in pure knowledge, but in
devotion to higher powers, personally conceived, and directly ap-
proached by appropriate offices (piija) and means (sddhand). In
the words of the Bhagavad Glta: " He who on earth doth not follow
the wheel (of activity) thus revolving, liveth in vain He that
doeth that which should be done, he is the true Monk, the true Yogi.
Smithsonian Miscellaneous Collections Vol. 80, No. 6

;
2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 80
not the recluse who refrains from actions Whatsoever thou

doest, do thou that as an offering to Me ; thus shalt thou be liberated.
.... He who offereth to Me with devotion a leaf, a flower, a fruit,
or water, that I accept Howsoever men approach Me, even
so do I welcome them, for the path men take from every side is
Mine."
In the earlier Vedic books there is a total absence of many of these
most fundamental features of Hinduism properly so called it is only ;
in the Brahmanas and Upanisads (and afterwards, much more defi-

nitely in the Epics) that the ideas of Samsara (the cycle of birth and
rebirth), Karma (causality), religious asceticism and Yoga, and
Bhakti (devotion to a personal deity) begin to appear, and the same
applies to the cults of Siva, Krishna, Yaksas, Nagas, innumerable
goddesses, and localized deities generally. It is natural and reason-
able to assume that these ideas and from the Vedic deities derive, not
Aryan tradition, but, as De la Vallee-Poussin expresses it, from " un
certain fond commun, tres riche, et que nous ne connaissons pas
*
parfaitement."
There much
to be said for Fergusson's view {Tree and Serpent
is
Worship, 244) that " Tree and Serpent worship," i. e., the worship
p.
of Yaksas and Nagas, powers of fertility and rainfall, " was the primi-
tive faith of the aboriginal casteless Dasyus who inhabited northern
India before the advent of the Aryans." But in using language of
this kind, a certain degree of caution is necessary for, in the nature ;
of things, it is only the popular and devotional aspect of these " primi-
tive faiths " of which we are able to recover the traces, and there may
well have existed esoteric and more philosophical phases of the same
beliefs. We do not know how much of Indian philosophy should
really be traced to Agamic rather than Vedic origins. Indians them-
selves have always believed in the existence of theistic scriptures, the
Agamas, coeval in antiquity with the Vedas ; and if the existence of
* For these groups of ideas as foreign to the Vedas, and for their indigenous
source, see De la Vallee-Poussin, Indo-Europeens et Indo-Iraniens; L'Inde
jusque vers 300 av. J. C, Paris, 1924, pp. 303, 315-6, 320, etc. ; Senart, E.,
Castes, pp. xvi-xvii ;
Jacobi, H., The Gaina Sutras, S. B. E., XXII, p. xxi
Keith, A. B., Religion and philosophy of the Veda, Harvard Oriental Series,
vols. 31, 32, pp. 132, 193, 258; Macdonell, A., Vedic Mythology, pp. 153, 1541
Vogel, J. Ph., Indian Serpent lore, ig26; Charpentier, J., Vber den Bcgriff
und die Etyuiologie von piija, Festgabe Hermann Jacobi, 1926.
It is to be noticed that all the clans particularly associated (so far as the
materials here relied upon are concerned) with Yaksa worship, are by no means
completely Brahmanised, and probably are not of Aryan origin (De la Vallee-
Poussin, LTnde p. 182).
;
NO. 6 YAKSAS — COOMARASWAMY 3
such scriptures is beyond proof, it is at least certain that religious
traditions,which must be spoken of as Agamic in contradistinction

to Vedic, are abundant and must reach far back into the past. This
past, moreover, has been proved by recent archeological discoveries
to have been much more ancient and to have been characterized by a
much higher culture than had been formerly recognized. And we
know so well the continuity of Indian racial psychology during the
historical period, that we cannot but believe that long before this
period begins the Indians had been, as they are today, essentially
worshippers of personal deities.
In the beginning, when Aryans and non-Aryans were at war, in
the period of military conquest and greatest social exclusiveness, and
before the two elements had learned to live together, or had evolved
a conception oflife covering and justifying all its phases, a divergence
between the two types of religious consciousness had been profound

in those days the despised worshippers of the sisna (phallus) might
not approach the Aryan sacrifice. As time passed the dividing lines
grew fainter, and end there was evolved a faith so tolerant and
in the
so broad that it could embrace in a common theological scheme all
grades of religious practise, from that of the pure monist to that of
savages living in the forests and practising human sacrifice.
Now, regarding the accomplished fact, it is not always easy to dis-
tinguish the separate elements that made so great a creative achieve-
ment possible. We are apt both to over- and underestimate the sig-
nificance of what we describe as primitive animism.
Hinduism, quantitatively regarded, is a worship of one deity under
various aspects, and of genii and saints and demons, whose aid may
be invoked either for spiritual or for altogether material ends. This
Hinduism, in the period we have referred to, broadly speaking, that
of the last three centuries before Christ, was not so much coming
into existence for the first time, as coming into consciousness and
prominence.
Dr. \'ogel, in Indian Serpent Lore, has very recently and very
admirably studied the old Indian (or perhaps we ought rather to say,
the Indian aspect of the widespread Asiatic) cult of Nagas or Dragons,
guardian spirits of the Waters.
In the following pages I have attempted to bring together, from
literary and monumental sources, material sufficient to present a

fairly clear picture of an even more important phase of non- and
pre-Aryan Indian " animism," the worship of Yaksas and Yaksis,
and to indicate its significance in religious history and iconographic
evolution.
;
4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8o
2. YAKSAS AND YAKSATTVA (" GENI-HOOD ")

The status of a Yaksa as typically represented (i) in the later
sectarian literature and (2) in modern folklore will yield an imper-
fect,and indeed an altogether erroneous idea of the original signifi-
cance of Yaksattva if not examined with cautious reservations. As
^
remarked by Mrs. Rhys Davids :
The myth of the yakkha, and its evohition still, I beHeve, await investigation.
The English equivalent does not exist. " Geni " (djinn) is perhaps nearest
(cf. Pss. of the Sisters, p. 30). In the early records, yakkha as an appellation is,
like naga, anything but depreciative. Not only is Sakka so called (M. i, 252),
but the Buddha himself is so referred to in poetic diction (M. i, 383).^
Wehave seen Kakudha, son of the gods, so addressed (Kindred Sayings,
II, 8) and in D. II, 170 the city of the gods, Alakamanda, is described as
;
crowded with Yakkhas ("gods")- They have a deva's supernormal powers.

.... But they were decadent creatures, degraded in the later era, when the
stories of the Jataka verses were set down, to the status of red-eyed cannibal
ogres.
And it may be added that it was only natural that in losing their
importance as tutelary deities, the Yaksas in popular folklore, influ-
enced no doubt by the prejudices already referred to as apparent in
the sectarian literature, should likewise have come to be classed with
the demoniac Raksasas.' Their fate in this connection may be com-
pared with that of the Devas at the hands of Zoroaster, or that of the
older European mythology under the influence of Christianity (e.g.,
in Saxo Grammaticus). Notwithstanding this, it is quite possible to
gather both from the sectarian and the semi-secular literature a great
deal of information incidentally presenting unmistakable evidences
of the Yaksas' once honorable status, their benevolence toward men.
^ Book
of the Kindred Sayings, I, 1917, p. 262. In the above citation, M. is
Majjhima Nikaya and D. is Dialogues of the Buddha. An excellent article on
Yakkhas in Buddhist literature will be found under Yakkha in the P. T. S.
Pali Dictionary.
' Elsewhere
the Buddha finds it necessary to say that he is not a Deva,
Gandhabba, or Yakkha (Angnttara Nikaya, II, 37).
' For gigantic or cannibal Yaksas see Kathasaritsagara, Tawney,
I, pp. 127,
337, II, P- 594- For the cult of Yaksas (Simhalese, Yakd) surviving as
" devil-worship " in Ceylon see Callaway, Yakkun Nattanawa, London, 1829
Upham, E., History and doctrine of Buddhism, 1829 Parker, Ancient Ceylon, ;
London, 1909, Ch. IV and Yaka, Yakkhas in Index (p. 153, a dead man speak-
ing in a dream says, " I am now a Yaka"). For an excellent general account
of non-Aryan deities, local and tutelary, beneficent and .malevolent, see White-
head, H., The village gods of South India, Oxford, 1916 ("in many villages
the shrine is simply a rough stone platform under a tree"), also Mitra, S. C,
Village deities of Northern Bengal, Hindustan Review, February, 1922, and
Enthoven, R. E., The folklore of Bombay, Pt. Ill, Tree and snake ivorship.
:
and the affection felt by men toward them. As remarked by Levi

(loc. cit. infra), " le Yaksa est essentiellement un personage divin
etroitement associe par la tradition aux souvenirs locaux . . . . ils
rappellent de bien pres nos saints patronaux."

The word Yaksa* is first found in the Joiininiya Brahinana (in,
" a wondrous
means nothing more than
203, 272), where, however, it
thing." In the sense of a " spirit " or genius, usually associated with
Kubera (the chief of Yaksas) it does not appear before the period of
the Grhya Sutras where Yaksas are invoked together with a numer-
ous and very miscellaneous host of other major and minor deities, all
classed as Bhutas," " Beings," in the Grhya ritual at the close of Vedic
studies f in a somewhat later book they are possessing spirits of
disease.* The Sankhayana Grhya Sutra mentions Manibhadra.
In the Satapatha Brahinana, Kubera is a Raksasa and lord of rob-
bers and evil-doers this may only mean that he was an aboriginal
:
deity, alien to Brahman orthodoxy. In the Sutras he is invoked with

Isana for the husband in the marriage ritual, and his hosts plague
children (cf. Hariti in her original character).

are represented and named at
The following Yaksas and Devatas
Bharhut :Yakho, Yirudhako Yakho, Gaiigita Yakho.
Su]:iavasu
Suciloma Yakho, Kupiro Yakho (Kuvera), Ajakalako Yakho;
Sudasana Yakhi, Cada (Canda) Yakhi Sirima Devata, Culakoka ;
Devata, INIahakoka Devata.

Yaksas by name or as a class are much more familiar figures in the
Epics. In the Rdmdyana, 3, n, 94, we find yaksattva amaratvani ca,
" spirithood and immortality " together, as boons bestowed by a god
or gods. Men of the Sattvik (" pure ") class worship the gods
(Devas), those of the Rajasik ("passionate") class, Yaksas and
^
The word Yaksa occurs in the following forms, which are here retained
in citations
Sanskrit, Vaksa. (/.) Vaksl, Yaksim: Pali, Yakkha. Yakkh'i. Yakkhiin:
Prakrit, Jakkha, Jakkhim; Sirhhalese, Yaka, Yakl.
The word is perhaps of indigenous non-Aryan origin. Tlie later Rdmdyana
proposes an explanation which looks like mere folk etymology : Brahma
created beings to guard the waters, and of these some cried " rak.samah," " let
us guard," and others " yaksamah," " let us gobble," becoming thus Raksasas
and Yaksas. The idea is perhaps derived from the big belly which is the most
constant feature in Yaksa iconography.

^ Siva
is " Bhiitesvara," and Yaksas are often called Bhutas the word Bhuta ;
may mean "those who have become (Yaksas)," cf. Mahdvamsa, Ch. X, verse
yakkha-bhilta, " those that had become Yaksas."
^
Sdnkhdyana Grhya Siltra, IV, 9; Asvaldyana G. S., Ill, 4; Pdi-askara G. S..
II, 12. (Keith, Religion and philosot'hy of the Jêda, p. 213.)
*
Mdnava Grhya Sutra, 11, 14: Keith, %b. p. 242.
Raksasas, those of the Tamasik (" dark ") class, Pretas and Bhutas
(Mahabharafa, 6, 41, 4) ;
in other words, the Yaksas are ranked below
the Devas, but above the goblins and ghosts and here distinguished
from Bhutas. But very often they are not clearly distinguished from
Devas and Devatas. The Yaksas are sometimes sylvan deities, usually
but not always gentle, like the Vanadevatas (Hopkins, Epic Myth-
ology, p. 57; Atanallya Suttanta).
Kubera or Kuvera (Vaisravana, Vaisramana, also in Buddhist lit-
erature Vessavana, Pancika, Jambhala, etc.),' is one of the Four

Great Kings (Maharajas), or EightGreat Devas, a Lokapala, Regent
of the North (sometimes, with Indra, of the East), and the chief of
all Yaksas, whence his epithets Yaksendra, Deva Yaksaraja, etc. He
is a god of power and productivity : worshipped especially for trea-
sure (as Dhanada, Vasuda, giving wealth).^ His city Alaka situated
on Mt. Kailasa (also the abode of Siva) is a magnificent walled town,
where dwell not only Yaksas, but also Kimnaras, Munis, Gandharvas
and Raksasas. Very possibly, as M. Goloubew (Ars Asiatica, X) has
suggested, the whole of the ceiling of Cave I at Ajanta may be re-
garded as a representation of the Paradise of Kuvera. When Kubera
repairs to a convention of the gods, he is accompanied by a great host
of Yaksas, collectively designated Vaisravana-kdyika-dcvas.
Kubera has many beautiful palaces, groves, gardens, etc., on
Mt. Kailasa. These need not be referred to in detail, but it may be
remarked of the grove Caitraratha that its trees have jewels for their
leaves and girls as their fruits."
The cult of the Lokapalas or Four Great Kings (N. Vaisravana,

E. Dhrtarastra, S. Virudhaka, W. Virupaksa) was extensively devel-
oped in Khotan, where they are represented as standing on demon
vahananis* Vaisravana is here very frequently represented with
* For Jambhala see Foucher, L'Iconographie bouddhiquc dc I'Inde, I,

p. 123,
and II, p. 51 his sakti is Vasundhara, the Earth-goddess. He may be sur-

;
rounded by eight Yaksinis, Bhadra, Subhadra, etc. (ibid., II, 85).

"
He might be styled Mammon but not in a bad sense of the word, for
:
from the Indian point of view wealth, prosperity and beauty are rewards of
innate virtue, of which, according to the doctrine of Karma, Mammon could
only be the dispenser. Cf. Mahabharata, 12, 74, 3 f.
^ Both motifs are of on account of their occurrence in decorative art,

interest
the Bharhut coping reliefs showing many forms of jewel-bearing creepers
(kalpa-latd), and medieval art, especially in Ceylon (ndrl-latil designs, plate 22,
fig. 3) many examples of creepers with girls as their flower or fruit. The latter
motif, too, may have some connection with the later Arab legends of the
Waqwaq tree.
*
Stein, Ancient Khotan, figs. 30, 31, and pi. II; Serindia, p. 870.
"
NO. 6 YAKSAS COOMARASWAMY 7
shoulder flames. In this connection it should be safe to identify the

flaming Kankall Tila figure (pi. i6, fig. 2) with Vaisravana; the
corpulent body in any case is that of a Yaksa, and the flames repre-
sent the fiery energy inherent in a king.
Of Kubera's Yaksa followers we learn a good deal: they possess
the power of assuming any shape, the females particularly that of a
very beautiful woman (so that an unknown beauty is asked if she be
the goddess of the district, or a Yaks!) ;
^
they are kindly, but can
fight fiercely as guardians (Kubera himself is a " world-protector,"
and it is chiefly as attendants, guardians and gate-keepers that the
Yaksas appear in Buddhist art, equally in India and in the Far East) ;
they are sometimes specifically grouped with Nagas, more often with
gods, Gandharvas and Nagas they are known as " good folk
;
(Pimyajana) and appear to be countless in number, though few are

individually named. Manibhadra (Manivara, Manicara, Manimat)
in the Mahabhdrata (5, 192, 44 f.) is a Yaksaraja, and Kubera's
chief attendant. He is invoked with Kubera as a patron of merchants ;
this may be the explanation of the statue at Pawaya, set up by a guild

(gostha) 2):
(pi. i, fig.
Ganesa isundoubtedly a Yaksa type, by his big belly and general

character : but he is not cited by name in any lists. He is effectively
and perhaps actually equivalent to Kubera or Manibhadra.^ But the
earliest representation of an elephant-headed Yaksa seems to be that
of the AmaravatI coping, Burgess, Stiipas of Amaravati and Jaggay-
yapcta, plate XXX, i (here pi. 23, fig. i ) ; and this is not a Yaksaraja,
but more like a guhya or gana. Ganesa is son of Siva, is him- who
self called Ganesa (Lord of hosts) in the Mahahhdrata. Ganesa as
elephant-headed deity does not appear in the Epic except in the intro-
duction which is a late addition. The figure of Ganesa begins to appear
quite commonly Gupta art, about 400 A. D., e. g., at Bhumara, plate
in
18, figure I at Deogarh (pilaster left of the Anantasayin panel).

;
There is some confusion of Yaksas and Raksasas, who according

to one tradition have a common origin both have good and evil quali- ;
benevolent and malevolent as the case may be very often the

ties, ;
same descriptions would apply to either, but the two classes are not
identical, and broadly speaking we find the Yaksas associated with
^
MahahhUrata, Vana Parva, Ch. CCLXIII (Draupadi).
'There exists a " story of the Mahayak?a Manibhadra" in MS. ; see Hoernle
in Congr. Int. Orientalistes, 12, Rome, 1899, Vol. I, p. 165.
^
Cf Scherman, Dickbauchtypen in der indischen Gotterzvelt, Jahrb.
.
as. Kunst,
I, 1724. Also M. F. A. Bulletin, No. 154.
Kubera, the Raksasas with Ravana, who is their chief. Yaksas as a

rule are kindly, Raksasas bloodthirsty.''
Yaksas are not only the attendants, but also, the bearers of their
Lord Vaisravana. They play, indeed, the part of bearers or sup-
porters in all kinds of situations where their attitude is one of friendly
service ; thus, they are constantly represented as supporting the four
legs of Kanthaka, on the occasion of the Abhiniskramana (Great
Renunciation, or Going Forth of the Buddha)." They bear, too, the
pavilion in which the Bodhisattva descends to take incarnation in the
womb of Maya Devi (pi. 21, fig. i). In connection with Vais-
ravana, and other deities, the Guhyas appear in crouching dwarf-
ish forms as supporters; in fact, as "vehicles" {vahanani) as in
plate 3, figure i, etc. Some of these types have been preserved with
remarkable Far Eastern art, in the case, for example, of the
fidelity in
Jikoku-Ten of the Kondo, Nara, Japan,^ so closely resembling the

Kubera from Bharhut (pi. 3, fig. i), and the Siva figure of the
Gudimallam lihgam (pi. 17, fig. i). In the case of Siva, the Yaksa
vehicle in later images (Nataraja, etc.) has come to be regarded as a
demoniac symbol of spiritual darkness (apasnidra purusa, or mala).
Kuvera is also " Naravahana," but the Naras here in question are
not men, but mythological beings variously described, sometimes as
bird horses, which may possibly explain the occasional representation
of winged Atlantes (pi. 13, figs. 2 and 3, also Foucher, L'art grcco-
houddhiquc fig. 314). The interpretation Naravahana = borne
. . . .,
by men, is later.
As Atlantes, supporters of buildings and superstructures (pi. 13,
figs. I, 2, 3), and as garland-bearers (pi. 23, figs, i, 2) Yaksas are
constantly represented in early Indian art (Bharhut, SaficI, Gandhiira.
etc.).Those who support Kuvera's flying palace are designated
Guhyas (Mahdblidrata, 2, 10, 3) Kuvera is Guhyapati. The Guhyas ;
are essentially earth-gnomes (cf. pi. 13, fig. i). The YaksinI of
Kathdsaritsdgara, ch. XXXVII, who carries a man through the air,
is called a Guhyaki.
Some Yaksagrahas (demon possessors, causing disease) are at-
tendants of Skanda, who is sometimes called Guha, a name which
^ For
a detailed summary of the Epic accounts of Kubera and the Yak.sas, see
Hopkins, Etîc Mythology, p. 142 ff., also pp. 30, 38, 57, 67 ff., 14s, 148, etc. See
also Waddell, Evolution of the Buddhist cult, J. R. A. S. Any connection with
the Greek Kabeiros is very improbable (Keith).
'^
E. g., Foucher, L'art greco-bouddique du GandhCira, i, pp. 357, 554 ff., and
figs. 182-4, ch. ; Stein, Scrindia, p. 858.
^
For the Nara figure see Nara Horyuj'i Okagami, Vol. 38, pi. 7, or Warner,
Japanese sculpture of the Suiko period, fig. 35.
;
NO. 6 YAKSAS— COOMARASWAMY 9
may be related to the Guhyas, attendants of Kuvera (Hopkins, Epic

Mythology, pp. 145, 229).
Yaksas (like Nagas) are sometimes regarded as constructive or
artistic genii: thus Hsiian Tsang, Bk. VIII, speaks of the Asokan
remains at Pataliputra as having been built by genii (Yaksas).'
Kubera himself can be regarded as the first smelter of gold."
Comparatively few individual YaksinTs are mentioned by name
the Mahahharata (3, 83, 23) speaks of a Yaksini shrine at Rajagrha
as " world-renow^ned." But it is beyond doubt that Yaksinis wêre
extensively worshipped, in part as beneficent, in part as malevolent
beings. In the latter aspect they do not differ essentially from their
modern descendants, such as the Bengali Sitala, goddess of smallpox,
or OlabibT, goddess of cholera. The Seven Mothers (who are in part
connected with Kubera), the Sixty-four Joginls, the DakinTs, and
some forms of Devi, in medieval and modern cults, must have been
Yaksinis. In Southern India, indeed, to the present day, nearly all
the village deities are feminine. MiUclksi, to whom as wife of Siva,

the great temple at Madura is was originally a daughter of
dedicated,
Kubera, therefore a Yaksini. Durga was originally a goddess wor-
shipped by savage tribes.
The case of Hariti is too well known to need a long discussion. To
sum up her story, she was originally a Magadhan tutelary goddess,
wife of Paficika and residing at Rajagrha she was not ill-disposed,;
for her name Nanda means She was called even in Hsiian
Joy.
Tsang's time the Mother of Yaksas, and the people prayed to her for
offspring. But Buddhist legend has it that she had begun to destroy
the children of Rajagrha by smallpox, and so earned the name of
Hariti, " Thief," by which she is known to Buddhism metaphorically, ;
she was said to " devour " them, and is represented as an ogress, and
it was as an ogress that the Buddha encountered her. The Buddha
adopts the expedient of hiding her last-born child (Piiigala, who had
been a human being in a previous life, the Yaksa birth being here a
penalty) ; she realizes the pain she has been causing others, and be-
comes a convert but as she can no longer seek her accustomed food,
;
the Buddha promises that she shall receive regular offerings from
pious Buddhists, as a patroness of children and fertility. This reads
more like an explanation or justification of a cult than a true account
.^Beal, Buddhist Records, II, p. 93. Cf. also Laufer, Cilralaksaiia, pp. 189,
190,where a late Tibetan author ascribes Asoka's works at Bodhgaj-a to Yaksas
and Nagas, and speaks of certain Indian medieval sculpture and paintings as like
the art-work of the Yaksas.
"
Hopkins, Epic Mythology, p. 146.
.
lO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8o
of its origin; probably this was the best way to provide an edifying
sanction for an ancient animistic cult too strong to be subverted.
HaritI is also constantly represented together with Paficika, forming
a Tutelary Pair (Gandhara, Mathura, Java, etc., pi. 15, fig. i ;
pi. 21,
figs. 3-5).^
A YakkhinI by name, or rather, epithet, Assa-mukhi (" horse-

faced ") plays an important part in the Padakusalamdnava Jataka.
There may be specific reference to this Jataka whenever a horse-
headed Yakkhini is represented on the medallions of Buddhist rail-
ings " (pi. 12, fig. i). But the Kimnaras and Kiriipurusas, and Gand-
harvas too, typically half-human, half-equine, are a class of beings
frequenting forests and mountains (cf. the valava-mukha Cetiya, of
Pandukabhaya, infra, p. 16) and as such are sometimes naturally
represented as a part of the scenery, and in such cases there need be
no reference to the Jataka.^
In the Mante ud&vadana
a Yaksini undertakes to bring about a
marriage, and to this end has the marriage " represented " (murtivai-
vCihikam karma, presumably in a painting)."
In the Jaina Bhagavatl Sutra (Hoernle, Uvdsagadasdo, Appendix)
Punnabhadda and Manibhadda are called powerful Devas, and they
appear together to those who practise certain austerities. Another
work gives the following list of " Devas " who are obedient to Vaisra-
mana Punnabhadda, Manibhadda, Salibhadda, Sumanabhadda,
:
^
For Hariti see Foucher, The Btiddhist Madonna, and Tutelary Pair, in The
beginnings of Buddhist art; L'art greco-boiiddhique dii Gandhara; Vogel, The
Mathura school of sculpture, A. S. I., A. R., 1909-10, p. 77; Walters, On Yuan
Chwang, I, 216; Beal, Records I, no; Waddell, Lammsm, p. 99;
. . . .,
Chavannes, in T'Aiing Pao, 1904, p. 496 f

"
Mitra, R., Buddha-Gaya, pi. XXXIV,
2 Foucher, in Mem. cone. I'Asie orien-
;
tale, III, 1919, pi. I; Waddell, Report on excavations at Paialiputra, pi. I.

Perhaps also Ajanta, Cave XVII (Griffiths, pi. 142, b).
*
At Bhaja, HIIA, r. corner; Mandor, HIIA, fig. 166. Kimnaras
fig. 27, lower
in Indian literature and art are of two types (I) horse-headed, as above, and
(2) half bird, half human (siren type). Both kinds are musical, and may be
classed in this respect with Gandharvas. The masculine horse-headed type is
rare: examples in Cat. Ind. Collections, Boston, V. Rajput Painting, No. CLIX
(called Gandharvas, one Narada), and in Arts et Archeologie khmers, II,
fig. 56, his. Most likely the horse-headed type is not a Kimnara at all.
*
In the Svayambhu Pur&na, De la Vallee Poussin, J. R. A. S., 1894, p. 315.
Here we have the normal connection of Yak§inis with human marriage. The
miirti-vaivahika motif appears also in Bha§a's SvapnavSsavadatta, and is repre-
sented in a Rajput Painting of the eighteenth century {Cat. Ind. Coll., V,
Rajput paintings, p. 189).
NO. 6 YAKSAS — COOMARASWAMY II
Caksuraksa, Purnaraksa, Savvana, Savvajasa, Savakama, Saniiddha,'

Amohe, Asamta. It may be remarked incidentally that nearly all
these names of Yaksas are auspicious, implying fullness, increase,
prosperity, etc.
As we have seen, Yakkhas are often called Devas in the Jaina
books, where, as Sasana Devatas, they are usually guardian angels.
But it is whether the " false and lying Devas " who
not at all clear
persecute the followers of Mahavira in the Uvasagadasao, §§ 93 f.,
224, etc., are to be regarded as Yakkhas or not. That they should be
so regarded in one case at least (§ 164) is suggested by the fact that the
Deva here appears in an asoga (asoka) -grove and takes possession of
objects laid on an altar. It may
remarked that the Deva of
also be
§ 93 is a characteristic power for

an expert shape-shifter, which is
Yakkhas the text speaks of the " Pisaya (Pisaca) form of the Deva,"
;
and it may be that the Yakkhas, like the more orthodox Brahmanical
deities had their sdnta and ugra forms. But even if these false and
lying Devas are Yakkhas, it need not be forgotten that their objection-
able qualities are emphasized in the interests of Jaina edification.
The Atanatiya Suttanta (Digha Nikaya, III 195 f.),' however,
speaks of good and bad Yakkhas, the latter being rebels to the Four
Great Kings (Kubera, etc.). If any of these assail a Buddhist monk
or layman, he is to appeal to the higher Yakkhas Vessavana himself ;
supplies to the Buddha the proper invocation, and gives a list of the
Yakkha chiefs; the list includes Ind(r)a, Soma, Varuna, Pajapati,*
Mani (-bhadda), Alavaka, etc. It will be observed that the first four
mentioned are orthodox Brahmanical deities ; but this is not the only
place in which Indra (Sakka) is spoken of as a Yakkha. Vessavana
(Kubera) goes on to say that there are Yakkhas of all ranks who do,
and others who do not believe in the Buddha, " But for the most part.
^
Lord, Yakkhas do not believe in the Exalted One."
Another list of Yaksas is to be found in the MahamdyfirV a work
which goes back to the third or fourth century A. D. In this list we
45, the Deva Samiddliisumana inhabits a idjayatana-

*
In Mahavai'iisa, i,
tree in the Jetavana garden at Savatthi he had been a man in Nagadlpa.

:
S. B. B., vol. 4 {Dialogues of the Buddha. 3). This text contains much vaki-
"
able information on Yakkhas.

' With Pajapati, cf. Prajapati, name of a Jogini, Pataini Devi temple (A.
S. I.,
A. R., Western Circle, 1920, p. 109).
* A similar distinction
of good and bad Yaksas is made in Mahavainsa, XXXI.
81, " Moreover, to ward off the evil Yakkhas the twenty-eight Yakkha chief-
tains stood keeping guard." The twenty-eight Yaksarajas are again referred to
in Lalita Vistara, Ch. XVI.
* Levi, S., Le catalogue gcographique de Yaksa dans Ic Mahamayfirl, ]. A.,
1915-
find Nandi ca Vardhanascaiva nagare Nandivardhane ,

" Nandi and
Vardhana, these twain, have their seat in the city of Nandivard-
hana " a Chinese commentator on the Avatamsaka Sutra has stated
;
that this city was in Magadha, as indeed the Sutra itself implies. All
this is of interest because two Yaksa statues (pi. 2, figs, i and 2)
have been found near Patna, and they bear inscriptions of which one
reads yakha ta vota naindi. The conclusion arrived at by Gangoly,
that the pair represent the tutelary Yaksas of Nandi vardhana may be
correct/ But the Mahdmayun list has also a Nandi Yaksa of Nandi-
nagara, separately mentioned. There are several Nandinagaras
known; one is frequently mentioned in the Saiici inscriptions. It
seems to me Patna figure designated as the Yaksa Nandi in

that the
the inscription may just as well be Nandi of Nandinagara as Nandi
of Nandivardhana this would leave the second statue unidentified,
;
as it is not named in the inscription. In the same list Manibhadra and

Purnabhadra are called brothers. Others mentioned include Visnu,
Karttikeya, Sahkara, Vibhisana, Krakucchanda, Suprabuddha, Dur-
yodhana, Arjuna, Naigamesa (tutelary Yaksa of Pancall), Makarad-
hvaja (=Kamadeva, the Buddhist Mara), and Vajrapani. The last
is said to be the Yaksa where is his

of Vulture's Peak, Rajagrha,
krtalaya (" made abode," evidently a temple) in the Yakkha Suit as ;
Sakka ( ? Indra), who is called a Yakkha of Mara's faction, may not

be the same as the Yaksa Vajrapani. Naigamesa is the well-known
antelope-headed genius, Indra's commander-in-chief, who both in
Brahmanical and Jaina mythology is connected with the procreation
of children.^
^
Gangoly, O. C, in 'Modern Review, Oct. 1919. Also Chanda, R., Four An-
cient Yaksa statues, Univ. of Calcutta, Anthropological Papers, 3 (Journ. Dep.
Letters, IV, Calcutta, 1921), and references there cited.
^
It will be seen that the list includes the names of orthodox Hindu deities,
Epic heroes, and others. Suprabuddha in Buddhist legend is the father-in-law

(rarely the grandfather) of the Buddha, and is one of the five persons who
suffered condign punishment for crimes committed against the Buddha or the
Order, one of the others being the Yaksa Nandaka. Krakucchanda is a former
Buddha.
Saiikara one of the well-known names of Siva, whose close connection with
is
Yaksas is in many ways, inter alia, by the existence of numerous tem-

shown
ples dedicated to him under names which are those of Yaksas, e. g., the Viru-
paksa temple at Pattadkadal. Siva's followers called Parisadas are huge-bellied
like Yak-sas. Cf. Hopkins, Epic Mythology, pp. 221-222.
For Naigamesa (ya) (Nejamesa, Naigameya, Harinegamesi) see Winternitz
in J. R. A. S., 189s, pp. 149 ff. Keith, Religion and Philosophy of the Veda,
;
p. 242. Naigamesa in the Epic is generally a goat-faced form of Agni. As

Harinegamesi he plays an important part in the conception and birth legend of
: " ;
NO. 6 YAKSAS — COOMARASWAMY I3
In Buddhist works the Yakkhas are sometimes represented as

teachers of good morals, and as guardian spirits.Thus in Thcra-
then-gatha, XLIV, Sanu Sutta/ Sanu had been the son of a YakkhinT
in aformer birth now this Yakkhini " controUing " (as SpirituaHsts
;
would say) Sanu, warns and advises his present human mother as
follows
Your son has a tendency to roam, wherefore bid him rouse himself. Tell him
what the Yakkhas say :
" Do nought of evil, open or concealed.

If evil thou doest or wilt do,
Thou shalt not escape from evil e'en though thou flee."
But more often, as in the Atandt'iyCi Sutfanta, the Yakkhas are said to
be unbelievers, to whom the ethics of the Buddhas are distasteful
they " haunt the lonely and remote recesses of the forest, where noise,
where sound, hardly is, where breezes from the pastures blow, hidden
from men, suitable for meditation. There do eminent Yakkhas dwell,
who have no faith in the word of the Exalted One."
In the Vijaya legend the aboriginal inhabitants of Ceylon are called
Yakkhas.'' One of Vijaya's men follows a bitch, who is the Yakkhini
Kuvanna in disguise ; she bewitches him, and all those who follow
him, but cannot devour them, as they are protected by charmed
threads. Vijaya follows, overcomes the YakkhitiT, and obtains the
release of the men; Kuvanna takes the form of a beautiful girl, and
Vijaya marries her (almost the Circe motif !). She enables him to
destroy the invisible Yakkhas who inhabit the land, and he becomes
Mahavira Kalpa Siltra, see Jacobi, S. B. E., XXII). In the Antagada

(in the
Dasao we him worshipped (Barnett, Antagada Dasao, p. 67, cited below,
find
p. 25). He is represented in an early relief from Mathura (Smith, Jaina stiipa
of Mathura, pi. XVII) with an inscription in which he is designated Bhagava

Nemeso also in some other early but mutilated reliefs in the Mathura
;
Museum., and regularly in the illustrations to the Jaina manuscripts of the

Kalpa Sutra (Coomaraswamy, Cat. Indian Collections, Museum of Fine Arts,
Boston, pt. IV).
Mara, and his hosts of deformed demons, is brilliantly represented at SancI,

north toraija, middle architrave, back (pi. 23, fig. 3). In a medieval relief
at Sarnath he is provided with a niakaradhvaja (Ann. Rep. Arch. Surv. India,
1904-05, p. 84) : as Kamadeva, with Rati, at Elura, in the Kailasa shrine, he
also has a niakaradhvaja.
'
Rhys Davids, Psalms of the Brethren, p. 48. Cf. ibid., p. 245, the older and
later attitude side by side, the Yakkha, though a cannibal, being invoked as the
guardian of a child.
' Dig ha Nikaya, III, 195 (Rhys Davids, Dialogues of the Buddha, Part 3, in
S. B. B., IV).
"
Mahavamsa, Ch. VII.
:
king. Later, he repudiates her and marries a human princess. She

returns to the Yakkhas, but is Her two children
killed as a traitress.
became the ancestors of the Pulinda (perhaps the Veddas, who are
still worshippers of Yakkhas; perhaps as ancestors?). In this story
the Yakkhas, though credited with supernatural powers, seem to be
regarded as aborigines themselves.
Not only may a human being be reborn as a Yaksa, but vice versa.'
A very interesting case of such a rebirth appears in the Indrakila
inscription, near Bezwada, of the ninth century. This inscription
occurs on a stele, sculptured with reliefs illustrating the Kiratarjuna
episode of the Mohabhdrata; the stele was set up by one Trikotti-
Boyu, who regarded himself as an incarnation of the friendly Yaksa
who at Indra's behest guided Arjuna to the inaccessible Indrakila hill,
there to wrestle with Siva and to receive the Pahipata astram. Extant
texts of the Epic do not mention any Yaksa, but some version of the
story must have known him, and Trikotti-Boyu regarded him as an
ancestor.^
3. YAKSAS AS TUTELARY DEITIES (PATRON SAINTS) AND

GUARDIAN ANGELS
In many cases Yaksas have been human beings attached to the
service of a community or individual, and, reborn as a spirit or geni,
continue to watch over those whom they had formerly served. Thus,
from a Tibetan source '
we get the following story connected with the
times of king Bimbisara, a contemporary of the Buddha
At and had been born
that time one of the gate-keepers of Vaisali liad died
again among the demons. He
gave the inhabitants of Vaisali the following
instructions " As I have been born again among the demons, confer on me the
:
position of a Yaksa and hang a bell round my neck. Whenever foe to the inhabi-
'
The doctrine of reincarnation is not Vedic, and in view of the suggestions
of indigenous origin that have been plausibly made, it is of interest to note how
constantly the idea of rebirth is connected with the Yaksa mythology, in which
a Yaksa may have been, or may again become a human being. Hodson, T. C,
The Primitive Culture of India, p. 7, and Lecture V, passim, shows that a
belief in reincarnation is widely spread amongst primitive tribes in India
(Khonds, Bhuiyas, Garos, etc.). The Lushais (p.- 105) desire to escape from
the mortal coil of reincarnation. Santals say that " good men enter into fruit-
trees " (Sir W. Hunter, Annals of Rural Bengal). According to a Buddhist
tradition Kuvera himself was once a very charitable Brahman (S. B. B., IV,
p. 193, note 4).
"
Sastri, H. K., The sculptured pillar on, the Indrakila hill at Bcsivada,
Ann. Rep. Arch. Surv. India, 1915-16.
"
Schiefner, A., Tibetan tales from the Kah-gyur (Ralston, p. 81).
5
;
NO. 6 YAKS AS — COOMARASWAMY 1
tants of Vaisali appears, I will make the bell sound until he is arrested or has
taken his departure." ' So they caused a Yak?a statue to be prepared and hung
a bell round its neck. Then they set it up in the gatehouse, provided with obla-
tions and garlands along with dance and song and to the sound of musical
instruments.
The same Tibetan sources show that the Sakyas honored a Yaksa
by name Sakyavardhana ("He who prospers the Sakyas") as a
tutelary deity. This tradition is recorded in the Tibetan Dulva f
we need not believe in the miracle, but there is every possibility that
there was a Sakya
tutelary Yaksa of
and tliat the Sakyas
the clan,
presented their children in the temple. Moreover, the Presentation is
four times illustrated at Amaravati (pi. 20, also Fergusson, Tree and
Serpent Worship, pis. LXIX, XCI, 4, and Burgess, Buddhist stiipas
. . frontispiece, detail left of center, and pi. XXXII, 2). Accord-
. .,
ing to the text,

was the habit of the Sakyas to make all new-born children bow down at
It
the feet of a statue of the Yaksa Sakyavardhana {Sakya-sphel or spcl) so ;
the king took the young child (the Bodhisattva, Siddhartha) to the temple, but
the Yaksa bowed down at his feet .... and when the king saw the Yaksa bow
down at the child's feet he exclaimed, " He is the god of gods," and the child
was therefore called Devatideva.
The same tradition is found in the Chinese Abhiniskramana Sutra

(the late sixth century Chinese version by Jfianakuti),' but the temple
is called a Deva temple, and the Deva's name is Tsang Chang, for
which the equivalent DIrghavardana is suggested. The story is much

more elaborated in the Lalita Vistara, Ch. VIII, where the temple is
full of statues of gods (Siva, Siirya), and all bow down to the child
this is obviously a later development.

In the Jaina Uttaradhyayana Sutra, Ch. Ill, 14-18, it is stated
as a general rule thatYaksas are reborn as men when their stock of
merit (acquired, of course, in a previous life on earth) is exhausted.
Not only human beings, but even animals may be reborn as tutelary
Yaksas. The following story of the Jaina saint JIvaka is related in the
Tamil classic, the Jwaka-cintamani : *
Jivaka rescues a drowning dog,
*
As regards the bell ; it should be observed that the voice of Devas and
Yaksas is often said to be like the sound of a golden bell (e.g., Samyutta
Ntkaya, Yakkha Suttas, §8 (Commentary), and Sakka Suttas, II, § 10 (Com-
mentary). For Yaksas with bells see plate 12, figure 2; plate 13, figure 3; and
plate 18. For a very similar story from the Divyavadana see Appendix.
^ Rockhill,
W. W., Life of the Buddha from Tibetan works in the Bkah-Hgyur
and Bstan-Hgyur, p. 17; Csoma de Koros, Analysis of the Kah-Gyiir, Asiatic
Researches, XX, p. 289. Cf. Watters, On Yuan Chwang, II, 13, 14, with some
other references, including Divyavadana.
^
Beal, S., Romantic history of Buddha, p. 52.
* Vinson, Lcgendes bouddhistes et djainas, Paris,
J., 1900, t. 2, p. 43.
2
l6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8o
or, to be more exact, recites to it the mantra of the Five Namaskaras,

whereby it is reborn as a deity, a chief of the Yaksas ; as such it is
called Sutanjana and lives in Candrodaya ("Moonrise") on the

white i\It. Sahga. Later, Jivaka is imprisoned by his enemies ; he calls
to mind Sutanjana, who immediately experiences a trembling which

brings Jivaka to hismind (cf. the heating or quaking of Indra's
throne when good men are in distress), and he hastens to the rescue.
He produces a great storm, and under cover of it carries off Jivaka
and takes him to his heavenly palace. Later, he bestows on Jivaka
three great spells (iiianfras) which bestow marvellous beauty, destroy
poison, and give the power of shape-shifting, and finally takes him
back to earth. There Jivaka erects and endows a temple and sets up
a statue in it.
A detailed story of Yakkhas is given in the Mahdvaiiisa, cha]iters

IX, X. It may be summarized as follows :
Prince Gamani had two attendants, Citta and Kalavela, respectively a herds-
man and a slave. He
love with the Princess Citta but it had been prophe-
fell in ;
would slay the Prince's uncles, who were then in power.

sied that the latter's son
However, the Princess became enceinte, and the marriage was permitted but it ;
was decided that if a son should be born, he should be put to death, and mean-
while Citta and Kalavela were executed for their part in the affair. " They were
reborn as Yakkhas, and both kept guard over the child in the mother's womb."
The child, a son, was duly born, and was called Pandukabhaya he was ex- ;
changed with the new-born daughter of another woman, and thus brought up in
safety away from the court (cf. the story of the infant Krsna). When the
young prince was once in sudden danger, the two Yakkhas appeared to save him.
Later on, Pandukabhaya captured a YakkhinI mare, described as valava-
rupa or valmrd-mukha, "mare-shaped" or " mare- faced " (cf. Assamukhl,
discussed below) her name was Cetiya, and she used to wander about the
;
Dhumarakkha mountain in the form of a mare, with a white body and red feet.
Pandukabhaya bored her nostrils and secured her with a rope she became his ;
adviser, and he rode her in battle. When at last established on the throne (in
.Anuradhapura), Pandukabhaya "settled the Yakkha Kalavela on the east side
of the city, the Yakkha Cittaraja at the lower end of the Abhaya tank. The
slave-woman who had helped him in time past (as foster-mother) and was
(now) reborn as (or of) a YakkhinI, the thankful (king) settled at the south
Within the royal precincts he housed the YakkhinI having the
gate of the city.
face of a mare. Year by year he had sacrificial offerings made to them and to
other (Yakkhas) but on festival days he sat with Cittaraja beside him on a
;
seat of equal height, and having gods and men to dance before him, the king
took his pleasure in joyous and merry wise With Cittaraja and Kalavela
who were visible,' the prince enjoyed his good fortune, he that had those that
^
had become Yakkhas for friends."
'
/. c, were represented by statues.
"
Alternativelv, "had Yakkhas and Bhutas for friends.'
NO. 6 YAKSAS COOMARASWAMY I7
4. SHRINES AND TEMPLES (CAITYA, AYATANA)

The haunt or abode {bhavanam) of a Yaksa, often referred to as
a caitya (Pali, cctiya, Prakrit, cciya) or ayatana (Prakrit, ayayana)
may be outside a city, in a grove, on a mountain or at a (//la.^ (shrines
of Punnabhadda and Moggara-pani those of the Indra's Peak ;
Yakkha, and the Yakkha Suciloma near Rajagaha mentioned in the

Saiiivutta NikCiya, Yakkha Suttas {Kindred Sayings, i, p. 264) and ;
the Yaksa shrine and image of Uttarddhyayana Sfitra, ch. XII,

S. B. E., XLV, p. 50, note), or by a tank (the Yakkha Cittaraja,
MahCivainsa, ch. X) ; or at the gates of a city (slave woman reborn
as a YakkhinI, Malulz'aiiisa, chapter X, and the tutelary Yaksa of
Wii.sfdT mentioned above) ; or within a city (shrine of Manibhadra,
KafhCisaritsCigara,, ch. XIII) or even within the palace precincts
(shrine of the YakkhinI Cetiya. Malidvaiiisa, ch. X). These shrines
are constantly spoken of as ancient, magnificent, famous, or world-
renowned.
The essential element of a Yaksa holystead is a stone table or altar
{vcvaddi, vianco) placed beneath the tree sacred to the Yaksa. The
hJiavanani of the Yakkha Suciloma at Gaya is particularly described
as a stone couch (better rendered as dais or altar) 1:)y or on which the
Buddha rested the words ; used are taiikita mailco, explained in the
commentary to mean a stone slab resting on four other stones {Sain-
yiitta Nikaya, Yakkha Suttas, ch. X, Kindred Sayings, i, p. 264). At
the Punnabhadda ceiya described below there were not only altars
(and probably an image) in an elaborate temple, but also a decorated
altar beneath an asoka-tree in the grove.
It was just such an altar beneath a sacred tree that served as the
Bodhisattva's seat on the night of the Great Enlightenment ; Sujata's
maidservant, indeed, mistakes the Bodhisattva for the tree-spirit him-
self {Niddnakatha). It is very evident that the sacred tree and altar
represent a combination taken over by Buddhism from older cults,
and in the case of the Bodhi-tree we see the transference actually in
process.
How often the hhavanas of the Yaksas mentioned in Buddhist and
Jaina' literature should be regarded as constructed temples it is hard
to say. Some, like the Punnabhadda ceiya. were certainly buildings,

independent of the altar beneath a sacred tree. In references to con-
structed temples supposed to have existed in the latter centuries pre-
ceding the Christian era there is nothing at all improbable ; some of
the dyafanas and caityas of the Epics are certainly buildings, and
sometimes contain statues. So. too. in Manu, 4, 39. The Candala
temple of Mahdhhdrata, 12. 121 (post-epical) has images and bells,
8 ;
and may have been a Yaksa shrine, or the shrine of a goddess.

Structural temple architecture was already far advanced in and before
the Kusana period.' The existence of images (and Yaksa images
are the oldest known images in India) in every case implies the ex-
istence of temples and a cult.
On the other hand it is quite certain that the word caitya some-
times means no more than a sacred tree, or a tree with an altar such ;
are designated caitya-vrksas in the Epics, and it is stated in the

Mahabharata, Southern Recension, 12, 69, 41 that such holy trees fit'.,
should not be injured inasmuch as they are the resorts of Devas,

Yaksas, Raksasas, etc. Even when as so often happens in Buddhist
the Buddha is represented as halting or resting at the
literature,
bhavanam of some Yakkha, it does not follow that a building is meant
the bhavanam may have been only a tree sacred to a Yaksa, and such
sacred trees are natural resting and meeting places in any village, as at
the present day. But in Safnyutta Nikdya, Yakkha Suttas, IV, it is
expressly stated that the bhavanam of the Yakkha Manibhadda was
called the Manimala caitya (the Jaina Silrya-praj napti says that the
Manibhadda ceiya lay to the northeast of the city of Mithila). As the
shrines of Manibhadda and Punnabhadda seem to have been the most
famous of all Yakkha shrines, it is most likely that the former as well
as the latter was a real temple, and indeed it is described as a temple
with doors and an inner chamber in K athdsaritsa gara, chapter XIII.
We know, too, that a statue of Manibhadda was set up at Pawaya,^
and this must have been housed in some kind of structure. Sakyavar-
dhana's shrine, too, in the Tibetan text and in one of the Amaravati
reliefs, is a temple: so also the krtdlaya of Vajrapani in the Maha-
mayuri list.
On the whole, then, we may be sure that in many cases Yaksa

shrines, however designated, were structural buildings. What were
they like? The passages cited in the present essay tell us of buildings
with doors, and arches {torane, which may refer either to gateways
like the Buddhist toranas, or, as the text has, " on its doorways,"
probably to stone or wooden pediments, with which we are familiar
from the Maurya period onwards) ;
"*
and of images and altars within
*Cf. HIIA, figs. 41, 43, 45, 62, 69, 70, 142: M. F. A., Bulletin, Nos. 144, 150:
Parmentier, L'Art khmcr primitif, p. 349, and Origine commune des architec-
tures dans I' hide et en Extreme-Orient, in fitudes Asiatiques.
See Garde, M. B., in Arch. Surv. India, Ann. Rep. 1914-15, Pt. I, p. 21 and
^
The site of Padmavalt, ib. 1915-16, p. 105 and PI. LVIl. See also p. 28.
* Cf. Smith, V. A., Jaina stupa
of Mathura, pis. XIX, XX; Coomaraswainy,
in M. F. A. Bulletin, No. 150 (August, 1927).
NO. 6 YAKSAS COOMARASVVAMY I9
the buildings. Indian styles of architecture, of course, are not sec-

tarian ; the style is that of the period. So that to discuss this question
fully would involve a discussion of all structural temple architecture
from the Maurya to the Kusana period inclusive; which would not
be altogether impossible, on the basis of literary references, and
representations in reliefs. This would take up too much of the space
at present available. But it may be observed that the Gujarat! com-
mentators gloss the word jakkhayayana by dyat thdnak deJiro, a little
domed temple/ This description would very aptly characterize the

little domed pavilions which are represented on Audumbara coins
from Kaiigra about the beginning of the Christian era, and on some-
what similar coins from Ceylon,' while a more elaborate structure of
the same type is seen in the Sudhamma Deva-sabha in the well-known
Bharhut relief (early second century B. C.).* Another example of a
" little domed temple " is the fire temple of the SancI relief, east torana,
left pillar, inner face, second panel. Cf also . HIIA, figure 145,
One of the detailed descriptions of a Yaksa holystead may be quoted
in full : this is the famous shrine of the Yak.sa Purnabhadra (Punna-
bhadda) of which a long account is given in the Aupapdtika Siitra*
Near Campa there was a sanctuary {cciya) named Punnabhadde. It was of

ancient origin, told of by nien of former days, old, renowned, rich, and well
known. It had umbrellas, banners, and bells ; it had flags, and flags upon flags
to adorn it, and was provided with brushes.^
^ Barnett, Antagada Dasao, p. 13, n. 5.

^ Audumbara J. A. S. B., LXVI, pt. I, 1897
coins. Smith, V. A., in Cunning- ;
ham, Coins of Ancient India, pi. IV, 2; HIIA, figs. 116, 117. Ceylon coins,
Pieris, P. E., Nagadlpa J. R. A. S., Ceylon Br., XXVII, No. 72, 1919.
,
^Cunningham, Siupa of Bharhut, pi. XVI; or HIIA, fig. 43.

* Leumann, E., Das Aupapatika
Sutra, erstes Upanga der Jaina, Abh. Kunde
des Morgenlandes, VIII, 2, 1883. The same account is implied in the Antagada
Dasao, the quotation above being taken from Barnett's rendering inserted in his
translation of the latter text.
The Jaina canonical works, like the Buddhist, may be regarded as good evi-
dence for the centuries immediately preceding the Christian era.
It may be remarked that Jaina ceiyas are distinguished (from those of Yak-
khas) as arhat ceiya.
^ Lo ma-hat tha: it seems to me that the rendering "brushes" may be due to
the translator's preoccupation with Jaina ideas.
Pali loma-hattha means "with hair erect" (horripilation)
in fear, astonish-
ment, or joy. May
not the suggestion be here simply " marvellous to behold,"
rather than the designation of an object? or could yak-tail fly-whisks {court),
more appropriate in a Yaksa shrine, have been meant?
It had daises (vcyaddi)^ 1)uilt in it, and was revercntiall)- adorned witli a
coating of dry cow-dung, and bore figures of the five-fingered hand painted in
go'sirsa sandal, fresh red sandal, and Dardara Sandal. There was in it great
store of ritual pitchers. On
(? beside, or above) its doorways were ritual jars
(vandaraghadc) and well-fashioned arches (toranc). Broad rounded long-
drooping masses of bunches of fresh sweet-smelling blossoms of the five colours
scattered therein. It smelt pleasantly with the glimmering reek of kalaguru,
fine kundurnkka, and iuriikka (incenses),'' and was odorous with sweet-smelling
fine scents, a very incense-wafer. It was haunted by actors, dancers, rope-
walkers, wrestlers, boxers, jesters, jumpers, reciters, ballad-singers, story-tellers,

pole dancers, picturei-showmen {mahkhc) i^
pipers, lute-players, and minstrels.
.... This sanctuary was encompassed round about by a great wood In
this wood was a broad mid-space. Therein, it is related, was a great and fine
Asoka-tree. It had its roots pure with kusa and vikusa grass Under-
neath this fine Asoka-tree, somewhat close to its trunk, was, it is related, a large
dais of ( ? resting upon) earthen blocks (pudhavlsild pattac). It (the dais)
^
an earthen or stone slab altar for the reception of offerings is the
I'eyaddi:
essential part of a shrine. Sometimes a symbol is placed on it. Later, when
images come into general use, it becomes the asana (seat or throne) or pit ha
(pedestal) of the figure. Altars are generally plain and smooth; but beautifully
ornamented examples are known, particularly one, Jaina, from the Kaiikali Tila,
Mathura (Smith, Jaina sttipa of Mathura, pi. XXII), and the outer vajrasana,
Buddhist, at Bodhgaya (Cunningham, Mahabodhi, pi. XIII), both of pre-
Kusana date.
In the Uvasagadasao, 164 (Hoernle, p. 107) the altar is called a masonry
platform {pudhavl-sUa-pattae ^
Sanskrit prthvi-sila-patfaka or pattaya, of.
the sild pattaain of the Malavikagnimitra, III, 79) ; Hoernle discusser the terms
at some length. Pudhavlsild might mean laterite. The words tahkitc-manco
"
are used in the Pali Yakkha Suttas, and rendered stone couch, but " altar
would be better.
^
The five fingered hand design is mentioned also elsewhere ; e. g., Mahavainsa,
XXXII, 4 (pancdngiilikd pantikd). Perhaps a five-foliate palmette would have
been thus designated.
* Picture-showmen ; probably those who exhibited scrolls (yainapata) illus-
trating the rewards of good and bad actions, to be realized in a future life. In
the Jaina Bhagavatl Siltra, XV, i, there is mentioned the heresiarch Gosale
Maiikhaliputte, whose second name refers" to his father's trade as a viankha
(cf. Hoernle, Uvasagadasdo, pp. 108, 121, notes 253, 273 and Appendix, p. i).
Patafijali, Malidbhdsya, in, 2, 11 1, refers to the exhibition of paintings of the
Krsna Lila, and to the use of the historical present in verbal explanation of
them; see Liiders, Sitz. k. Ak. Wiss., Berlin, 1916, pp. 698 ff. ; also Keith, A. B..
The Sanskrit drama (but Keith's rejection of the spoken explanation is probably
mistaken). In Visakhadatta's Mudrardksasa, Act. i, Canakya's spy adopts the

disguise of an exhibitor of yainapata (Prakrit, jamapadaam). Cf. tlie modern
Javanese Wayang Beber (Groeneveldt, W. P., Notes on the Malay Archipelago
and Malacca, compiled from Chinese sources, Batavia, 1876; Krom, N. J., in Ars
Asiatica, VIII, pi. LIX, and the similar Siamese exhibitions cited by Kram-
risch, Visnudharmottaram, Calcutta, 1924, p. 5, from the Siamese Sdratha
Pakdsini, pt. II, p. 398).
'
NO. 6 YAKSAS — COOMARASWAMV 21
was of goodly proportions as to breadth, length, and height and it was black ;
.... smooth and massive, eight-cornered, like the face of a mirror, very de-
lightful, and variously figured with wolves, bulls, horses, men, dolphins, birds,
snakes, elves, runi-deer, sarabha-deer, yak-oxen, elephants, forest creepers, and
padmaka creepers It was shaped like a throne, and was comforting
.... comely.
In those days, at that time, there arrived the reverent elder Subhamme
amidst a company of five hundred friars he travelled on and on, journeying in
pleasantness, he came to the city of Campa and the sanctuary Punnabhadde he
took a lodging such as was meet, and abode there. People came out from
Campa to hear his preaching.
The Anfagada Dasdo, chapter 6, in connection with the garland-

maker Ajjunae provides interesting details regarding the cult and
shrine of the Jakkha Moggara-pani. The following ahstract includes
all that is pertinent to our study :
Outside the city of Rayagihe (Rajagrha) Ajjunae possessed a beautiful

flower-garden. Some way from
garden there was a shrine (jakkhayayana)
this
sacred to the Jakkha Moggara-pani; this shrine "had belonged to Ajjunae's
grandfather, great-grandfather, and great-great-grandfather, and had passed
through a line of many men of his race " (by whom it had been supported in
past generations). "In it there stood a figure of the Jakkha Moggara-pani
holding a great iron mace a thousand palas in weight." Every morning, before
plying his trade, Ajjunae would go
garden with baskets and cloths to
to the
gather flowers then " with the chiefest and best flowers he would approach the
;
jakk'hciyayaita of the Jakkha Moggara-pani, fall upon his knees, and do rever-
ence." On a certain festival day he took with him his wife Bandhumai.
Meanwhile a certain gang of roughnecks from Rayagihe had made their way
Ajjunae and his wife, they plan
to the shrine to take their pleasure there; seeing
to bind him and take possession of her. To this end they hid themselves behind
the doors; when Ajjunae had made his offerings, they seized him as arranged,
and worked their will on his wife. Ajjunae reflected, " Verily I have been from
childhood a worshipper of my lord the Jakkha Moggara-pani now if the ;
Jakkha Moggara-pani were present here, could he behold me falling into such
ill-fortune? Then the Jakkha Moggara-pani is not present here: 'tis plain this
is but a log." Moggara-pani, however, became aware of Ajjunae's thoughts, and
took possession of his body having done so he seized the iron mace, and smote
;
down the six villains and the woman.

Ajjunae, still possessed by the Jakkha, now went about killing six men and a
woman everyday. The matter was brought to the king's notice. He proclaimed
that people should stay at home, and not go out of doors about their usual tasks.
A Jaina ascetic then arrived. Despite the king's orders and the danger, the pious
merchant Sudamsane cannot be dissuaded from going out to pay his respects to
the ascetic. The Jakkha meets and threatens him but Sudariisane, without fear, ;
immediately makes full profession of the monastic vows, and thus, as it were,
armed in the Lord, the Jakkha cannot approach him, but comes to a halt, staring
'
Translation by Barnett, 1907, p. 86. I have restored the original jakklia and
jakkhdyayatia in place of Harnett's " fairy " and " fairy-shrine."
: '
fixedly athim; then he abandons the body of Ajjunae, and returns to his own
place with the mace. Ajjunae falls to the ground, but on recovering himself,
accompanies Sudaiiisane and likewise takes the vows.
Here we find both the cult, patron-saint, and possession features

well displayed ; it is also clear that the Jakkha shrine is a building
with doors, and it is of interest to note that the statue is of wood,
and that it is provided with a club (cf, pi. 12, fig. 3). It is hardly
necessary to point out that the statue is not the Jakkha ; the latter
appears suddenly, and carries off the club with which the statue is
provided. The name Moggara-pani signifies, of course, " Club-
bearer." The antiquity of the shrine and simple nature of the cult
remain, and so, too, the fact that the worshipper regards the Jakkha
as his natural protector ; but the Jakkha is represented as a fierce
creature, without the sense to know when to stop —rather like the
giants of European fairy-tales. But he subdued by the new-

is easily
made Jaina monk ; and from the Jaina point of view the story is a
highly edifying one.
A characteristic and almost essential feature of Hindu and Buddhist
shrines an enclosing wall or railing {prakara, vcdika, etc.). The
is
following story related in the DJimmnapada Atthakathd (Burlingame,

E. W., Buddhist legends, H. O. S., Vol. 28, p. 146) refers to the
building of such an enclosure in the case of a tree worshipped with
desire for children
At Savatthl, we named Great-Wealth Maha-

are told, lived a householder
Suvanna. He was ample means
rich, possessed of great wealth, possessed of
of enjoyment, but at the same time he was childless. One day, as he was on his
way home from bathing at a ghat, he saw by the roadside a large forest tree
with spreading branches. Thought he, " This tree must be tenanted by a
powerful tree-spirit." So he caused the ground under the tree to be cleared,
the tree itself to be inclosed with a wall (p&kara), and sand to be spread within
the inclosure. And having decked the tree with flags and banners, he made
the following vow :
" Should I obtain a son or a daughter, I will pay you great
honor." Having so done, he went on his way.
Another story, in the Kah-gyur (Schiefner, Tibetan tales, IX) re-
lates how
a childless Brahman had recourse to the deity of a great nyagrodha-tree
(banyan), near the city called thence Nyagrodhika. He caused the ground
around it to be sprinkled, cleansed, and adorned. He then filled the space with
perfumes, flowers, and incense, and set up flags and standards. Then, after
having entertained eight hundred Brahmans and bestowed upon them material
for robes, he prayed to the tree-haunting deity, " Be pleased to bestow on me a
son." In case the request were granted, he would continue to offer the like
honors for a year, but if not, he would cut down the tree and burn it. The tree
deity, who was in favor with the Four Great Kings, betook himself to the
Maharaja Rastrapala, Virudhaka, Virupaksa, and Vaisravana; and the matter
was ultimately arranged by the aid of Sakra and Mahabrahma.
Another and later instance may be cited in the Mdlavikagnimitra,

V. I, where a bhitti-handho, or hhittivcdikahandha is built round an
asoka-tree.
Elaborate structures built round the Bodhi tree are represented in
numerous from Bharhut, SancT, Mathura, and AmaravatI,
reliefs
and there no reason to suppose that structures of this kind were
is
made for the first time after the Yakkha hhavanam (for such it was)
at Uruvela became the Bodhi tree of Gautama.
Yaksa caityas, etc., are constantly described as places of resort, and
suitable halting or resting places for travellers Buddhist and Jaina ;
saintsand monks are frequently introduced as resting or residing at

the haunt of such and such a Yaksa, or in such and such a Yakkha
ceiya (Punnabhadda ceiya, ut supra; the Buddha, in many of the
Yakkha Suttas of the Samyutta Nikaya). Amongst other caityas or
groves mentioned in Buddhist literature, the following may be cited
as having been in all probability sacred to the cult of a local divinity:
(i) the Capala caitya given to the Buddha by the Vajjians (Lic-
chavis) of Vaisall (Watters, On Yuan Chwang, II, 78) (2) the
Supatittha cetiya in the Yatthivana or Staffwood, where Buddha
stayed on his first visit ; it is stated, indeed, that this was the ancient
place of abode of Supatittha, the god of a banyan tree (Watters, ibid.,
II, 147), (3) the grove of sal-trees belonging to the Mallas, where the
Parinibbana took place. Here the couch {uttarasisakain) on which

the Buddha lay must have been a dais or altar originally intended for
the reception of offerings. In some reliefs, tree spirits are seen in
each of the two trees. (4) The Vajjian (Vaisali, Licchavi)
caityas referred to by the Buddha {Mahaparinihhana Suttanta, and
Angnttara Nikaya, VII, 19) when he repeats the conditions of future
welfare for the Vajjians, exhorting them not to allow the " proper
offerings and rites as formerly given and performed at the Vajjian
cetiyas to fall into desuetude." Buddhaghosa (Sumahgala Vildsinl)
regards these as having been Yakkha cetiya, and it can hardly be
doubted that this was so in most or all cases. With reference to the
Saradanda cetiya at Vaisall, where the Buddha was staying on the
occasion of stating the conditions of Vajjian welfare, he says that
" this was a vihdra erected on the site of a former shrine of the
Yakkha Saradanda."
In the same way Gujarat! commentators of Jaina texts interpret,
no doubt correctly, the ceiyas referred to, as Jakkha shrines. But the
Dijipalasa ceiya N. E. of the Vaniyagama suburb of Vaisall may be
separately mentioned. Here, in the Uvdsaga Dasdo^ § 2f., we find
'
Hoernle, UvdsagadasSo, II, p. 2.
;
Mahavira ^
in residence. The same ceiya is called a park {ujjdna) in
Vipaka Sutra, and elsewhere a ceiya of the Nala clan.
lect. i, § 2,
As Mahavira was a son of the chief of this Ksattriya clan, Hoernle
assumes that the ceiya must have heen sacred to the previous Jina
Parsvanatha. But even if we regard this Jina as historical, there
could have existed no Jaina cult {pfijd) in the time of Mahavira, and
it is much more likely that this was a Jakkha shrine or park. When,
further, the son of a pious householder of Vaniyagama takes the vows
of a lay adherent, and renounces willing offerings to " the Devas, or
objects of reverence to a heterodox community," it is probable that
Jakkha ceiyas are included. But here the commentory cites ceiya as
" idol," and mentions Virabhadra and Mahakala.
5. WORSHIP (PUJA) IN YAKSA SHRINES

Offerings to Yaksas, with a long list of other beings, are referred to
Grhya Sutras as being made at the close of Vedic studies
in several
the Sahkhayaim Sraiita Sutra, I, 11, 6, mentions Manibhadra. The
Asvalayana Grhya Sutra, I, 12, describes what is called a caitya-
(vandana) by householders. Hillebrandt,' followed by Keith,
ofifering
assumes that caityas erected as funeral monuments to teachers and
prophets are intended, butit is much more likely that the reference is
in themain to Yaksa caityas.

The Mahdbhdrata mentions that the flowers ofit'ered to Yaksas,
Gandharvas, and Nagas make glad the heart, hence they are called
suuianasas, eumenides ; such flowers being other than the sharp-
.scented. thorny and red flowers used in magical rites (Hopkins, Epic
Mythology, p. 68). The incense made from deodar and Vatica
^
As remarked by Hoernle the terms ceiya and ujjana, vana-sauda, vana-
khanda = grove or park, are interchangeable.
"
Rltual-Liicratnrc, Grundriss, III, 2, p. 86. It is quite possible that Hille-
brandt (like the author of the P. T. S. Pali Dictionary) ignores here the com-
mon meanings of caitya, other than funeral mound. I cannot help suspecting
too that when Keith {Religion and Philosopliy of the Veda, p. 73) remarks
that " Buddhist literature knows .... Yaksas who live in relic mounds," a
pre-occupation with the idea of funeral mounds (which are but one kind of
caitya) underlies the statement, which seems to be founded only on a misinter-
pretation of the collocation Yakkha-cetiya.
It is true that the word caitya is said to be derived from a root ci meaning
to build or heap up but as used
; in the Epics and early Buddhist and Jaina
literature, it means any holystead, altar, shrine, grove, temple, etc. May it not
be derived from cif, with the sense therefore of an object to be meditated upon
or attended to?
The Epic uses tlie word cdilka when Bauddha cetiyas (stupas) are specifically
meant; and in Jaina works, Jaina ceiyas arc distinguished as ArJiat ceiya.
robusta is liked by all deities but sallaklya incense is disliked by the

;
gods and suitable only for the Daityas. i\Iilk and flowers should be
offered to the gods, who The
take only the perfume of the latter.
appearance of flowers is Nagas use
acceptable to Raksasas, but the
them as food. On the other hand the food of Yaksas and Raksasas
is meat and spirituous liquor (Hopkins, ibid., pp. 68, 69). Here
again, as is generally the case, the Yaksas are given a spiritual rank
intermediate between that of the gods (Devas) and the lower spirits.
Manu (XI, 96) says that meat and intoxicating drinks are the food
of Yaksas, Raksasas and Pisacas. In the MeghadiJta, H, 3, Yaksas
are described as drinking wine produced from kalpa-tvees, in the
company of fair damsels cf the Bacchanalian Yaksa groups of
: .
Mathura (pi. 14, fig. i) and those of the ceiling of Cave I at Ajanta.
The prospector, before digging for treasure in Northern India,
makes offerings of meat, sesamum seeds, and flowers, to Kuvera,
Manibhadra, etc. {Mahdbhurafa, 14, 65, 11).
In connection with a YaksinT shrine at Rajagrha it is mentioned in
the Mahdbhdrata (3, 84, 105) that there was a daily service.
A passage omitted from the description of the Punnabhadde ceiya
cited above informs us that this sanctuary
was meet for the and supplications of many prayerful folk meet for
praj'crs ;
largesse, and respect

worship, celebration, veneration, offering, meet to be ;
waited upon with courtesy as a blessed and auspicious sanctuary of the gods,
divine, truth-telling, truth-counselling (or, surely satisfying the desires of its
worshippers). Miracles were manifested therein, and it received shares in thou-

sands of sacrifices. Many people came to worship the sanctuary Punnabhadde.
In the Antagada Dasdo, loc. cit., pp. 86 ff., the garland-maker

Ajjunae every day, before practising his trade, repairs to the temple
(jakkhayayanc) of the Yakkha Moggara-pani, with flower-offerings
of great worth, falls upon his knees, and does reverence.
Harinegamesi (see note on p. 12) is represented in the .Infagada

Dasdo (loc. cit.. p. 67) as receiving pnjd:
Sulasa was from childhood a worshipper of tlie god Harinegamesi. She
caused to be made an image of H., and every morning she bathed .... per-
formed the customary lustratory rites, and wîth a still moist robe made flower-
offerings of great worth, fell upon her knees, did reverence .... By the lady
Sulasa's devotion, veneration, and obedience the god H. was won over. So in
compassion for the lady Sulasa the god H., made both her and thee to become
pregnant at the same time.'
'
Here " thee "
refers to Queen Devai, whose living children are given to
Sulasa. when Queen Devai longs for children of her own, her husband
Later,
Kanhe (Krsna) Vasudeva worships Harinegamesi; the latter's throne quakes,
he looks down, and sees Vasudeva whose mind is fixed on him. He appears
to Vasudeva, " clad in robes of the five colours bearing hells," and promises
that Devai shall bear a child.
:
In the beautiful Jaina Tamil classic, the Jlvaka-cintdmani (Vinson,

J., Legendes houddhistcs et Djainas, Paris, 1900, t. 2, p. 43) the
grateful JIvaka erects a temple for the up a Yakkha Sutaiijana, sets
Statue, and dedicates a town (the rents whereof would support the
service of the temple) then he has prepared a drama relating to the
;
history of the Yaksa, and most likely we should understand that this
drama was presented in the temple on special occasions for the
pleasure of the deity.
The tutelary Yaksa at Vaisali, as we have seen, was worshipped
with oblations, dance and song, and the sound of musical instruments.
Later books appear to show that Yaksa worship and some par-
ticular Yaksas retained their prestige throughout the medieval period.
In these texts we find a cult of the same general character, and can
glean some further details. In the Kathasaritsagara, part I, chapter
XIII, we find:
" In our country, within the city, there is the shrine of a powerful
Yaksa named Manibhadra, established by our ancestors. The people
there come and make petitions at this shrine, offering various gifts,
in order to obtain various blessings." Offerings (of food) are re-
ferred to, which it was the duty of the officiating priest to receive and
eat. The anecdote turns upon the interesting fact that the Yaksa
temple was regularly used as a temporary jail for adulterers.
Numerous other and incidental references to Yaksas and Yaksinis
will be found in the same work, passim (e. g., in ch. XXXIV, story of
the Yaksa Virupaksa).
The equally late Parisistaparvan of Hemacandra (thirteenth cen-
tury) Canto 3, has a story of two old women, Buddhi and Siddhi
" Buddhi had for a long time continued to sacrifice to a Yaksa, Bhola
(or Bholaka), when the god, pleased with her devotion, promised her
whatever she should ask," etc. A
little further on we find a human
being, Lalitariga, " disguised as a statue of a Yaksa." ' The same

text, Canto 2, eighth story, describes an ordeal undergone by a woman
justly accused of adultery. " Now there was a statue of the Yaksa
Sobhana of such sanctity that no guilty person could pass through
between its legs." The lady (like Guinevere in a similar predicament)
frames an oath which is literally true but essentially false. " While
the puzzled Yaksa was still at a loss to know how to act," she passed
through his legs.
Devendra, in the Utiaradhyayana tlkd (
Jacobi, p. 39, Meyer,Hindu
tales, p. 140), Story of Domuha, tells of a lady named Gunamala
'Jacobi, H., Sthaviravali Charitra, Bib. Ind., CalcuUa, 1891, pp. 33, 37.
NO. 6 VAKsAS — COOMARASWAMY 2rj
who was unhappy because she had no daughter. And she vowed an
"
oblation (nvdiyam) to the Yaksa called Mayana .... a daughter

was born of her She gave the oblation to the Yaksa."
In the Prabandhacintamani, another Jaina story book, about 1419
A. D., we find a Yaksa by name Kapardin invoked by a Jaina layman,
acting on the advice of his Guru.* The Yaksa bestows wealth on his
supplicant, and then relates the circumstances to his sons, " in order to
manifest in their hearts the power of religion"; the Yaksa himself
is a worshipper of the Jina. It is clear that Jainism and Yaksa wor-
ship could be as closely interrelated as Buddhism and Hinduism have
often been.
Rites for attracting YaksTs are mentioned in the Kathasaritsagara,
chapter XLIX. These rites are performed in cemeteries, and are
evidently Tantrik. The beautiful Yaksis Vidyunmala, Candralekha.
and Sulocana are said to be the best among them. A certain Adityasar-
man, living in Ujjayinl, obtains the last as his wife, and lives with
her in Alaka their son Gunasarman is sent back to the human world,
;
and becomes a great king.
6. YAKSA WORSHIP A BHAKTI CULT

The reader cannot fail to have observed that the facts of Yaksa
worship summarized above are almost identical with those charac-
teristic of other and contemporary Bhakti (devotional) cults. It is,
in fact, a great error to assume that the term Bhagavat {" worship-
ful ") applies only to Visnu, and Bhaktd {" devout worshipper ") only
to worshippers of Visnu.* The rise, or, as it would be better to say,
the coming into prominence of Bhakti cults in the centuries immedi-
ately preceding the beginning of the Christian era was not an isolated
sectarian development, but a general tendency. All forms of belief
were involved, Buddhism no less than others.'
Not only is Vasudeva (Visnu) styled Bhagavat, but also the Four
Great Kings, the Maharajas, Regents of the Quarters, amongst whom
^Prabandhacintamani, Trans, by C. H. Tawney, London, 1901, p. 203.

^
As might be gathered from Bhandarkar, R. G., Vaisnavisin, Saivism, and
minor religious systems (Grundriss indo-arische Ph. und A.).
* For the Bhakti character of even early Buddhism, see De la Vallee-Poussin,
loc. cit. pp. 334 ff. The Majjhima Nikaya, i, 142, has " He who has faith
(srdddha) in Me and love (prema) for Me will attain to heaven." So too
Saivism, " Even after committing all crimes, men by mental worship of Siva
are freed from sin" (Mahabharata, 13, 18, 65). Both assurances are altogether
in the spirit of the Bhagavad Glta.
is Kubera, Regent of the North, himself a Yaksa' (and, as Vais-

ravana. frequently styled Bhagavat in the Mahdhharata), a Naga,'
and the Buddha himself.'' The Pawaya image of the great Yaksa
Manibhadra has a dedicatory inscription,^ in which the deity himself
is styled Bhagava and the members of the gostha (corporation) for
whom the image was set up speak of themselves as Mdnihhadra-

bhaktas. Nêmesa, too, is called Bhagava (Mathura inscription al-
ready cited). Thus, both the designation Bhagavat and the use of the
term Bhakti are seen to be common to most, as they probably were
to all of the contemporary faiths."
Apart from these questions of terminology it will be evident that

the facts of Yaksa worship correspond almost exactly with those of
other Bhakti religions. In fact, the use of images in temples, the prac-
tice of prostration, the offering of flowers (the typical gift, constantly
mentioned), incense, food, and cloths, the use of bells, the singing of
hymns, the presentation of a drama dealing with the Llla of the
Hindu worship even at the present
deity, all these are characteristic of
day." Only the nature of the food is peculiar, and this may be attrib-
uted to the relationship of Yaksas with Raksasas nor will it be for- ;
gotten that animal sacrifices and the use of strong liquors still per-
sists in some Sakta cults. Nothing of this cult type is to be found in
the Vedas.
7. YAKSA SOURCES IN BUDDHIST ICONOGRAPHY

Yaksas, as we have seen, may be represented by independent cult
images, or in connection with other sectarian systems, as attendants.
*
Panini, IV, 3, 97, speaks of Bhakti directed towards Maharajas, not in a
political sense, but with reference to the Four Great Kings (see Bhusari
in Ann. Bhandarkar Inst., VIII, 1926, p. 199). For Manibhadra as a Lokapala
see Vogel, Indian Serpent Lore, p. 10.
'"
The Naga Dadhikarna, in an inscription at Mathura, Liiders' list, No. 85.
^ Already at Bharhut, in the inscription Bhagavato Saka Munino Bodho, and
on the Piprahwa vase, Bhagavato sakiyamuni.
*
Garde, M. B.. The site of Padiinidvafh A. S. I., A. R., 1915-16, Gangoly,
O. C. in Modern Review, Oct. 1919; Fouchcr, in J. B. O. R. S. Chanda, Four ;
ancient Y'aksa statues. Text of the Brahmi inscription: .... gaiisfhya Maui-
bhadrabhaktagarbhasukhitah Manibhadrasya pratima. pratisthapayamti
"
For the meaning of Bhaga7)at, " Adorable," " Blessed," " Worshipful," etc.,
see Grierson, The translation of term Bhagavat, J. R. A. S., 1910;

the
Schrader, ibid., 191 1, p. 194; Hopkins, Epic use of Bhagavat and bJiakti, ibid.,
1912; Govindacharya Svainin, ibid., p. 483.
For an admirable account of the daily office in a modern temple,

" see
(Burgess, J.), The ritual of Ramcsvarani. Indian Antiquary, XII, 1883.
NO. 6 . YAKSAS COOMARASWAMY 29
guardians, and worshippers. But not only have both classes of figures
their own intrinsic and aesthetic interest (pi. i, fig. i, and pi. 8, for
example, are magnificent works), they are also of importance as fac-
tors in the development of Indian iconography generally. The force
of tradition is strong, and Indian art like other arts has always by
preference made use of existing types, rather than invented or
adopted wholly new ones. The case is exactly parallel to that of
religious development, in which the past always survives. We have
to do with a conscious sectarian adaptation, accompanied by an uncon-
scious, or at least unintentional, stylistic evolution.
In early Indian art, so far as cult images are concerned, one icono-
graphic type stands out predominant, that is the standing figure with
the right hand Sometimes the right hand
raised, the left on the hip.
holds a flower, or caurT, or weapon sometimes the left grasps the ;
robe, or holds a flask, but the position of the arms is constant. We are
here, of course, concerned only with two-armed images ; those with
four or more arms do not appear before the second century A. D.,
when the fundamentals had already been established. Stylistically, the
type is massive and voluminous, and altogether plastically conceived,
not bounded by outlines ;
the essential quality is one of energy, with-
out introspection or spiritual aspiration.
Of this type are the early images of Yaksas, and Yaksis, whether
independent or attendant. And it is also this type which provided the
model for the cult images of other deities, such as Siva or Buddha,
when the necessities of Bhakti determined the appearance of all dei-
ties in visible forms.

Making only a passing reference to the close formal relationship
recognizable between the oldest known Siva image, that of the Gudi-
mallam lihgam (pi. 17, fig. i), and the Yaksas of Bharhut and Saiici,
and to the facts that the Nyagrodha, Udambara, or Asvattha tree may
be identified with \'isnu, and that Siva, Saiiikara, Karttikeya, etc..
are all Yaksas in the Mahamayuri list, I propose to speak here only
of the part played by the Yak.sa type in evolution of Buddhist types.
In the case of the Buddha figure, as I have recently treated the sub-
ject at length in the Art Bulletin (Vol. IX, pt. 4), I shall only point
out the stylistic continuity presented in the series Parkham image :
(pi. I, fig. I ) one of the Yaksas from Patna (HIIA, fig. 67) Buddha
; ;
in the Lucknow Museum (HIIA, fig. 79) Bodhisattva in Philadelphia ;
(Art Bull., loc. cit., fig. 50) ; Friar Bala's image at Sarnath (pi. 17,
fig. 2) ; Gupta image in the Mathura Museum (HIIA, fig. 158). In
such a series the relationships are very evident, and there is no room for
the insertion of any Hellenistic type.
The Bodhisattvas Padmapani, Vajrapani and Maitreya may be dis-
cussed in greater detail.
The earliest Buddha triads are represented, as in plate 9, by a

Bodhi-tree supported by two Yaksas, each with an expanded rose-
lotus (padnia) in hand, or by a symbol (the wheel) between similar
Yaksas with a caitrl (pi. 10, fig. i). Yaksas with a lotus in hand
appear as guardian figures {dvarapalas) at Sahci (pi. 8) and else-
where (pi. 7). Now, a Yaksa with a padma in hand can only be
described adjectivally as padma-pani; can it be doubted that the
Bodhisattva Padmapani (a form or designation of Avalokitesvara),
whom we find a little later attendant on the Buddha or as an inde-
pendent Buddhist deity, is the same historically and iconographically,
as padma-pani Yaksa of the earlier sculpture? The caur'i-
the
bearing Yaksas (HIIA, figs. 84, and 85 right), too, are the same as
those of the earlier compositions, but we cannot as a rule give them
a name.
The case of Vajrapani is more involved.^ The one obvious vajra-
pani of Indian mythology whose weapon is the thunderbolt
is Indra,
already in the Vedas. In Buddhist mythology Indra is known as
Sakka (San. Sakra), and he plays a conspicuous part in the Buddhist
legend visiting or aiding the Buddha on various occasions." Buddha-
ghosa ^ tells us that Vajrapani is the same as Sakka and Sakka, upon ;
occasion (Yakkha Suttas, 2) may be called a Yakkha. But Sakka is

never himself a Bodhisattva.
On the other hand Vajrapani, independently of Indra, is called a
Yaksa in the Mahamayurl list, where he is said to be the Yaksa of
Vulture's Peak, Rajagrha (thework krtdlaya seems to imply that
therewas a temple). A Tibetan version of the Vinaya speaks of a
Yaksa Vajrapani (Gnod-sbyin Lag-na-rdo-rje). And in the Lalita
Vistara, XV, 66, we have a " benevolent lord of the Guhyakas,
^ For Vajrapani in addition to references cited below, see also Vogel, Le

Vajrapani greco-hoiiddhique, B. E. F. E. O. XI, 191 1, p. 525, where it is ob-
served that Vajrapani and Indra are not necessarily always one and the same
persons. M. Foucher has already fully established the Yaksa origin of the
Bodhisattva Vajrapani (L'Art greco-bouddhiqiie ., II, pp. 48-64). See also
. . .
Senart, E., Vajrapani dans les sculptures du gandhara, Congr. Int. Orientalistes,
14, Alger, 1905, pp. 111-131.
" For a full and valuable discussion of Indra as Sakka, see Mrs. Rhys Davids,
Introduction to the Sakka-pahha Suttanta, SBB., III., p. 294.
Waddell, Evolution of the Buddha cult, p. 118, citing Csoma de Koros,

Analysis of the Dulva, Asiatic Researches, XXj 64.
^Commentary on the Amba{tha Sutto, cited SBB, II, 117.
NO. 6 YAKSAS — COOMARASWAMY 3I
Vajrapani " who appears in the air on the occasion of the Abhinis-
kramana (Going Forth of the Buddha), and who, as remarked by
Foucher, " desormais le quittera pas plus que son ombre," becoming,
Buddha's guardian angel/ This Vajrapani is not the same
in fact, the
as Sakka, who is independently present on the same occasion.
This Vajrapani is constantly represented in Gandharan reliefs, and
sometimes in those of Mathura, illustrating scenes from the Life, sub-
sequent to the Going Forth, c. g., Foucher, loc. cit., figs. 191, 195, 197,
199. At his first appearance he is called a " benevolent Lord of the

Guhyakas, vajra in hand." Sometimes he holds a caxiri as well as a
vajra; moreover, this Vajrapani is generally represented as nude to
the waist and without any turban or crown, thus not as a great king,
as Indra should be. Moreover, this Vajrapani and Sakka are often
present together in one and the same scene (pi. 21, fig. 2).
Perhaps the earliest appearance of a Vajrapani in a Buddha triad
may be the example in the Museum of Fine Arts, Boston (HIIA,
fig. 85, left) and here we are in doubt whether to call him Yaksa or
;
Bodhisattva. It may be doubted whether the Bodhisattva Vajrapani

had been recognized so early. The only early independent image which
may be a representation of the Vajrapani, who is not Indra, is a frag-
ment from Mathura, illustrated in plate 15, figure 2.^
Thus there was actually a Yaksa Vajrapani, not identical with
Indra, but having an independent, pre-Buddhist cult ; this Yaksa be-
came the Buddha's guardian angel and attendant, and finally came to
be called the Bodhisattva Vajrapani, who sometimes appears in
Buddha triads, and is sometimes the object of separate worship
(HIIA, fig. 299).
As regards Maitreya, the earliest of the Bodhisattvas to be desig-
nated as such, there is less to say. His cliaracteristic emblem is the
auirta (" nectar ") fiask, held in the left hand. It will perhaps occur
to the mind of the reader that there are both Bacchanalian Yaksas,
and Bachhanalian Nagas, who hold a cup or flask in their hands and ;
as in verbal imagery nothing is more characteristic of Buddhism

than the reinterpretation of an old phrase in the interests of present
edification (cf. Lalita Vistara, VII, 91, "with the Water of Life
(amrta) shalt thou heal the suffering due to the corruption of our
mortal nature"), so here, perhaps, we have a literal example of the
pouring of new wine into old bottles.
'Foucher, L'Art greco-boiiddhique du Gandhara, I, 368: and cf. ibid., II,
pp. 48-64.
' Vogel, The Mathura school of sculpture, A. S. I., A. R., 1909-10, p. 76 and
pi. XXVIIIb.
3
8. WOMAN AND TREE MOTIF

Enough has been said in the course of the present article, or will be
found in the accompanying illustrations, to indicate the intimate con-
nection subsisting between spirits and trees/ For the. rest it will
suffice in the present connection to recall the Epic passage, " god-
desses born in trees, to be worshipped by those desiring children,"
such goddesses being designated as dryads (Vrksaka, Vrddhika).
There is no motif more fundamentally characteristic of Indian art
from first to last than is that of the Woman and Tree. In early
sculptures (reliefs on pillars of ' gateways and railings at Bharhut,
Bodhgaya, Sand, and Mathura) the female figures associated with
trees are voluptuous beauties, scantily clothed, and almost nude, but
always provided with the broad jewelled belt (mekhala) which ap-
pears already on the pre-Maurya terra-cotta figures of fertility god-
and which the Atharva Veda (6, 133) tells us was a long-life
desses,'
(ayiisya) charm. Sometimes these dryads stand on a vehicle {vaha-

nam) such as a Yaksa (Guhya), elephant, or crocodile (makara).
Sometimes they are adorning themselves with jewels, or using a
mirror. Very often they hold with one hand a branch of the tree
under which they stand, sometimes one leg is twined round the stem
of the tree (an erotic conception, for lata is both " creeper " or
"vine," and "woman," and cf. Atharva Veda, VI, 8, i, "As the
creeper embraces the tree on all sides, so do thou embrace me").
Sometimes one foot is raised and rests against the trunk of the tree.
Sometimes there are children, either standing beside the dryad
mother, or carried astraddle on her hip. Of the trees represented the
^
For pre- and non-Buddhist trees, tree-spirits, and sacred groves generally,
see Hopkins, Epic Mythology, p. 6 f., and Keith, Religions and Philosophy of
the Veda, pp. 184, 185. Trees and tree-deities play but an insignificant part in the
Rg Veda and even in the Atharva Veda (Macdonnell, Vedic Mythology, p. 154)
but even here they are connected with human life and productivity the beings ;
inhabiting trees being called Gandharvas and Apsarases. The Athama Veda,
of course, contains many elements incorporated from aboriginal non-Aryan
sources. It is perhaps also significant (in view of possible Sumero-Dravidian
connections) that in Babylonian tradition immortality and productiveness are
original functions of the tree of Fortune (Ward, Seal Cylinders of Western
Asia, pp. 233, 237, etc.).
opiate 4, fig. 2; pi. s; pi. 6, figs. 2, 3; pi. 11, figs, i, 2, 3: pi. 14, fig. 2;
pi. 19; pi. 22, figs. I, 2.
'Also the so-called Earth goddess of Lauriya-Nandangarh (HIIA, fig. 105) :
this nude goddess, who is represented also in very early terracottas (see M. F. A.
Bulletin, No. 152), may not be a Yak§T.
NO. 6 .
YAKSAS — COOMARASWAMY 33
asoka and mango are most usual. At first sight, these figures seem to
be singularly out of place if regarded with the eyes of a Buddhist or

Jaina monk.^ But by the time that a necessity had arisen for the
erection of these great monuments, with their illustration of Buddhist
legends and other material constituting a veritable Bihlia Pauperum,
Buddhism and Jainism had passed beyond the circle of monasticism,
and become popular religions with a cult. These figures of fertility
spirits are present here because the people are here. Women, accus-
tomed to invoke the blessings of a tree spirit, would approach the rail-
ing pillar images with similar expectations ; these images, like those
of Nagas and Yaksas often set up on Buddhist and Jaina sites, may
be compared to the altars of patron saints which a pious Catholic
visits with prayers for material blessings.
From these types of YaksT dryads ' are evidently derived three
types iconographically the same, but differently interpreted : the
Buddha Nativity, the asoka-tree dohada motif in classical literature,
and the so-called river-goddesses of medieval shrines.
^
The array of dryads at Mathura produces on the mind an effect like that of
Asvaghosa's description of the beautiful girls in Siddhartha's palace garden,
who " with their souls carried away by love .... assailed the prince with all
manner of stratagems" (Buddhacarita, IV, 40-53).
But it may be said to be characteristic of Indian temples that the exterior
displays the world of sensuous experience (cf. Konarak), while the interior
chambers are plain and severe, or even empty (cf. the air-liiigam at Cidam-
baram) and this arrangement, even for a Buddhist shrine, is not without its
:
logic.
I have scarcely mentioned and have not illustrated the many interesting
reliefs and paintings in which tree spirits are represented, not by a complete
figure beneath a tree, but as half seen amongst the leaves, patresv ardhakaydn
abhinirmaya (Lalita Vistara) a face, hand, two hands, or half body emerging
:
from the branches. Representations of this kind occur already at Bharhut, and
survive in the eighteenth century Buddhist painting of Ceylon. The spirits
thus represented may be male or female as the case requires.

'
That the Vrksakas of the railing pillars are properly to be described as
Yaksis is proved by the inscriptions accompanying the similar figures at Bharhut
(cf. Vogel, in A. S. I., A. R., 1906-07, p. 146). Vrksaka is, of course, legiti-
mate, but hardly more than a descriptive term. Some with musical instruments
should perhaps be described as Gandharvis, or even Apsarasas, but none are
represented as actually dancing, and to call them dancing girls is certainly an
error.
Hoysala bracket figures, however, which preserve the motif of woman and
tree,supported by a dwarf Yaksa, are often in dancing positions, and accom-
panied by drummers (Smith, H. F. A., fig. 163; others at Palampet and Beliir).
I. The miraculous birth of the Bodhisattva Siddhartha/ as is well-

known, took place in the Lumbini garden near Kapilavastu and on
the road between that city and Devadaha. The tree of which the
branch, " bending down in response to her need," served Mahamaya
as support, is variously called a sal-tree (Niddnakatha) , mango
(Asokdvaddna) ,
plaksa {Lalita VistaraY and asoka-tree {Divyd-
vaddna, and here plate 20) In the Divydvaddna Asoka himself
. is repre-
sented as visiting the site and conversing with the genius of the tree,
who had been a witness to the Nativity ; so that the tree had originally
been, or at least had come to be regarded as having been the abode of
a tree-spirit when Mahamaya halted beneath it. It is, no doubt, the
spirit of the tree that bent down the branch to meet Mahamaya's
hand ; indeed, in the drawing of a relief almost identical with our
plate 20, reproduced in Burgess, Buddhist stupas of Amaravati and
Jaggayyapcta, plate XXXII, a hand appears visibly from amongst the
branches of the Nativity tree. The Buddha himself is sometimes aided
in just this way, by a hand put forth from a tree, for example, when
he emerges from the waters of Lake Panihata {Lalita I'istara,
Ch. XVIII), and after crossing the River Nairanjana (Amaravati
relief, Vogel, Indian serpent lore, pi. VII, a).
We and should not regard Mahamaya, consid-
certainly need not
ered from the point of view of the literature, as having been herself
a Vrksaka; but iconographically, as she is represented in Gandharan
^
The Nativity is a stock subject in Buddhist art, Gandharan, Amaravati, and
later. Cf. Foucher, Beginnings of Buddhist Art, pis. Ill, IV; L'Art greco-
bouddhique du Gandhara, I, pp. 300 ff. and II, pp. 64-72 ; L' I cono graphic houd-
dhique de I'Inde,I, p. 163 and fig. 28: HIIA, fig. 104, upper right hand corner:
Krom, Life of the Buddha, p. 74 (with complete list of representations).

The Amaravati reliefs not only come nearest to the Vrksaka type, but also
suggests that the Nativity had been represented in Indian art (without the
child) previous to its occurrence in Gandhara (with the child).
Another version of much interest appears at the back of a Chinese Buddha
image of date 457 A. D. (Northern Wei) (Burlington Magazine Monograph
on Chinese art. Sculpture, PI. 4, D). There are two ranges; above we
have the tree, female attendant, Maya standing, the child emerging from her
and three Devas, one with a cloth, ready to receive it below, the First Bath
side, ;
and the Seven Steps. As the First Bath is here performed by polycephalous
Nagas, which are rarely met with in Gandhara, but are highly characteristic for
Mathura, there is a probability of direct dependence on an Indian original.
'
In the Lalita Vistara version, the tree is evidently regarded as a caitya-tree,
for it is adorned with coloured cloths and other offerings.
and Amaravati reliefs and elsewhere/ the step is very easy from a
Vrksaka holding the branch of a tree and in the hanche ("hip-
shot ") pose, to that of Mahamaya giving birth to the child, who
was miraculously born from her side.^ The addition of attendant
deities and later a further complication of the scene by a repre-
sentation of the Seven Steps, etc., would present no difficulty. The
literary versions are probably older that the oldest known sculptures
of the Nativity
;
' how far each may be dependent on the other can
hardly be determined. In any case, it is certain that the sculptor had
ready to hand a composition almost exactly fulfilling the require-
ments of the text, so far as the principal figure is concerned.
2. The dohada motif. The, in India, familiar conceit that the

touch of a beautiful woman's foot is needed to bring about the blos-
soming of the asoka-tree seems to be equally a form of the YaksT-

dryad theme one railing pillar, J 55 in the Mathura Museum, repre-
;
sents a woman or Yaksi performing this ceremony* (pi. 6, fig. 3) and

the motif survives in sculpture to the eighteenth century (pi. 19.
fig. 2), if not to the present day. In Kalidasa's Meghaduta the exiled
Yaksa speaks of himself as longing for his wife no less than the
asoka-tree desires the touch of her foot. Even in the Malavikagni-
* The formula was certainly not, as suggested by Foucher, L'Iconographie

bouddhique, I, 164, created " par I'art superieur des artistes Indogrecs " ; it is
only possible that they were the first to put in the attendant figures, but we can-
not be sure of even this. Even the crossed legs, described by many European
writers, grotesquely enough, as a dancing position, are taken over from the
Yaksi-dryads. Le Coq, Bilder- Atlas, figs. 153 and 156 not only describes Maha-
maya as being in " Tanzerinnenstellung," but also a dryad from Bharhut, who
with both arms and one leg is clinging to her tree, while her weight is rested on
the other foot dance under either of these circumstances would
(pi. 4, fig. 2) ; to
not only be a remarkable acrobatic feat, but in direct contradiction to the whole
pose. To stand with crossed legs, particularly when leaning against a tree, is in
India a position of rest and therefore not inappropriate (as a dancing pose
would be) to the representation of a miraculously painless parturition.
The motif has been well discussed (with reference to this and other misunder-
standings) by Berstl, Indo-koptische Kunst, Jahrb. as. Kunst, I, 1924; where a
Western migration of the motif is also recognized.

^ It is perhaps worth remarking that Cunningham once " erroneously
identi-
fied " one of the Mathura railing dryads " with Maya standing under the sal
tree " (Vogel, Cat. Arch. Mus., Mathura, p. 6).
'The legend of the miraculous birth is found already in the Acchariyabbhuta
Sutta, No. Majjhima Nikaya, thus considerably antedating
123, in the the
Nidanakatha version (Chalmers, in J. R. A. S., 1894). The Four Devas are
mentioned.
* Vogel, Catalogue, pp. 44, 153; La belle et I'arbre asoka, B. E. F. E. O., XI,
191 1 Cf. [Gangoly, O. C], A brass statuette from Mathura, Rupam, 2, 1920.
;
:
mitra, where Malavika, a mortal woman, is to perform the ceremony,

the scene takes place beside a " slab of rock " under the asoka-tree,
and this shows that the tree itself was a sacred tree haunted by a
spirit/
The word dohada means a pregnancy longing, and the tree is repre-
sented as feeling, like a woman, such a longing, nor can its flowers
open until it is satisfied. Thus the whole conception, even in its latest
form as a mere piece of rhetoric, preserves the old connection be-
tween trees and tree and human life. spirits,
3. The The dryad types with makara vehicles

River-goddesses.^
(pi. 6, figs. I and 2, pi. 14, fig. 2, and pi. 19, figs, i and 2) bear
an intimate relation, not amounting to identity, with the figures of

river-goddesses Gaiiga and Yamuna, with makara and tortoise vehicles
placed at the doorways of many northern medieval temples. I pro-
pose to discuss this subject more fully elsewhere.
CONCLUSION
The observations collected in the foregoing pages may be sum-
marized as follows
Kuvera and other Yaksas are indigenous non-Aryan deities or
genii, usually beneficent powers of wealth and fertility. Before
Buddhism and Jainism, they with a corresponding cosmology of the
Four or Eight Quarters of the Universe, had been accepted as ortho-
dox in Brahmanical theology. Their worship long survived, but in
purely sectarian literature they appear only to serve the ends of edifi-
cation, either as guardians and defenders of the faith, or to be pointed

to as horrible examples of depravity.
Yaksa worship was a Bhakti cult, with images, temples, altars, and
offerings, and as the greater deities could all, from a popular point of
view, be regarded as Yaksas, we may safely recognize in the worship
Nagas and goddesses) the natural source
of the latter (together with
of the Bhakti elements common
to the whole sectarian development
which was taking place before the beginning of the Kusana period.
The designation Yaksa was originally practically synonymous with
Deva or Devata, and no essential distinction can be made between
Yaksas and Devas every Hindu deity, and even the Buddha, is spoken
;
' Malainkclgtiimitra, Act. Ill; cf. Raghuvainsa, VIII, 62.

^ River-goddesses : Smith, V. A., History of Fine Art in India and Ceylon,
pp. 160, 161 and figs. Ill, 112; Maitra, A. K., The river-goddess Gahga, Riipam,
6, 1921 ; Vogel, Gahga et Yamuna dans I'iconographie bouddhique, Etudes
asiatiques, 1925 (the best discussion) ;Zwei unbekannte IVerke der
Diez, E.,
indischen Plastik in Ethuographisch Museum, Wien., Wiener Beitrage zur
Kunst und Kultur Asiens, I, 1926.
NO. 6
'
YAKSAS —COOMARASWAMY 37
of, Upon occasion, as a Yaksa. " Yaksa " may have been a non-
Aryan, at any rate a popular designation equivalent to Deva, and
only at a later date restricted to genii of lower rank than that of the
greater gods. Yaksa concept has played an important
Certainly the
part in the development of Indian mythology, and even more cer-
tainly, the early Yaksa iconography has formed the foundation of
later Hindu and Buddhist iconography. It is by no means without

significance that the conception of Yaksattva is so closely bound up
with the idea of reincarnation.
Thus the history of Yaksas, like that of other aspects of non-
Aryan Indian animism, is of significance not only in itself and for
its own sake, but as throwing light upon the origins of cult and
iconography, as well as dogma, in fully evolved sectarian Hinduism

and Buddhism. And beyond India, if, as is believed by many, charac-
teristic elements of the Christian cult, such as the use of rosaries,
incense, bells and lights, together with many phases of monastic
organization, are ultimately of Buddhist origin,^ we can here, too,
push back their history to more ultimate sources in non- and pre-
Aryan Indian pujas.
Adherents of some " higher faiths " may be inclined to deprecate
or to resent a tracing of their cults, more of dogmas, to sources
still
associated with the worship of " rude deities and demons " (Jacobi)
and " mysterious aboriginal creatures " (Mrs. Rhys Davids). But if
the Brahmans in fact took over and accepted from popular sources
the concept of devotion to personal deities, and all that this implied,
do we not sufficiently honor these thinkers and organizers of theo-
logical systems in recognizing that they knew how to utilize in the
service of more intellectual faiths, and to embody in the structure of
civilization, not only their own abstract philosophies, but also the
" forces brutes mystiques " (De la Vallee-Poussin) of pre-Hindu
Hinduism? And if some elements of ancient Hindu cult, perhaps of
millennial antiquity, are still preserved in the Christian office, this is
no more than evidence of the broad unity that underHes religious ten-
dencies and acts everywhere and always pagan survivals in all cur- ;
rent faiths are signs of fulfillment, rather than of failure. And in

India it becomes more than ever clear that thought and culture are
due at least in equal measure both to Aryan and indigenous genius.
*
See Garbe, Indian und das Christenttim; BerstI, Indo-koptische Kunst, Jahrb.
as. Kunst. I. T024.
EXPLANATION OF PLATP:S
Plate i
1. The Yaksa Kunika (the Parkham image now in the Mathura Museum) :
height Photo by Johnston and Hoffmann.

8' 8".
The date and identification of this figure have been matters of great con-
troversy.^ All that can be safely said is that the inscription is in charac-
ters generally corresponding to those of the Asokan and Piprahwa vase

inscriptions. Almost the only significant part of the text in the reading
of which all students agreethe name Kunika. This name has since
is
been found on the so-called statue of Manasa Devi at Mathura,^ which

is named in the inscription as that of a YaksinT, sister of Kunika. These
data appear to confirm the view long held, that the Parkham image
(so-called from the place of its discovery) represents a Yaksa and dates
from the Maurya period. When first discovered, the Parkham image
was being worshipped by the villagers as a Devata, the Baroda fragment
(HIIA, fig. 15) as a Yakheya. See also Chanda, R., in Mem. A. S. I.,
vol. 30.
The Parkham image is of great importance as the oldest known Indian
stone sculpture in the round ; it which can be fol-
establishes a formulae
lowed through many succeeding centuries. A female statue from Besna-
gar, now in the Indian Museum, Calcutta, height 7' 7", and perhaps repre-
senting a Yaksi, is also contemporary (see HIIA, fig. 8), so too, or but
little later, is a colossal female caurl-bearer from Didarganj near Patna
(HIIA, fig. 17). There is, or was, another Yaksa (or king) figure at
Deoriya, near Allahabad (see reproduction in my Origin of the Buddha
Image, Art Bulletin, 1927, Pt. 4, fig. 47) here it can be seen clearly that
;
the left hand is placed on the hip further, the figure wears a turban, and
;
is sheltered by an umbrella. The Deoriya figure must be of about the

same (Maurya) date as the Parkham image.
2. The Yaksa Bhagavata Manibhadra, set up by a guild of Manibhadrabhaktas,
at Pawaya, Gwaliar State, now in the Gwaliar Museum, First century
B. C. Photograph by the author.
Plate 2
I, 2. The Yaksa Nandi, and another Yaksa or king; perhaps the Yaksi Nandi
of Nandinagara, or the pair may be the Yaksas Nandi and Vardhana of
Nandivardhana. Patna, second century B. C, now in the Museum at
Patna. A. S. photographs.
^
Mr. Jayaswal (J. B. V, 1919) attempted to prove that the inscrip-
O. R. S.,
tion included the name of King Kunika Ajatasatru, and he

identified and dated
it accordingly about 618 B. C. (according to others this Saisunaga king died
about 459 B. C). Fatal objections to Mr. Jayaswal's views are raised by
Chanda, Four Ancient Yaksa statues, in the Journal of the Dept. of Letters,
Calcutta University, Vol. IV, 1921, where other references will be found.
^ For the figure
of " Manasa Devi," probably also of Maurya date, see Ann.
Rep. Arch. Snrv. India, 1920-21, pi. XVIII, and ibid., 1922-23, p. 165.
Plate 3
1. The Yaksa Kuvera (Kupiro Yakho), Bharhut, second century B. C, now

in the Museum, Calcutta. The vahanam, not well seen, is a crouch-
Indian
ing dwarf demon (Guhyaka?) with pointed ears. India Office photograph.
2. The Yaksa Supavasu, Bharhut, same date vahanam, an elephant. Now in the
;
Indian Museum, Calcutta. India Office photograph.
Plate 4
1. A Yaksl or Devata from Bharhut, found at Batanmara : vahanam, a running
dwarf. India Office photograph.
2. Culakoka Devata, from Bharhut : vahanam, an elephant. Now in the Indian
Museum, Calcutta. India Office photograph.
Plate 5
1. Yaksl or Devata from Bharhut ; vahanam, a horse accompanied by a dwarf

with a water-vessel. Now in the Indian Museum, Calcutta.
2. Yaksi or Devata: human (?) vahanam. Bodhgaya. India Office photograph.
Plate 6
1. The Yaksi Sudarsana, from Bharhut: vahanam, a makara. Now in the In-
dian Museum, Calcutta. India Office photograph.
2. Yaksi under asoka-tree; vahanam, a makara. From Mathura, now B. 51
in the Lucknow Museum. L. Mus. photograph.
3. Yaksi under asoka-tree, with one foot pressed against its stem (dohada
motif). From Mathura, in the Mathura Museum. A. S. photograph.
Plate 7
Yaksa with padma in hand (padma-pani) and auspicious pair (mithuna, ;
Yaksa and Yaksi?). At Amin, near Thanesar. Second century B.C.

A. S. photograph.
Plate 8
Guardian Yaksa at the base of a pillar, north torana, Sand. The panel above
shows the worship of a sacred tree (caitya-vrksa) in a grove (the
Venuvana at Rajagrha) though the theme is here Buddhist, the relief
;
serves very well to illustrate some of the descriptions of Jakkha ceie

cited above. First half of first century B. C. India Office photograph.
Plate 9
Part of the north torana, Sanci. The three uprights of the lower series consti-
tute a Buddha triad, with, in the center, the Buddha represented by the
Bodhi-tree, and on each side a padmapani Yaksa (prototype of the
Bodhisattva Padmapani). First half of first century B. C. Photograph
by Johnston and Hoffmann.
Plate 10.
I. West torana, Saiici, showing Yaksa (Guhya) Atlantes. Two panels of the
right hand pillar show the worship of caitya-trees. India Office photograph.
2. Upper part of north torana, Sand, with a cauri-bearing Yaksa showing- also
;
a symbol (often but wrongly styled vardhamana). There was originally

a Buddha triad consisting of a Dhamnacakka between two Yaksas. First
half of first century B. C. Photograph by Johnston and Hoffmann.
Plate ii
1,2. Front and rear views of a dryad bracket (Vrksaka and mango-tree) east
torana, Sanci first half of first century B. C. Photographs by the author.
;
3. Dryad (Yaks! or Vrksaka) putting on an earring; with banyan (?) tree.

Framed in a "
caitya-window " niche. Amaravati, second century A. D. or
earlier. Mfiseum? India Office photograph.
British
4. Yaksa bearing a garland, from rail-coping, Amaravati, second century A. D.
British Museum? India Office photograph.
Plate 12
1. Kusapadalamanava Jataka, with the Yaksi Assamukhi. Railing medallion

from Pataliputra, early second century B. C, now in the Indian Museum,
Calcutta. There are similar medallions at Sanci (Stupa II) and
Bodhgaya. Indian Museum photograph.
2. Yaksa (?) with bell (cf. fig. 29, right). Terracotta, about first century
A. D. Museum of Fine Arts, Boston. M. F. A. photograph.
3. Yaksa (?) held by the right arm, not seen in the photograph, is a broad
:
club; thus the Yaksa might be described as mudgara-pani (cf. the Yaksa
Moggarapani, supra). Terracotta, Maurya or earlier? Museum of Fine
Arts, Boston. M. F. A. photograph.
4. Yaksa (?) holding a ram; perhaps a bucolic divinity, a kind of Ksetrapala.
Terracotta, from Ujjain, probably Kusana, first or second century A. D.
Author's collection. M. F. A. photograph.
Plate 13
1. Yaksas (Guhyas) as Atlantes, Bharhut, Ca. 175 B. C. Indian Museum,

Calcutta. India Office photograph.
2. Winged Yaksas (Guhyas) as Atlantes; from a railing pillar at Bodhgaya,
about 100 B. C. Photograph by Johnston and Hoffmann.
3. Yaksas as Atlantes, Graeco-Buddhist, from Jamalgarhi. One is winged, and
provided with a bell. In Lahore Museum. India Office photograph.
Plate 14
I. Bacchanalian Kuvcra, Kusana, late second century A. D. From Mathura, in
the Mathura Museum. A. S. photograph.

2. Yaksi or Vrksaka (so-called river-goddess Ganges) originally one of a pair
from a doorway (forming the upper parts of the jambs) vahanmn,
:
a ntakara; tree, a mango. Gupta, about 400 A. D. From Besnagar, now

Museum of Fine Arts, Boston. M. F. A. photograph.
Plate 15
I. Pancika and Hariti, the Tutelary Pair, patron deities of wealth and fertility.
Graeco-Buddhist, from Sahri-Bahlol, now in the Lahore Museum.
Early second century A. D. A. S. photograph.
NO. 6 YAKSAS COOMARASWAMY 4I
2. Yaksa (?) Vajrapani from Mathura. Kusana; early second century A. D.?
Height of the fragment, i' 9". Now E 24 in the Mathura Museum. A. S.
photograph.
Plate 16
1. Yaksa, on railing to pillar, KaiikalT Tila, Mathura. Probably first century

A. D.
2. Yaksa, probably Vaisravana, with flames, from the Kahkall Tlla, Mathura,
same date. Both after Smith, Jaina stupa of Mathura. Both in the Luck-
now Museum.
Plate 17
r. Parasuramesvara lirigam (Siva), Gudimallam, about 100 B. C. For com-

parison with Yaksa types from Bharhut, etc. A. S. photograph.
2. Colossal Bodhisattva (Buddha), of Mathura manufacture, set up by Friar
Bala at Sarnath, 123 A. D. For comparison with Yaksa types, plate i,
figure I, and plate 2, figure i. A. S. photograph.
Plate 18
1. Ganesa, with chain of bells ; from Bhumara.

Gupta, about fifth century.
A. S. photograph.
2. Dvarapala, a Yaksa, with chain of bells. South Indian, Cola, about the tenth
century. Property of C. T. Loo.
Plate 19
1. YaksT,on door-jamb at Tadpatri makara vahanam. The tree is now

;
much conventionalized and proceeds from the makara's mouth.

The parrot (Kamadeva's vahanam), perched on the Yaksi's arm, is
a further indication that the makara in these associations is rather to be
connected with Kamadeva than regarded as a river-symbol. Parrots or
parrokeets are represented already on the shoulders of the voluptuous
Yaksls from the Bhutesar side in Mathura : and in the Lalita Vistara.
Ch. XXI, some of the apsarasas, Mara's (Kamadeva's) daughters, tempt-
ing the Bodhisattva, are said to have parrokeets or jays perched on their
heads or shoulders. Smaller Yak.sa (Guhya) Atlantes on right side
(cf. plate 13). A. S. photograph.
2. Yaksi, on door jamb of the Subrahmaniya temple at Tanjore, eighteenth
century. Makara vahanam; the tree much conventionalized; the Yaksi
holds a parrot and is pressing one against the trunk of the
foot
(presumably) a.soka-tree (dohada motif). Photograph by the author.
Plate 20
The conception and nativity of Siddhartha. Upper right, the Dream of Maya
Devi (Mahamaya) (Incarnation of the Bodhisattva in the form of a
white elephant) one female attendant also sleeping, and the Four Great
;
Kings, the Lokapalas (Kubera, etc.), occupying the four corners of the
chamber, on guard. Upper left, The Interpretation of the Dream; Maya
Devi seated. King Suddhodana enthroned, two Brahman soothsayers
seated below. Lower right, the Nativity; Maya Devi under the asoka-
tree, supporting herself by one hand (woman and tree, or yaksi motif),
with one attendant; to her proper right, the Four Great Kings holding
from her
a cloth on which the presence of the infant, miraculously born
right side, is indicated by two small feet. The stool represents the First
Bath. Lower left, Presentation at the Shrine of the Yaksa Sakyavard-
hana, as related in the Tibetan Dulva ; Mahaprajapati, aunt of the child,
holding the infant in the cloth, where itspresence is again indicated by
the two small two female attendants, one with an umbrella. The
feet ;
shrine of the tutelary Yaksa consists of a tree and altar, the Yaksa
visibly emerging from the altar and bowing to the child. From Amara-
vatl, late second century A. D. ; now in the British Museum.
Another representation of the same subject, also from AmaravatI, is illus-
trated in Fergusson, J., Tree and Serpent Worship, PI. LXIX here ;
the Yaksa is leaning forward from a sort of booth which may be called
a temple, and bowing to the child. A third example (Burgess, Buddhist
stnpas of Amaravati and jagayyapeta, frontispiece, detail) resembles that
of our Plate 20. A fourth, ib. PI. XXXII, 2, differs from our Plate 20
only in minute details.
Plate 21
1. Maya Devi's dream, Descent of the Bodhisattva, in the form of a white

elephant. The elephant is seen in a pavilion, supported by four Yaksas.
Amaravati, late second century A. D. India Office photograph.
2. The visit of Indra. On the right, the Yaksa Vajrapani above, Indra stand-
ing below. Kusana, second century A. D., Mathura. Property of
L. Rosenberg, Paris.
3. 4. Paiicika and Hariti, from door jambs. Kusana, Mathura, first or second
century A. D.
5. Paiicika and Hariti. Kusana, Mathura, first or second century A. D.
6. Scene from the Buddha's life the Buddha, nimbate, in center, the Bodhi
:
tree above him on the proper right, four women, of whom two at least
;
are represented as tree spirits. I cannot identify the scene. Amaravati,

late second century A. D. British Museum ? India Office photograph.
Plate 22.
1. Yaksi (vrksaka, dryad) bracket, from the Kaiikali Tila, Mathura. Kusana,
first century A. D. Lucknow Museum. L. Mus. photograph.
2. Yaksi, Madura, seventeenth century. Photograph by Dr. Denman W. Ross.

3. Nari-lata, ivory, Ceylon, eighteenth century. Colombo Museum. Author's
photograph.
4. Yaksa, probably Kubera; now C 18 in the Mathura Museum. Author's
photograph.
Plate 23
I, 2. Yaksa (gana) garland-bearers. One with an elephant's head, suggesting
Ganesa. Amaravati, late second century A. D. Madras Museum? India
Office photographs.
3. Palace of Kamadeva, a dance of Yaksas. Central architrave, back face of
north iorana, Sarici, about 100 B. C. India Office photograph.
"
APPENDIX
I
I to Professor Walter Eugene Clark the following tale of a Yaksa, found

owe
in the Divyavadana, 275, et seq. A certain man was the keeper of a sulka-
sala or toll-house. When he died, he was reborn among the Vyada-Yaksas. He
appeared to his sons in a dream and told them to make a yaksasthdita and attach
a bell. He said that the bell would ring if anyone tried to smuggle merchandise
past without paying toll. A man tried to smuggle in a yamall of fine cloth con-
cealed in the stick of his umbrella. The bell kept ringing and the merchants
were detained till he confessed.
This is very like the Vaisali story cited above, pp. 14, 15. The yaksasthana
may have been a separate shrine, or more likely a shrine made within the toll-
house presumably there was an image, and the bell was hung round its neck.
:
II
The well-known Besnagar kalpa-drinna capital, representing a banyan having

below its branches three money bags, and a conch, lotus, and jar, from which
square coins are welling up, probably represents Kubera in his capacity of
Dhanada, " Wealth-giver." The banyan-tree is mentioned in Mahavamsa, X, 89
Sahkha and Padma personified as lords of wealth are
as specifically his abode.
amongst the eight treasures of Kubera (Harivainsa, 2467 and 6004, and
Visnudharmottara, III, 53). The conch with coins or vegetation rising from
it occurs as a symbol elsewhere.
Ill
Page 2, note i, add: It is perhaps significant of the orthodox Vedic Brahman-

ical attitude towards the Yaksa cult that in Baudhayana Dharmasdstra. I, 5, 9
caitya-irksas are mentioned in a list of objects of which the touch causes

defilement requiring purification.
IV
Yaksa of the Klratarjunlya story (p. 14) :The Yaksa, described as a follower
of Kubera, appears in Bharavi's drama Klratarjunlya, guiding Arjuna to the
Indrakila (see H. O. S., Vol. 15).
The shrine of Kamadeva in Mrcchakatika, I, 32, is situated in a grove

(Kdmadcva adanaujjana^^ Kdmadcva ayatana udhyana).
SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 80, NO. 6, PL. 1
1 2
Yaksas, from Pfirkliam and Pawaya.
(For cxphination, see pages 7, 29, 38)
Yaksas, from Patna.

(For e.xplanatioii. see pages 12, 38)
Yaksas, from Bharluit.

(I'lir explanation, see pages 8, 39)
SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 80, NO. 6,<PL. 4
1 2
Yaksis or Devatas, frf)in Bharhut.
(For exiilanaticin, sec pages 32, 35, 39)
1 2
YaksTs or Devatas, from Bharliut and Bodhgaya.
(For explanation, see pages 32, 39)
Yaksis, from Bliarhut and Matlnira.

(For explanation, see page.s 32, 39)
Yaksa and iiiithiiiia, from Ami;

SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 80, NO. 6, PL.
Yak.sa, at Sand.
(For explanation, see pages 29, 30, 39)
pq -z
VOL. 80, NO. 6, PL. 1 1
SMITHSONIAN MISCELLANEOUS COLLECTIONS
Yaksi bracket. West torana. Sanci.
Yaksi, AmaravatT. Yaksa, AmaravatT.

The Yaksi Assamukhi. Yaksa with bel
4
Yak.'? a.
(For e.xplanatioii, see pages lo. 15, 22, 40)
Yaksas as Atlaiites or Caryatides.
I, Bharhut. 2, Bodhgaya. 3, Jamalgarhi.

(For explanation, see pages S. 40)
Vakâs from .\[athura

(l~or exi)lanatii)n, see pages 7.^1)
Siva-lirigam ; Gudiinallam. Bodhisattva (Buddlia). from Matliura, at

Sarnath.
(For explanation, see pages 8, 29, 41)
SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 80, NO. 6, PL. 1!
1 2
Ganesa ; Bhumara. Yaksa dvarapala, S. Indian.
(l"or explanation, see pages 7, 15, 41 )
SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 80, NO. 6. PL. 19
Vak.Ms, from Tadpatri and Tanjore.

( l-'tir explanntidii. see pages 32. 36, 41)
inception and Nativit\ if I'.iuldha; Ainaravati.
(For explanation, see pages 32, 34, 41, 42)

Descent of the Bodhisattva. \'isit of Indra Vajrapani a1)ove

:
AmaravatT. Mathiira.
Kubera : Mathura. HaritI ; Mathura.
y >
^4 > ^/
Kuhera and Hariti ; Mathura. Scene from Ruddha's life: .AmaravatT.
( l-'or explanation, see pages 8, cj, lo. 31. ^2)
'<:^''^
^
*% _'
Gana sâiiaiul-bcarers ; Amaravati.
Gana "arlaiul-bearers ; Amaravati.
vfW^ ''^ V '

'J', .''vu.*
Alara in darbar, with a dance of Yaksas : SancT.

( For explanation, see pages 7, 8, 13. 42)

SMC 80 Coomaraswamy 1928 6 1-43

Uploaded by

Copyright:

Available Formats

SMC 80 Coomaraswamy 1928 6 1-43

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

SMC 80 Coomaraswamy 1928 6 1-43

Uploaded by

Copyright:

Available Formats

SMITHSONIAN MISCELLANEOUS COLLECTIONS

VOLUME 80. NUMBER 6

the religious outlook of medieval and modern India is so profoundly

medieval and modern Hinduism is one thing, Vedic Brahmanism

Smithsonian Miscellaneous Collections Vol. 80, No. 6

2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 80

not the recluse who refrains from actions Whatsoever thou

in the Brahmanas and Upanisads (and afterwards, much more defi-

NO. 6 YAKSAS — COOMARASWAMY 3

such scriptures is beyond proof, it is at least certain that religious

traditions,which must be spoken of as Agamic in contradistinction

between the two types of religious consciousness had been profound

literary and monumental sources, material sufficient to present a

4 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8o

2. YAKSAS AND YAKSATTVA (" GENI-HOOD ")

crowded with Yakkhas ("gods")- They have a deva's supernormal powers.

NO. 6 YAKSAS — COOMARASWAMY 5

and the affection felt by men toward them. As remarked by Levi

rappellent de bien pres nos saints patronaux."

deity, alien to Brahman orthodoxy. In the Sutras he is invoked with

children (cf. Hariti in her original character).

Devata, INIahakoka Devata.

constant feature in Yaksa iconography.

erature Vessavana, Pancika, Jambhala, etc.),' is one of the Four

The cult of the Lokapalas or Four Great Kings (N. Vaisravana,

* For Jambhala see Foucher, L'Iconographie bouddhiquc dc I'Inde, I,

and II, p. 51 his sakti is Vasundhara, the Earth-goddess. He may be sur-

rounded by eight Yaksinis, Bhadra, Subhadra, etc. (ibid., II, 85).

^ Both motifs are of on account of their occurrence in decorative art,

NO. 6 YAKSAS COOMARASWAMY 7

shoulder flames. In this connection it should be safe to identify the

(Pimyajana) and appear to be countless in number, though few are

this may be the explanation of the statue at Pawaya, set up by a guild

Ganesa isundoubtedly a Yaksa type, by his big belly and general

18, figure I at Deogarh (pilaster left of the Anantasayin panel).

There is some confusion of Yaksas and Raksasas, who according

benevolent and malevolent as the case may be very often the

Kubera, the Raksasas with Ravana, who is their chief. Yaksas as a

Jikoku-Ten of the Kondo, Nara, Japan,^ so closely resembling the

NO. 6 YAKSAS— COOMARASWAMY 9

may be related to the Guhyas, attendants of Kuvera (Hopkins, Epic

the village deities are feminine. MiUclksi, to whom as wife of Siva,

lO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 8o

A YakkhinI by name, or rather, epithet, Assa-mukhi (" horse-

Chavannes, in T'Aiing Pao, 1904, p. 496 f

tale, III, 1919, pi. I; Waddell, Report on excavations at Paialiputra, pi. I.

Caksuraksa, Purnaraksa, Savvana, Savvajasa, Savakama, Saniiddha,'

§ 93 is a characteristic power for

45, the Deva Samiddliisumana inhabits a idjayatana-

tree in the Jetavana garden at Savatthi he had been a man in Nagadlpa.

able information on Yakkhas.

find Nandi ca Vardhanascaiva nagare Nandivardhane ,

seems to me Patna figure designated as the Yaksa Nandi in

as it is not named in the inscription. In the same list Manibhadra and

is said to be the Yaksa where is his

Sakka ( ? Indra), who is called a Yakkha of Mara's faction, may not

Epic heroes, and others. Suprabuddha in Buddhist legend is the father-in-law

Yaksas is in many ways, inter alia, by the existence of numerous tem-

p. 242. Naigamesa in the Epic is generally a goat-faced form of Agni. As