THE "WEBBED FINGERS" OF BUDDHA
By Jitendra Nath Banerjea
The Indian Historical Quarterly
Vol 6:4, December, 1930, p717-727
p. 717
According to authoritative Buddhist tradition,
Sakyamuni Gotama, when he was born, was endowed with
32 signs(1) of great men which prognosticated his
future career. One of these signs is variously
described in Pali and Sanskrit Buddhist Texts, as
Jala-hatthapado (Mahapadana-Suttanta, Rhys Davids;
Digha Nikaya, vol. II, p. 17), Jalangulihastapadah(2)
(Lalitavistara, p. 106, ed. by Lefmann) and
jalavanaddhahastapadah (Mahavyutpatti, p.6). This
sign has been rendered into English by Prof.
Grunwedel as 'his fingers and toes have a web
between.'(3) Dr. Rhys Davids, however, translates the
Jala-hatthapada of
________________
1 One who has these signs on his body is destined by
fate, according to early Buddhist tradition, to be
either a ruler or a saviour of mankind. A detailed
list of these signs 'which probably date back to
mythological origin and were originally attributed to
Devas' is given in the following works among others:
Digha vol. XI, 17-19; Ibid., III, 142; Suttanipata,
1040 sq.; Dhammapada, 352; Milindapanha, 10;
Lalitavistara, pp. 105f.; Mahavyutpatti, pp. 5f. with
slight variations here and there occurs also in
Siamese, Ceylonese, and Chinese traditions about the
person of Buddha noticed by the scholars named
Alabaster (The Wheel of the Law), Spence Hardy
(Manual of Buddhism), and S. Beal (Romantic History
of Buddha).
2 In Dr. R. L. Mitra's edition (Bibliotheca Indica
Series, p. 121), the reading is Jangulikahastapadah.
He translates it thus:'his fingers and toes joined
with webs,' and adds this note 'the web is noticed
only at the root of the toes and the fingers, the
skin between them rising slightly in a thin web'
(Lalitavistara, translation, Bibliotheca Indica
edition, p. 143 and 170).
3 Buddhist Art, p. 161.
p. 718
the Mahapadana-Suttanta as 'with hands and feet like
a net' and adds the following foot-note to it: "Like
a lattice, says the commentary, and explains this to
mean that there is no 'webbing' between fingers and
toes, but that these are set in right lines, like the
meshes of a net."(1)
Now, the artists of the Gupta period are supposed
by practically every scholar interested in Indian art
to have portrayed this feature of the webbed fingers
in many of the numerous images of Buddha fashioned by
them. Dr. A. K. Coomaraswamy refers to the 'webbed
fingers' as one of the characteristic features of the
Buddhas of the Gupta period.(2) He further remarks,
while describing the Mankuwar image, 'the fingers are
webbed, as in several other early Gupta Examples.'(3)
But, are the fingers really webbed? The correct
answer to this question can only be ascertained, if
we put under close observation a few well preserved
Buddha specimens of the Gupta period. We may refer
first to the unique and interesting Mankuwar image
mentioned above.(4) Its left hand rests upturned on
the lap, while the right hand is raised in the
abhayamudra with the palm spread outwards, fingers
not being set close to one another, but placed
slightly apart. It appears, however, that there is
the suggestion of a thin continuous web at the back
of the palm, joining, as it were, each finger with
the other. But though at the first sight, this is
apparent, the case is really not one of 'webbing.'
The Gupta artist, in order to safeguard the
preservation of these free-standing slender fingers,
not only left the intervening space near the back of
each uncarved, but the part of the stone-block in
between the outspread palm and the upper arm was not
chiselled away. However, a con-
___________________
1 Dialogues of the Buddha, pt, II, p. 14. Cf, the
commentary: jalahatthapado ti na cammena patibaddha
angulantaro. Ediso hi phanahatthako purisadosena
upahato pabbajjam pi na labhati. Mahapurisassa pana
catasso hatthanguliyo panca pi padanguliyo ekappamana
honti, tasam pana ekappamanatta jalalakkhanam
annamannam pativijjhitva titthanti, ath'assa
hatthapada sukusalena vaddhakina yojitajalavata-
panasadisa honti, tena vuttam jalahatthapado ti.
2 History of Indian and Indonesian Art, p. 74 & f.n.2
3 Ibid., p. 241.
4 Ibid., pl. XLIII, fig. 162. The image is inscribed
and dated in the year 448-49 A.D.
p. 719
summate master of his art as he was, he showed a fine
edge along the top of the fingers in order to give a
beautifying effect to his image. This finely carved
line carried over one finger tip to the other(1) led
the art-critics of the modern age to describe it as
webbing. But that it was far from the original
intention of the artist can be proved if we compare
with the image just described, the two beautiful
metal images, one, the famous copper Buddha of
Sultangunj, now in the art gallery of the Birmingham
Museum, and the other, a bronze Buddha in the
collection of the Boston Museum, both of the fifth
century A.D.(2) The right hand of the latter is
unfortunately broken, but its left hand, holding the
hem of the garment with fingers, one detached from
the other, is well preserved. Both the hands of the
former, however, are whole, the right one being in
the usual abhaya pose, while the left one is holding
the hem as above, but in a different manner. The
fingers of the only remaining hand of the Boston
image do not seem to show the slightest suggestion of
'webbing', while those of the Sultangunj Buddha,
especially the right hand ones, just suggest the
so-called webbed connection in their lower ends.(3)
The fact appears to be that the casters of these
metal images relied on the durability of the material
in which they worked and had thus no necessity for
completely joining the fingers of their image by
means of the so-called 'web' for their safety. The
greyish Chunar sandstone on which the stone artists
of the Gupta period worked, was, however, from the
point of view of durability, much inferior to metal
and thus their choice of the material led them to
adopt this peculiar device. This observation of ours
will explain why even the portion between the chest
and the top section of the upper arm beneath the
armpits of the Mankuwar image which is fully in
__________________
1 In very rare cases, the upper one or one and a half
digits of the fingers are carved in the round while
in the lower section they are joined one with the
other by means of this so-called 'web.'
2 Coomaraswamy, op. cit., pl. XLI, 160 and pl. XL,
159. Regarding the Boston Buddha, the learned
author says that the typically Gupta bronze of
figure 159 said to have been found in Burma is
probably of Indian origin'; op. cit., p. 171.
3 The original image could not be observed. But from
the plate, it seems that the fingers of the two
hands are treated in a different manner; the
detachment in the case of those in the left seems
to be more marked than in the case of the others in
the right.
p. 720
the round (unique in the case of a seated Buddha type
of the Gupta age) and which had a halo (now almost
gone) behind its head and shoulders only, is left
uncarved by the skilful artist. Need I be more
explicit and say that this was done only for giving a
greater protection to the detached arms of the image
and not for indicating a possible web between these
two sections of Buddha's body?(1) In the case of the
other seated figures of Buddha Sarnath, which are in
very high relief, their back always resting on the
Prabhabali, generally rectangular in shape, the
artists were in no fear of any sudden damage to the
arms. A remark may be made in this connection, with
regard to the left palms of some of the Gupta Buddhas
shown holding the hem of the garment in a partially
closed fist. When the fingers are slightly apart,
they appear, at first sight, connected with the
'web.' But on close observation, it can be seen that
this supposed connection is not really the 'web'
which it looks like. The Gupta artist merely left the
portion of the stone inside the bent palm unscooped
and the fingers were shown in very high relief their
inner side resting on the unscooped block. But
usually their first digits were carved in much higher
relief than the remaining portion, and it is this
feature which gives it the specious appearance of the
web.
Our suggestion about this peculiar feature of a
good many of the Gupta Buddhas is corroborated by an
interesting peculiarity of the early Mathura ones.
Dr. Vogel, while describing the Katra
Bodhisattva-Buddha, writes, "The hand (in the abhaya
pose) is connected with the back-slab by means of a
projection carved with decorative pattern"(2) (really
a lotus?). This is nothing but a different manner of
ensuring the safety of the fore-arm, palm and the
fingers of the image, on the part of the Mathura
sculptors.
To elucidate our point further a reference may be
made to some typical mediaeval Visnu images. The
four-armed Sthanaka-
____________________
1 It can on no account be suggested that these
uncarved portions near the back are edges of
garments, for had the case been so, at least a
trace of it, however slight, would have been left
on the arms along the line.
2 Mathura Museum Catalogue, p. 47, pl. VII. The
standing Buddha no. A 4 in the Mathura Museum also
possesses this peculiarity; ibid., p. 49, and pl.
XVa.
p. 721
murti of Visnu in the Mathura Museum(1) has his front
hands hanging down, the right hand being in the
varada pose with a lotus mark in the centre of the
palm. The back of the palm rests on a full-blown
lotus the stalk of which rises from the double
petalled padma (visvapadma) of the pedestal on which
Visnu stands. The left hand holding the Sankha is
also in a pose suggestive of varada and the back of
its palm rests on a lotus as described above. It may
not be quite apparent in the plate, but a close
observation of similar Sthanakamurtis of Visnu in any
of the museums will prove that the fore-arm, which is
slightly tilted forward thus making an angle with the
upper arm, is by this lotus device (which is itself
attached to the prabhavali by the portion of the
original stone here ornamentally carved in the shape
of a tapering cylinder) joined to the back slab for
its proper preservation. The fingers also resting on
the lotus blossom were thus saved from possible
damage. There are very few stone images there, which
have their fingers free and outstanding and
completely set apart, one from another.
If we refer to the Gandhara Buddha images we find
that our contention is supported by their evidence in
a peculiar manner. The hands are usually in the
dhyana or dharmacakra pose in the case of the seated
images; where these are in the abhaya pose, they are
almost invariably never thrust much forward from the
body of the image and the fingers are usually set
very close. Similar observation can be made with
regard to the standing Buddha images from
Gandhara.(2) But wherever the fore-arms were made out
of slabs other than the main one out of which the
torso and other parts of the image were carved, they
were the very first to get damaged during the
dilapidation of the monasteries and stupas.(3) But in
this connection, an
___________________
1 M. M. Cat., p. 101 and plate XVIII.
2 But compare Grunwedel, op. cit., P. 169 fig. 117,
where the fingers are set apart, but the fore-arm
with the hand in the abhayamudra is folded up and
practically attached to the upper arm and the
torso, thus minimising the possibility of damage to
the palm and fingers.
3 A.S.I.A.R., 1906-7, p. 115: "As in the case of
large Buddha figures, the hands of the larger
Bodhisattvas also were not carved out of the same
block as the main body of the figure, but were from
the beginning carved separately and added to the
finished statue" (Spooner), In this manner, the
artist was no doubt 'saved
p. 722
interesting reference may be made to a broken hand
which was discovered by Dr. D. B. Spooner in the
course of his excavations at Sahri Bahlol. He wrote
about it: "Thus the hand shown in fig. 9 of plate
XXXV is remarkable for the well defined webbing
between the fingers, one of the physical
characteristics of the Buddha figure(1)......" It is
needless to point out that the learned author
misinterpreted the evidence at his disposal. The only
conclusion that can be made from this fragmentary
piece is that in very rare instances the Gandharan
artists partially adopted the device so frequently
used by the Gupta artists in carving the hands of the
images; the hand, in this particular instance,
however, could never have been the hand of a Buddha,
because there are clearly marked bracelets on its
calf, and thus might possibly have belonged to a
Bodhisattva image.(2)
A few more references to the Buddha images of the
Gupta period may be deemed necessary for further
strengthening my hypothesis. If we study closely some
typical Buddha specimens where other poses of the
hand such as dharma-cakra, dhyana and bhumisparsa are
shown, we find that this 'webbed connection' of the
fingers is absent whether the palms are attached to
the body or the fingers are set close together (as in
bhumisparsa and dhyana-mudras). The delicately
beautiful Buddha in the Sarnath Museum(3) has his
hands in the dharmacakra pose;it seems that this
mudra in the early stages of its representation in
the Indian art of the Gupta period was a combination
of two distinct poses of the hand, viz., vyakhyana or
chin and jnana,(4) the right hand being in the
vyakhyana and
______________
a great labour of cutting away the vast amount of
superfluous stone', but he decidedly impaired the
durability of his images. The indigenous Mathura and
Gupta artists, on the other hand, surmounted this
difficulty in a thoroughly artistic and at the same
time practical manner.
1 A. S. I. A. R., 1906-7, pl. XXXV, fig. 9. For
'webbing' in the fingers of a few Bodhisattva
images in the Ind. Mus., see the foot-note above.
2 Cf. the seated Bodhisattva with the hands in the
dharmacakramudra, ibid., pl. XXXIII(b).
3 D. R. Sahni: Sarnath Museum Catalogue, p. 70-1, pl.
X; also reproduced in Coomarswamy, op. cit., pl.
XLII, 161.
4 Referred to under these names in iconographic
literature, compiled at a comparatively late period
; cf. T. A. G. Rao, Elements
p. 723
the left in the jnana pose. Now, the right hand of
our specimen shows traces of the so-called 'webbing',
while the left hand does not. The peculiar position
of the fingers in the former needed this sort of
protection, whereas that of the latter did not.
Again, if the Gupta artist really interpreted the
jala-hattha-pada of the texts as webbing of the
fingers and toes, then why did he not depict the so-
called 'fine net-like membrane' on the back of the
toes of his images? Further, the Gupta sculptors,
curiously enough, did almost always omit one of the
most important laksanas invariably met with in
earlier and later images of the Buddha, viz. the
urna. Is it not strange that they have particularly
selected a characteristic which as interpreted by
Buddhaghosa would be impossible to be shown in
plastic or pictorial representations without marring
their beauty or give it a plastic shape after
completely misinterpreting it?
I may add here a few words with regard to the
correct interpretation of the Jala-hattha-pada and
its variants in the early and late texts. The
Buddhist commentators correctly interpreted this
characteristic. Who was then responsible for this
misinterpretation? (1) We may refer here to the
English rendering of the
_________________
of Hindu Iconography, vol. I, pt. I, pp. 16-17,pl. V,
figs. 15&16. Might this combined representation of
the mudras in many early Buddhas specially of the
Gupta period as well as in many Buddhas of the
medieval eastern school of sculpture symbolise the
attainment of the bodhijnana first and then the
expounding of the same to the various beings? In
Gandhara, we do not see this manner of representing
the dharmacakra (cf. A. S. I. A. R., 1906-07, pl.
xxxll(b); this is the usual manner here).
1 From Buddhaghosa's manner of introducing his
explanation of this interesting laksana, we find
that he was afraid that people might misinterpret
it on account of the various meanings of the word
jala and so he begins na cammena patibaddha
angulantaro. The jalavanaddhahastapadah in the late
compilation Mahavyutpatti and
jalabaddhavajrangulipanipadatalata in Hodgson's
list collected from the Nepalese literature of late
date, prove that the learned commentator was quite
justified in his apprehension. It may be argued
that the wrong rendering of the term was in vogue
during the time of Buddhaghosa himself. But, then,
Dharmapala who was possibly two centuries later
than Buddhaghosa, gives its correct interpretation
(see p. 725, f.n. 2) and we can infer that even as
late as the 7th cent.
p. 724
Siamese, Sinhalese and Chinese texts by eminent
scholars. Alabaster quotes Burnouf's rendering of
this sign as "His toes and fingers are marked with
lines forming a net-work."(1) The Siamese way of
describing this sign is, "The palms and soles are
exquisitely marked and the fingers set so close that
no drop of water can pass between them"(2); the
Sinhalese: "The palms and soles appeared like richly
ornamented windows."(3) Beal, on the other hand,
renders the Chinese translation of this sign from the
Sanskrit original, thus: "The fingers and toes
severally connected with a fine net-like
memebrane"(4). The fact is that the word jala had,
among various meanings, both the senses of a net and
a latticed window.(5) Buddhaghosa skillfully uses
both these senses in explaining the significance of
the term Jalahatthapada and he is certainly right.
The word jalangulihastapadah of Lalitavistara can,
without any great stretch of imagination, be very
well explained as 'the fingers and toes marked with
jalas or uniform and parallel lines
________________
A. D. the misinterpretation did not take place. If it
be still insisted that it actually happened during
the earlier commentator's time it may be answered
that Buddhaghosa was an Indian of the 5th century
A.D. (cf. Kern, M. of Buddhism, p. 125) and thus
could certainly have seen some of the Gupta Buddhas,
if not any of the Gandhara ones, bearing the
peculiarity discussed in this paper. He saw the
danger of its misconstruction by the pious but
uncritical worshippers of the images,and hastened to
give the correct significance of mahapurusalaksana.
It is very likely, if not certain, that this
suspicious appearance of the palms of the Buddha
images led to the change in the meaning of this sign
in a much later period.
1 Albaster, The Wheel of the Law, Appendix, p. 313.
2 Ibid., p. 113. He adds this curious note to it:"It
is added that this peculiarity arose from 'his
having steadily established himself in the four
elements of benevolence, Sangkhrihawatthu'. These
are the Sanskrit Sangrahavastuni, defined as
almsgiving, agreeable speaking, kind acts, unity in
that which is for the general good."
3 Sp. Hardy, Manual of Buddhism, p. 368.
4 S. Beal, Romantic History of Buddha, p. 55.
5 Cf. Varahamihira's Brhatsamhita, (Vizianagram Sans.
Series), ch. 57, v, I Jalantarage bhanau etc. and
Utpala's comment on it.
p. 725
as are to be found in the meshes of a net or the
lattice of a window'(1). This seems to be a very
obvious interpretation; that it was so long
misunderstood by well-known scholars was perhaps
partly due to the Gupta sculptures and partly to the
equivocal significance of the term jala. But the
blame for it can on no account be laid at the door of
the artists; certainly it was not their fault that 'a
simple craftsman's device' misled generations of
art-critics of the modern age.
For the purpose of ascertaining the true
significance of the 'webbed fingers' of Buddha, I
examined several times very closely not only the
Buddhas and 'Bodhisattvas of different periods in the
collection of the Indian Museum, but also the images
of various
_________________
1 We may refer here to another interesting
explanation of this term given by Dr. Stede in his
Pali-English Dictionary, p. 116 (jala), on the
authority of the commentary of Vimana-vatthu:
"having net-like hands and feet (one of the 32
marks of a Mahapurisa, probably with reference to
long nails, Digha, II, 17 (see Dial. II, 14, note 3);
cf. Jalitambanakhehi, Vimanavatthu 81, 16 (explained
at Vim. Vat. Atthakatha 315: Jalavantehi
abhilohita-nakkehi. Tena Jali (v. l. jala-) hatthatan
Mahapurisa-lakkhanan tambanakhatan anuvyanjanan ca
dasseti)." But why Dr. Stede uses the expression
'with` reference to long nails' is not clear. The
passage in the Vimanavatthu is so mam muduhi panihi
ti muduhatthatam mahapurisalakkhanam vadati.
Jalavantehi etc. Here two of the 32 greater marks
viz. muduhatthata and jalihatthata and one of the 80
lesser ones viz. tambanakhata are mentioned. There is
not the slightest reference here to another lesser
sign viz. tunganakhata. Then, why should it be
referred to, to explain jalihattha when the sense of
the latter is obvious? Dhammapala like Buddhaghosa
clearly lays down that jali means 'covered with
net-like lines', and by his use of the single word
jalavanta, makes himself perfectly clear; it is never
jalabaddha or jalanaddha with him, as in the texts of
a much later period.
Reference may be made here to Varahamihira's
mention of damanibhabhis cadhyah (Brhatsamhita, ch.
60, Purusalaksana, v. 47) in the course of his
enumeration of the marks on human palms, and Utpala's
comment on it. The latter writes, damanibhi
rajjvabhabhis cadhya isvara bhavanti. There can be no
doubt that the characteristic significance of the
words jala, jali or jalavanta of the early Pali texts
and their commentaries are referred to here, though
in a slightly different manner.
p. 726
other gods and goddesses there. This close inspection
has convinced me that there: is a danger of our
misunderstanding many such 'simple craftsman's
devices', one of which is the subject-matter of the
present paper, if we study images only from their
reproductions in the publications on Indian art and
the museum catalogues. However excellent the
reproductions might be, there is always just the
chance of some detail, perhaps very simple from the
image-maker's point of view, but extremely important
from that of the study of the images themselves,
being not prominently shown there, or, assuming a
peculiar appearance which is far removed from its
real character.(1)
I may mention here that an interesting Sanskrit
text Samyak-
_________________
1 To refer to a typical case: The excellent
reproduction the of the Sarnath Buddha with his
hands in the dharmacakra mudra in Dr.
Coomaraswamy's famous book History of Indian and
Indonesian Art leaves us in some doubt as to
whether the palms are connected with the chest by a
portion of the original stone block left there
purposely by the artist. But my study of a Sarnath
Buddha in the lndian Museum (S. 49, Cat. II, p. 26)
with his hands in the said pose led me to find out
that the right palm which is in the Vyakhyana pose
and about 1 1/2 inch away from the right breast of
the figure, the left one in jnana pose practically
resting on the left breast, is joined with the
torso in a manner just referred to. I am sure that
this is almost invariably the case with the other
Gupta Buddhas with their hands in the same pose.
Similar is the case with the mediaeval Buddha
figures in the collection of the Indian Museum,
whatever the poses of their hand might be--abhaya,
varada, or dharmacakra (the last really being a
combination of vyakhyana and jnana?).
One observation more about the presence of
these-called webbed fingers in Gandhara art. Exhibits
Nos. 9 (4915, r. h. abhaya, 1. h. holding lotus), 10
(4946, r. h. abhaya, l. h. holding a vase) in the
Gandhara room, Indian Museum, show their r. h.
fingers connected together by `a thin membrane'
(certainly it would appear so in reproduction)
resting on a hexagonal wedgeshaped stone thus
connecting the back of the palm with the torso, but
leaving the lower portion of the palm and the calf
thoroughly detached; in one case the wedge is absent.
The figures are those of the celestial Bodhisattvas.
Ganhara Buddhas nos. 4905 and 12 in the same room
bear the same peculiarity in a slightly modified
manner.
p. 727
Sambuddhabhasita-Pratimalaksana by name brought
by me friend and colleague Dr. P. C. Bagchi of the
Calcutta University from the Durbar Library, Nepal,
which is being edited by me, though containing many
interesting and important details about the Buddha
Pratima is silent about this so-called "webbed
fingers".