Pramāṇa · Search · About · API

The Beginnings of Buddhist Art and Other Essays in Indian and Central-Asian Archaeology c10: The Six-Tusked Elephant: An attempt at a chronological classification of the various versions of the Shaḍḍanta-Jātaka

The Beginnings of Buddhist Art and Other Essays in Indian and Central-Asian Archaeology c10

Language
English

The Six-Tusked Elephant: An attempt at a chronological classification of the various versions of the Shaḍḍanta-Jātaka

c10.2

The close relation which exists between the written and the figured forms of the Buddhist tradition has no longer to be proved. It is known by experience. Rare indeed are those narratives of Buddha's miracles whereof no illustration has yet been discovered; still more rare are the images which do not at once find their commentary in the texts already published. And thus we have naturally come to speak of the help which, on numerous details of exegesis, the texts and monuments reciprocally lend (²). All the same, it is to be observed that until now we have principally made use of the first to explain the second. In fact the two sorts of documents seem to be unequally matched: and the muteness of the stones will never, in the estimation of philologists, be able to equal (as regards the extent and variety of the information which can be derived from them) the verbosity of the writings. However, there is one point in which the sculptures have an advantage over the manuscripts, namely the permanent fixity of their testimony. Such as they were when they left the hands of the

c10.3

(¹) Extract from *Mélanges Sylvain Lévi*, Paris, 1911. (²) Cf. *Une liste indienne des Actes du Buddha* in the *Annuaire de l'École pratique des Hautes Études, Section des Sciences religieuses*, 1908, a paper of too technical a character to be translated here.

c10.4

workman, such are they still to-day; or at least, if they are likewise subject to mutilations and susceptible, strictly speaking, of being counterfeited, no attempt at rifacimento or interpolation, that scourge of Indian literatures, could in their case pass unperceived. Guaranteed against the insidious address of the diasceuasts, they are equally so against the individual fancy of their own authors, who are forcibly restrained by the material conditions of their technique. It results from this that they can be arranged with perfect assurance in chronological order and dated with a sufficient approximation. It is in this sense that we are able to say with Fergusson, that « in such a country as India, the chisels of her sculptors are ...immeasurably more to be trusted than the pens of her authors(¹) ». It is in virtue of this advantage that the figured versions seem to us able in their turn to render some service to the written accounts of the same legend. In short, after having so often applied the texts to the interpretation of the monuments, we should like on this occasion to essay the application of the monuments to the chronology of the texts.

c10.5

¹ FERGUSSON, History of Indian and Eastern Architecture, Preface to the first edition, 1876, p. VIII (2d edit., 1910, p. x).

c10.6

For this purpose we will direct our attention to a celebrated legend, which, however, it may not be useless briefly to recall to the reader, that of the « elephant with six tusks » (Skt. Shaḍdanta, Pâli Chaddanta, Chinese Lieu ya siang). Of course, this marvellous animal was none other than one of the innumerable past incarnations of our Buddha; and he lived, happy and wise, in the company of his two wives and of his troop of subjects in a hidden valley of the Himâlayas. However, the second wife, wrongly believing herself slighted for love of the first, gives herself up to death in an access of jealous fury, making a vow one day to avenge herself upon her husband for his supposed want of affection. In the course of her succeeding existence she becomes, thanks to some remnant of merit, queen of Benares, and possesses the gift of remembering her previous birth. She astutely obtains from the king permission to despatch against her former husband the most skilful hunter in the country, with orders to kill him and bring back his tusks as a proof of the success of his mission. The man does, in fact, succeed at great risk of his life in striking the noble elephant with a deadly arrow. But the soul of the Bodhisattva is inaccessible to any evil passion : not content with sparing his murderer, he voluntarily makes a present of the tusks whereof the man had come to rob him. When the hunter finally brings back to the queen this mournful trophy, she feels her heart break at the sight of it.

c10.7

Such is this touching story, reduced to its essential and most generally reported features : for it is known under multiple forms. We know, in particular, that it appears in the Pâli collection of the Jâtaka (nº 514). Since 1895 M. L. Feer has compared with this text, point for point, the Sanskrit account in the Kalpadrumâvadâna and two Chinese editions, taken, the one from the Lieu tu tsi king (Nanjio, nº 143) and the other from the Tsa pao tsang king (Nanjio, nº 1329); but, with perhaps excessive prudence, he was careful not to draw any conclusions from this detailed comparison (¹). More recently the translation of the *Sûtrâlaṅkâra* of Açvaghosha, so excellently rendered by M. Ed. Huber from the Chinese of Kumârajîva, has made accessible to us a new and most important version (¹). Finally, a publication by M. Ed. Chavannes has placed at the disposal of Indianists generally both a complete translation of the two texts quoted by M. L. Feer, and also a translation of the corresponding passage of the *Ta che tu luen* (Nanjio, nº 1169), ascribed to Nâgârjuna (²). So much for the literary sources of our study (³). If we now turn to the works of art, we observe that we have been no less fortunate in having preserved to us at the same time a medallion from Barhut (⁴), another from Amarâvatî (⁵), a lintel from Sânchî (⁶), a fragment of a frieze from Gandhâra (⁷), and finally two frescoes from Ajaṇṭâ, the one

c10.8

¹ Journal Asiatique, Jan.-Feb. 1895. For the version of the Kalpadrumâ-

c10.9

*vadâna*, cf. the Sanskrit Ms. 27, fol. 232 vº-240 vº of the Bibliothèque Nationale and RAJ. MITRA, *The Sanskrit Buddhist Literature of Nepâl*, pp. 301-303. — We refuse to take into account the commentary of vv. 26-27 of the *Dhammapada*, which, as Mr. FEER also remarks, has scarcely any feature in common with the *Shaḍḍanta* legend.

c10.10

(¹) Ed. HUBER, *Sûtrâlaṅkâra*, Paris, 1908, ch. XIV, nº 69, pp. 403 sqq.

c10.11

(²) Ed. CHAVANNES, *Cinq cents contes et apologues extraits du Tripitaka chinois*, three volumes (1911). The story nº 28 (I, p. 101) represents the passage in question from the *Lieu tu tsi king*; the two other extracts will appear in vol. IV. Strictly one might connect with it the story nº 344, which also presents the characteristic trait of the gift of the tusks, but in quite different surroundings. We are happy to take this opportunity of thanking M. CHAVANNES, whose great kindness permitted us to make use of the relevant pages of his work prior to publication.

c10.12

(³) As to nº 49 (not yet published in the *Bibl. Indica*) of the *Bodhisattvâvadânakalpalatâ*, we cite it merely for record : for this narrative is missing from the only ms. (*Sanscrit 8*) of the Bibliothèque Nationale (see below, p. 204, n. 1).

c10.13

(⁴) A. CUNNINGHAM, *Stûpa of Barhut*, 1879, pl. XXVI, 6.

c10.14

(⁵) J. BURGESS, *Buddhist Stûpas of Amaravati and Jaggayyapeta*, 1887, pl. XIX, 1.

c10.15

(⁶) Rear face of the middle lintel of the southern gate; cf. J. FERGUSSON, *Tree and Serpent Worship*, 2ᵈ ed., 1873, pl. VIII.

c10.16

(⁷) *Art gréco-bouddhique du Gandhâra*, fig. 138 (fragment of the counterin Cave X, and the other in Cave XVII (¹). The identification of these bas-reliefs and of these paintings is fortunately no longer matter for reconsideration, except, perhaps, in detail (²). From the very fact that the meaning of these images has once for all been recognized, they have taken their place side by side with the texts in the capacity of independent and trustworthy witnesses to the divers forms which the legend has successively assumed. Altogether we find ourselves in possession of no less than twelve versions, of which six are provided by art and six by literature. These twelve versions are, if we may say so, so many successive « stages » of the tradition : the precise problem is to classify these various stages in their chronological order.

c10.17

We must admit that, if we were reduced solely to the historical data relative to the texts, the enterprise would be almost desperate. It is easy to contest the orthodox belief, according to which the stanzas of the Jātaka all fell from the lips of Buddha himself; it is much less easy to replace it by more satisfactory assertions concerning the exact time of the composition of these gāthās, which are certainly very ancient, more ancient at times than Buddhism. Their commentary (aṭṭhakathā), according to the confession even of the monks of Ceylon, has existed under its present form

c10.18

march of a staircase, derived from the hill of Karamâr; Lahore Museum, nº 1156).

c10.19

(1) Ajaṇṭā, Cave X : J. GRIFFITHS, The Paintings in the Buddhist Cave-temples of Ajaṇṭā, 1896, I, pl. 41 and fig. 21; cf. J. BURGESS, Notes on the Buddha Rock-temples of Ajaṇṭā, 1879, pl. VII, 2, and Arch. Survey of Western India, IV, pl. XVI. — Cave XVII, GRIFFITHS, ibid. fig. 73 and pl. 63.

c10.20

(2) Cf. for example, infra, p. 194, n. 1 and p. 195. The majority of the published descriptions are in error in speaking of more than one hunter : it is, of course, question of the same individual, represented in various attitudes and at different moments.

c10.21

only since the Vth century A. D.; but in their view this could only be the translation into Pâli of a prose which was quasi-contemporaneous with the verses (¹). Of the Kalpadrumâvadâna all that we can say without imprudence is that this versified amplification does not bear the marks of high antiquity. As to the dates at which the Chinese translations were made and which, according to the information kindly communicated by M. Chavannes, extend from the end of the IIIrd century to that of the Vth of our era, they naturally can furnish us only with a terminus ad quem. Thus, as far as the texts are concerned, practically every extrinsic element of chronological classification is lacking. Happily we are a little better served, as regards the images. Each of these forms part of a whole to which either votive inscriptions or technical considerations permit us to assign a determinate epoch. It is established that the bas-reliefs of Barhut and of Sânchî go back to the IInd or Ist century B. C. (²). Those of Gandhâra and of Amarâvatî are by common accord attributed to the Ist or IInd of our era (³). It is to the same epoch at the latest that, on the strength of the inscriptions and the style, Messrs. Burgess and Griffiths ascribe the archaic paintings of Cave X at Ajanṭâ : on the other hand, the same authorities bring the decoration of Cave XVII down to the beginning of the VIth century (⁴). Certainly these are only approximate dates : but it is a good thing to have even so much, and we must consider ourselves fortunate, if we succeed, by using

c10.22

(¹) Cf. RHYS DAVIDS, Buddhist Birth-Stories, 1880, Introduction, pp. 1-11. (²) See above, pp. 4, 34, 67. (³) Cf. Art gréco-bouddhique du Gandhâra, p. 42. (⁴) For the « Cave X » see GRIFFITHS, loc. cit., pp. 5 and 32; BURGESS, Notes, p. 50; for the « Cave XVII » GRIFFITHS, ibid., p. 5; BURGESS, ibid., p. 61 (cf. p. 57).

c10.23

these figured monuments as so many land-marks, in dating some of our texts with a similar degree of approximation. Nay, were we not able to call to our aid these hitherto unutilized auxiliaries, it would be wiser to surrender in advance every attempt at historical classification.

c10.24

Certainly we should not for that reason remain completely disarmed before the confused mass of these often divergent versions; and it would be our part to introduce — by recourse, for want of anything better, to some internal principle of coordination — an order at least theoretical. It is indeed the favourite occupation of folklorists thus to draw up genealogical trees of what they have decided to call « families of tales ». But, if the enterprise is possible, and the pastime permissible, it goes without saying that the result can be of value only upon a double condition, namely that we shall have known how to choose the topical detail which must act as main-spring for the establishment of the series, and that we shall have well observed and followed out, in the arrangement of this series, the natural course of human affairs. Now, in the case of the *Shaḍḍanta-jātaka* we are in no wise puzzled to discover at once the characteristic trait and the way in which to use it. It is a well recognized law that successive versions of narratives of this kind have a tendency continually to outdo each other in the direction of increasing edification. The usual effect of this pious inclination is, let us say in passing, to destroy by degrees the whole salt of the story together with its probability and its ingenuousness, while substituting for it compositions whose insipidity is sweetened to the point of nausea. Nevertheless there is no religious literature, and the Buddhist less than any other, which, its original raciness once evaporated, escapes this deplorable and fatal invasion of convention and artificiality. Now what, in the theme with which we are at present concerned, is the essential point, wherein exactly its edification lies? In order that we may not be accused of choosing arbitrarily and to suit the necessities of the case, let us appeal to the *Lalitavistara*, which happens to sum it up in a verse (¹): at the time of his previous birth as the elephant Shaḍdanta (it is the Gods themselves who subsequently remind the Bodhisattva, in order to encourage him to follow his vocation) « thou didst sacrifice thy teeth of dazzling beauty, but morality was saved. » This is indeed the point of the story, which has caused it to be ranged under the category of the « perfection of morality », or better, « of goodness » (²): it is the surrender by the elephant of his beautiful ivory tusks, as sanction to the pardon granted to the hunter who has just mortally wounded him. But there is more than one way of returning good for evil, and it can be done with more or less good grace. In this particular case the virtuous elephant might have limited himself to allowing his enemy to work his will; or, better, he might have facilitated the operation for him; or finally, which quite attains to the sublime, he might have done the deed himself for the advantage of his murderer. It is evidently in the order of this increasing generosity that, in theory, the various versions will have to be classified.

c10.25

¹ *Lalitavistara*, ch. XIII, 40; ed. LEFMANN, p. 168, l. 9: *Parityaji te ruciraçubhadantâ na ca tyaji çîlam*. — Naturally it is this same point that is emphasized in the résumé of Hiuan-tsang to which reference will be made below, p. 199.

c10.26

² *Çîla-pâramitâ* : this is the classification of the introduction to the *Jâtaka* (ed. FAUSBØLL, I, p. 45; trans. RHYS DAVIDS, p. 55) and of the *Lieu tu tsi king* (CHAVANNES, *Cinq cents contes*, I, pp. 97 sqq.).

c10.27

In fact, if we recur to the written accounts which have been preserved to us, we remark that the protagonist adopts in turns one or other of these attitudes at the culminating moment of the narrative : « Rise, hunter, take thy knife (khura, Skt. kshura), and cut from me these teeth before I die », is the extent of what the elephant says in stanza 31 of the Jātaka ; and his interlocutor does not let him repeat the invitation. The Lieu tu tsi king considers it only right to add a little moral homily. But with the prose commentary of the Jātaka things become more complicated. The animal has attained a size so monstrous, that it is only with great difficulty that the man succeeds in raising himself up to the root of its tusks, and even there, though instead of the hatchet of a savage (the use of which would, in fact, have been disastrous to the ivory) he now uses a more perfect instrument, the saw (kakaca, Skt. krakaca), he vainly exhausts himself with cruel efforts : his victim himself must come to his aid. In order to make things more pathetic, the monastic editor does not recoil before the most flagrant contradictions. The elephant is already so weak that he cannot raise his silver trunk to take hold of the saw ; and he has to call all his senses together, in order to beg the hunter to give him the handle of it ; after which — as it is generally agreed that the Bodhisattva is by his very nature endowed with supernatural strength — he instantly saws through his two tusks (for here (') they are no more in number than two), like the tender stems of a

c10.28

(1) M. L. FEER (loc. cit. p. 50 and p. 77, note 1) has observed the same thing in the Kalpadrumâvadâna, in spite of the persistence in the title of the traditional name of « six-toothed »; but it is to be noticed that the word plantain! In the *Kalpadrumâvadâna*, the *Ta che tu luen* (which besides is simply a very summary résumé) and the *Tsa pao tsang king*, the hero does not even trouble himself to borrow from his murderer any instrument whatever: he himself breaks off his tusks, according to the first two accounts against a rock, according to the third against a big tree. But to the *Sûtrâlaṅkâra* belongs the palm for spontaneity in the action of the martyr : it is simply « by slipping his trunk round his teeth » that this time the elephant pulls them out, not without pain or grief, while the hunter respectfully waits for him to present them expressly to him. Further than this it is impossible to go.

c10.29

Thus, then, we obtain a first classification of all our texts. Theoretically it is unassailable; practically we must not form any illusions as to its historical value. If notwithstanding we proceed to arrange the figured monuments according to the same criterion, the chances of arriving by their intervention at a less conjectural result assume immediately a better aspect. In fact we are not long in perceiving that the order thus obtained coïncides exactly with that already forced upon us by the purely archæological data. At the head of them there always comes, in its simplicity, the medallion of Barhut: on the left the hunter, having put down his bow and arrows, sets about cutting off the elephant's tusks with a rude saw (¹). The latter has kindly crouched

c10.30

danta occurs in the text very frequently in the plural and not in the dual. On the other hand, it is unfortunately impossible to know what was said on this particular point by texts of which we no longer possess more than the Chinese translation.

c10.31

(¹) See above, p. 39. Perhaps it is worth while to remark that, in the Barhut version, the cause of the drama is evidently the same as in the *Lieu tu tsi king*, that is, the gift of a lotus to the first wife, if at least, as is said in the *Kalpadrumâvadâna*, she did not receive two, one to decorate each of her temples. This reason is cited by the prose commentary of the Jâtaka down to further the wishes of his enemy and to render his task less difficult (pl. XXIX, 1). The case is the same in Gandhâra and at Amarâvatî, where in addition we see represented the episode of the hunter hiding in a ditch, in order to wound the elephant in the stomach with an arrow (pll. XXIX, 2 and XXX, 1). The fresco of Cave X of Ajaṇṭâ shows us likewise, in the words of Mr. Griffiths (loc. cit., p. 32), « the huge six-tusked elephant lying down and a hunter engaged in cutting off the six tusks ». It is, as a matter of fact, six tusks — more or less distinctly separated, but always carefully noted — that the elephant has in all these representations, except that from Gandhâra. But, when we pass on to the painting of Cave XVII, the picture is changed : « the huge white Elephant King », says Mr. Griffiths (ibid., p. 37), is standing, « with only one tusk, upon which he rests his trunk, while a man kneels and makes profound obeisance before him ». In reality (cf. pl. XXX, 2), the elephant, to whom the artist no longer lends more than his two normal teeth, has already torn out one, and is about, as it is written in the Sûtrâlaṅkâra, to twist his trunk round the second, in order to pull that out in its turn. And during this time the hunter, in adoration before him, awaits the accomplishment of the magnanimous sacrifice. There is, as we see, a striking parallelism of development between our two kinds of documents; and it is continued from one end to the other of the two series.

c10.32

If now we bring the two lists together, we obtain, always only as a subsidiary one : of the first, a very ingenious one, that he advances, — and according to which the great elephant one day, unintentionally, by shaking a çâla tree in full blossom, caused to fall on his second wife, who was standing to windward, only twigs of wood, dry leaves, and red ants, while the first, who was to the leeward, received flowers, pollen and green shoots — there is no more question at Barhut than in the texts, except this particular commentary.

c10.33

by virtue of the same principle and by the simple intercalation of the various versions (¹) in the position respectively belonging to them, the following combination :

c10.34

I. Stanzas of the Pâli Jâtaka :

c10.35

The hunter cuts off the teeth with a knife.

c10.36

II. Medallion of Barhut (IIⁿᵈ century B. C.) :

c10.37

The hunter cuts off the elephant's teeth with a saw.

c10.38

III. Medallion of Amarâvatî IV. Fresco of Ajaṇṭâ, Cave X (Ist-IInd century A. D.) : V. Counter-step of Gandhâra

c10.39

The same version as at Barhut.

c10.40

VI. Lieu tu tsi king (trans. by Seng-houei, d. 280) :

c10.41

The same version (the instrument is not specified).

c10.42

VII. Prose Commentary of the Jâtaka (rendered into Pâli in the Vᵗʰ Century) :

c10.43

The elephant himself saws off his teeth.

c10.44

VIII. Kalpadrumâvadâna :

c10.45

The elephant himself breaks off his teeth against a rock.

c10.46

IX. Ta che tu luen (trans. by Kumârajîva between 402 and 405) :

c10.47

The same version as in the Kalpadrumâvadâna.

c10.48

X. Tsa pao tsang king (trans. by Ki-kia-ye and T'an yao in 472) :

c10.49

The elephant himself breaks off his teeth against a tree.

c10.50

XI. Sûtrâlaṅkâra (trans. into Chinese by Kumârajîva towards 410) :

c10.51

The elephant himself pulls out his teeth with his trunk.

c10.52

XII. Fresco of Cave XVII of Ajaṇṭâ (VIᵗʰ century) :

c10.53

The same version as in the Sûtrâlaṅkâra.

c10.54

(¹) It will be noticed that the final list differs slightly from that which we drew up at the beginning of this study. On the one hand, we have had to leave aside the lintel at Sânchî, which, treated too decoratively, did not supply us with any information upon the precise point which we are now considering; on the other hand, the tenor of the commentary of the Jâtaka has shown itself so divergent from that of the text that we have had to divide this source into two. On the whole, then, we always reckon twelve versions, five artistic and seven literary.

c10.55

Such as it is, the chronological table thus obtained is at least worthy of being taken into consideration; and the hope occurs to us that we may have restored, in accordance with the natural play of the religious conscience, the different phases of the evolution of the story. In fact, it is not that we have thus arbitrarily arranged all the accessible documents: it is they, which, when interrogated on a definite, capital point, have spontaneously and without any violence or solicitation on our part, arranged themselves in the order indicated above. As far as the images are concerned, this series is not only in conformity with their historic succession on the whole: it takes into account, in a surprising manner, their proximity as well as their aloofness in time, grouping together at the beginning the four which resemble each other, and reserving the sole variant to quite at the end. Then, as regards the texts, the impression of confidence and security, which arises from this spontaneous classification, would be still further increased, if we made our inquiry apply equally to such or such other accessory episodes of the legend. It is not, indeed, the manner of giving the tusks only, it is a whole group of concomitant details, which concur in determining theoretically, for one who knows how to read them, the order of priority of the various narratives. Take the one which comes at the head of the list, that is, the rhymed account of the Jātaka; you will observe that there everything takes place in accordance with the customary rules of elephant-hunting. The hunter hides in a ditch; at the cry of the wounded animal all his companions flee; remaining alone in the presence of the man, the elephant advances to kill him: the fact that it stops on recognizing on him the colour of the monastic coat is the sole sign of the Buddhist adaptation of the ballad. Beginning with the *Lieu tu tsi king* (nº 6), it is no longer sufficient that the clothes of the hunter should be naturally of a reddish-brown, like those of that hunter (¹) from whom Buddha formerly borrowed his first monk’s coat : henceforward the man will deliberately disguise himself as a monk, in order to inspire confidence in his prey. But, since he now employs this infallible means of approaching within easy reach, there is no longer need for him to hide in ambush : and in fact, beginning with the *Kalpadrumâvadâna* (nº 8), he ceases to have recourse to this obsolete proceeding. At the same time, as he has approached openly, it will be necessary that by a refinement of pity his victim should defend him against the vengeance of his first wife, if not from the rest of the herd : this is what the Bodhisattva fails not henceforth to do (nº 9-11). Soon — with nº 10 — scruples are aroused in the mind of the hunter, thus protected : he no longer dares to lay his sacrilegious hand on the tusks of the « Great Being », for fear that it may fall from his body. Finally, in the *Sûtrâlaṅkâra* (nº 11), to these interested fears is added a real and too legitimate repentance. Thus is seen how a striving after increased edification has by degrees modified a whole concordant assemblage of details : and so it is not, as might be imagined, for an isolated reason, but by a whole sheaf of proofs, if we had time to consider them more closely, that the order of the preceding table would be justified.

c10.56

¹ And doubtless, of all people of low caste : for the costume of his order of mendicant brothers Buddha would quite naturally have chosen the coarsest material of the cheapest colour. At least we do not see that the tradition relative to the *kâshâya*, if it had any meaning, can at the bottom signify anything else. For its variations in form cf. also *Art gréco-bouddhique du Gandhâra*, p. 369.

c10.57

Does this mean that we must blandly accept for the known documents all its features, and that, on the other hand, in order to fix the date of every new version, it will be sufficient to refer it to the corresponding degree on this chronological scale? In the case of a figured monument we should be rather inclined to believe so, provided that it is upon inquiry verified whether by chance it were not a case of some more or less archaizing imitation. As soon as it is a text that is concerned, the question becomes much more delicate, and from the very beginning we fall again into our difficulties. For the most part the table furnishes us with nothing more than simple presumptions, and these need still to be correctly interpreted. It affirms, for example, that the *Sūtrālaṅkāra* represents the state of the legend current from the Vth century of our era; and of this fact we have, in truth, two indisputable proofs. The one, of an artistic order, is the fresco of Cave XVII of Ajaṇṭā (VIth century). The other, literary, but by a happy chance dated exactly as belonging to the second quarter of the VIIth century, is nothing less than a passage from Hiuan-tsang: the story of the Shaḍdanta, gathered by the great pilgrim at Benares, is, as M. S. Lévi has already pointed out in his admirable article on the *Sūtrālaṅkāra et ses sources*, « an exact and faithful résumé of the story of Açvaghosha »(¹). What are we to deduce from these statements? As the name of the author scarcely allows us to bring the work lower down than the IInd century of our era, must we hasten to conclude, as we might be tempted to do, that the account of the « white elephant with six tusks » is only a late addition? This story forms a part of the XIVth chapter. Now M. Ed. Huber warns us in his preface « that one of the first catalogues of the Chinese Tripitaka, the Li tai san pao ki, drawn up in A. D. 597, gives only ten chapters » to the Sûtrâlankâra. Besides, we feel to what a degree this collection of tales (which, like that of the Jâtakamâlâ, must at a very early date have been used by Buddhist sermon-writers for the needs of their daily preachings) was ill-defended against interpolations... — This is all very well and good; and after all the thing is possible: but surely the place assigned to the Sûtrâlankâra in our list by its conception of the Shaḍdanta-jâtaka does not authorise us to conclude from it anything of the kind. What, in fact, does it prove? That this text already contains the form which the legend had assumed in the imagination of the artist painters of the VIth, and in the memory of the guides of the VIIth, century. And in what way does it prevent the poetic talent of Açvaghosha from having been the first to put into circulation the elaborate version which, as we have just seen, was coherent in all its parts and destined to have great success and definitely to supplant the far too primitive account of the stanzas of the Jâtaka? Two or three centuries may not have been too much for this literary production to become popular in its turn; and here we find positively no peremptory reason invalidating its authenticity. The best course, with a view to the solution of this question — as of the question how far the Chinese translation is adequate to the Sanskrit original — is to leave it to the future, especially now that we may hope for everything from the discoveries of manuscripts in Central Asia (¹).

c10.58

¹ Cf. M. S. LÉVI, *Açvaghosha, le Sūtrālaṅkāra et ses Sources*, in the *Journal Asiatique*, July-August 1908, p. 175. Stanislas JULIEN (I, p. 360) translates in fact: « The elephant tore out his tusks », and WATTERS (II, p. 53) says exactly the same. According to BEAL (II, p. 49) he « broke off his tusks ». M. CHAVANNES admits that this second translation might literally be possible: but, not to mention that the sense of « breaking » is given in the dictionary of Couvreur as a secondary meaning, it is that of « tearing out » which corresponds to the description of the attitude in the Sûtrâlankâra, and its representation in the fresco of cave XVII of Ajaṇṭâ.

c10.59

On the other hand, there is a point on which we believe we may already risk a categorical affirmation : we mean the manifest divergence which is seen between the version of the verse text of the *Jātaka* (nº 1) and that of the prose commentary (nº 7). This divergence is not to-day remarked for the first time (²): what we have here is only one more striking experimental demonstration of it. Read afresh with reference to our list the text of Fausböll's edition (V, pp. 37 sqq.), and you will quickly perceive that the editor of the commentary in its present form knew a state of the legend analogous to that reflected in the works numbered 8 to 11; that, if he did not follow these latter right to the end, it was because he was hindered at each moment by his text, whose ancient particulars held him back, *nolens volens*, on the incline down which he asked nothing better than to glide; and that finally he applied himself as well as he could to inserting between the lines of the ancient story ornaments borrowed from the later legend. Henceforth you will hold the secret of the strange liberties which he takes with the letter of the stanzas; and you will have only to note point by point, as they occur, the most flagrant of his offences. You will smile at the palpable cunning with which, from the first line (p. 37, l. 1), he transfers the name of the elephant, Chaddanta, to the lake near which the latter dwells, and a

c10.60

(¹) It is known that Prof. LÜDERS has already announced the discovery of fragments (still unedited) of the Sanskrit text of the *Sûtrâlaṅkâra*.

c10.61

(²) It is sufficient to refer here to Prof. LÜDERS in *Göttingische Gelehrte Nachrichten*, 1897, p. 119, and M. E. SENART'S article on *Les Abhisambuddhagâthâs* in the *Journal Asiatique*, May-June 1901, pp. 385 sqq.

c10.62

little further (p. 41, l. 23) glosses his « six tusks » by « two tusks of six colours »; for you know that the latest mode was to ascribe to him only one pair. Where the good monk will perhaps seem to go rather far, is when he translates *khura* by *kakaca* (p. 52, l. 9), and unblushingly essays to make you believe that knives are saws, in other words, that chalk is cheese. But soon you will content yourself with shrugging your shoulders before this strange and systematic perversion of the text which he was supposed to interpret; the fact is that you read his hand in advance and see why, before allowing the hunter to descend into the ditch specified by stanza 23, he believes it necessary to clothe him in the *kāshāya* of a monk (p. 49, l. 8); why, when according to stanza 24 the whole troop is scattered to the « eight cardinal points », he considers it more suitable to detain by the side of the wounded one at least his faithful wife (p. 50, l. 9); why, a few lines further down, he has her brutally driven away, for fear she should punish the assassin (p. 50, l. 19), etc. And when finally to stanza 32 — which states merely that the hunter took his knife, cut off the elephant’s tusks, and departed — he openly opposes (p. 52) the absurd and pathetic account which we have already analysed (¹), the measure is heaped up and the cause decisively heard. If the *gāthās* have all the characteristics of an ancient popular plaint, which the barbarity of the proceeding employed by the hunter to get possession of the ivory forces us to declare anterior to the Barhut medallion,

c10.63

(¹) See above, p. 193.—It on all these points we have not referred to vol. V of the English translation carried out under the direction of Professor COWELL, it is because the metric version of Mr. W. FRANCIS (either through blind confidence in the commentary or on account of the necessities of the rhyme) seems to regard it as a duty to palliate all the divergencies between the prose and the verse. Thus it is that the beginning of stanza 32 becomes on page 29 : « The hunter then the tusks did *saw* » etc.

c10.64

that is to say, to the IIⁿᵈ century B. C., it is no less evident that their *aṭṭhakathā* was not merely translated into Pāli, but also accommodated to the taste of the times by a cleric of the Vᵗʰ century of our era. It is a chasm of at least seven centuries that opens before our eyes between texts which at times some persons have desired to believe contemporaneous.

c10.65

Thus, whether we arrive at simple points of interrogation or at real certainties, according to the case, it is worth while to take note of these first results. It is well known that in matters of chronology the Indianist is accustomed to be satisfied with very little. He can no longer neglect the data afforded by a comparison of the texts and the monuments, wherever they lend themselves to it. We have certainly chosen a relatively favourable specimen for our attempt : but as regards more than one *jātaka*, and even more than one miracle of Buddha, it would already be possible to draw up a table analogous to that whose spontaneous generation we have just encouraged. We may augure that these studies in detail, in proportion as excavations and new editions supply their constituent elements, will come to each other's aid, and that by a series of tests chronological data will in the end become more and more precise. From that time it would no longer be of such or such a particular episode, but of the whole Buddhist legend that we should succeed in distinguishing the successive states. If it is permissible even to print prognostications which are still so vague, we should be very much surprised if we did not see reproduced, in a general way, the fact dominating the present list of the versions of the *Shaḍḍanta-jātaka*. In fact, it is self-evident that these latter divide nearly equally into two large groups, profoundly divergent from one another, between which the Singhalese commentator of the Jātaka vainly endeavoured to construct a bridge. The six first are closely connected with the old native tradition : the five last proceed no less unanimously from a new spirit, which probably filtered into India through its north-west frontier, as a result of foreign invasions. Thus, this table would be before all an excellent illustration of the « crisis » which a succession of great political upheavals at last, a short time after the beginning of our era, provoked in the Indian conscience, and which has already been described in a masterly fashion, by M. Sylvain Lévi, writing of Açvaghosha (¹).

c10.66

¹ Loc. cit., pp. 73-74. — Since the above article was written Prof. RAPSON has been so good as to have copied by one of his pupils. Mr. W. H. B. THOMPSON, under his direction and for our use, the version of the Shaddantâvadâna from the Bodhisattvâvadâna-kalpalatâ, which is lacking in the Paris ms. (cf. above, p. 188, n. 3), according to the mss. Add. 1306 and 913 in the University Library at Cambridge. The kind communication of this copy has enabled us to prove the identity of this version — with the exception of three interpolations — with that of the Kalpadrumâvadâna. It appears that the author of the latter collection restricted himself to reproducing, without however (in any way) informing the reader of the fact, the work of Kshemendra, except that on two points he has lengthened the narrative of his predecessor, which in his opinion was too much abbreviated. This fact, however unexpected it may be, naturally does not change anything in our conclusions, as far as concerns the general chronology of the successive forms of the legend : it only causes us to think that the Kalpadrumâvadâna and Bodhisattvâvadâna-kalpalatâ agree in preserving for us the version of the canon of the Mûla-Sarvâstivâdins, which, as we know (cf. above, p. 151, n. 2), usually serves as a basis for the poetic lucubrations of Kshemendra. On the other hand, it supplies us with an excellent illustration justifying the reservations expressed above concerning the chronology of the texts : here, in fact, we are dealing with a well-known author, who wrote at the beginning of the XIth century, and who yet makes use of a version older than that of the Sûtrâlankâra. Thus it was wise on our part to consider as an acquired result only the demonstration of the difference of time between the stanzas of the Pâli Jātaka and their commentary. We are happy to be able on this last point to connect with the already cited evidence of M. SENART and Professor LÜDERS that of Prof. OLDENBERG (Nachrichten der k. Gesellschaft der Wissenschaften zu Göttingen, Phil.-hist. Klasse, 1911, pp. 441 sqq.).

c10.67

I. — From CUNNINGHAM, *Stūpa of Bharhut*, pl. XXVI, 2 ; for the description cf. above, pp. 39 and 194-5.

c10.68

II. — From a photograph taken by the author at the Madras museum in December 1896. The number and variety of the episodes collected together on this single medallion, among trees and rocks used as frames, give it, in contrast to the simplicity of that at Barhut, an especially entangled and confused appearance. — 1. On the lower part, to the right, we see the miraculous elephant with six tusks, standing between his two queens, of whom the first, on his left, holds over his head a parasol, whilst on the right the second flourishes a fly-flapper. — 2. He moves in the direction of the lotus pond, which occupies the bottom of the picture, and where we see him sporting with a numerous company ; the apparently female pachyderm who is coming precipitately out of the pond on the left and who then seems to crouch in order to throw herself down some precipice, would perhaps be intended to awake the remembrance of the jealous wife and her suicide? — 3. Whatever may be the fact concerning this detail, the story is now continued on the right, in the upper portion of the medallion. The great elephant is depicted standing at the moment when he crosses the fatal ditch in which lurks the hunter, whose bust only is to be seen between the animal's legs. — 4. A little more to the left the elephant, whose fore. part only is shown, is kneeling, in order that the hunter may cut off his tusks by the aid of a saw furnished with a curved spring, much more elaborate than the tool used at Barhut. — 5. Finally, right at the top, the latter carries away, on the two ends of a pole balanced on his right shoulder, the *spolia opima* of the Bodhisattva. It is curious to observe that the tusks are twelve in number, six (2×3) at each end of the pole! Here and there indications of antelopes and deer, while lending animation to the scene, only add to the crowding.

c10.69

1. AT BARHUT

c10.70

2. AT AMARĀVATĪ

c10.71

I. — A frieze from the Lahore museum (no. 1156. height m. 0,16), which formerly decorated one of the counter-steps of a staircase on Karamâr Hill; from a photograph taken by the author (cf. *Art g.-b. du Gandh.*, I, fig. 158). The elephant has only one pair of tusks. — 1. On the left he is wounded in the stomach by the arrow of the hunter hidden in a ditch. — 2. He then kneels down, to allow his teeth to be sawn off. — 3. Finally, on the right, the hunter, twice represented, brings back on his shoulder his bundle of ivory, and then offers it to the royal pair of Benares. We shall note the striking contrast between the distributive order of the episodes according to the ancient Indian school and according to the Indo-Greek school of Gandhâra, there crowded together inside the same pannel, here deployed one by one along a frieze.

c10.72

II. — From GRIFFITHS, *The Paintings in the Buddhist Cave Temples of Ajaṇṭā*, pl. 63 (fragment). For the description and interpretation of the attitude of the great white elephant cf. p. 195. The hunter is represented twice, first prostrate at the elephant's feet, with his head on the ground, then still squatting, but already balancing on his shoulder the double burden of tusks, which the magnanimous animal has just delivered to him.

c10.73

1 IN GANDHĀRA

c10.74

2. — AT AJANṬA

Linked texts (11)