Here we have a simple account of Buddhaghosha1 and his literary labours written by a man, himself a priest, and who may well have known Buddhaghosha during his stay in Ceylon. It is true that the statement of his writing the same book three times over without a single various reading, partakes a little of the miraculous; but we find similar legends mixed up with accounts of translations of other sacred books, and we cannot contend that writers who believed in such legends are therefore unworthy to be believed as historical witnesses.

The next question which has to be answered is this, Did Buddhaghosha's Parables, and the whole of the commentary in which they are contained, form part of the 'Arthakathâ' which he translated from Singhalese into Pâli. The answer to this question depends on whether the Dhammapada formed part of the 'Pitakattaya' or not. If the verses of the Dham

1 The Burmese entertain the highest respect for Buddhaghosha. Bishop Bigandet, in his Life or Legend of Gaudama' (Rangoon, 1866), writes: "It is perhaps as well to mention here an epoch which has been, at all times, famous in the history of Budhism in Burma. I allude to the voyage which a Religious of Thaton, named Budhagosa, made to Ceylon, in the year of religion 943 400 A.C. The object of this voyage was to procure a copy of the scriptures. He succeeded in his undertaking. He made use of the Burmese, or rather Talaing characters, in transcribing the manuscripts, which were written with the characters of Magatha. The Burmans lay much stress upon that voyage, and always carefully note down the year it took place. In fact, it is to Budhagosa that the people living on the shores of the Gulf of Martaban owe the possession of the Budhist scriptures. From Thaton, the collection made by Budhagosa was transferred to Pagan, six hundred and fifty years after it had been imported from Ceylon."

mapada were contained in the canon, then they were also explained in the Singhalese 'Arthakathâ,' and consequently translated from it into Pâli by Buddhaghosha. Now it is true that the exact place of the Dhammapada in the Buddhistic canon has not yet been pointed out; but if we refer to Appendix iii., printed in Turnour's edition of the 'Mahavansa,' we there find in the third part of the canon, the Sûtra-pitaka, under No. 5, the Kshudraka-nikâya, containing fifteen subdivisions, the second of which is the Dhammapada.

We should, therefore, be perfectly justified in treating the parables contained in Buddhaghosha's Pâli translation of the Arthakathâ,' i. e. the commentary on the Dhammapada, as part of a much more ancient work, viz. the work of Mahinda, and it is only in deference to an over-cautious criticism that I have claimed no earlier date than that of Buddhaghosha for these curious relics of the fable-literature of India. I have myself on a former occasion1 pointed out all the objections that can be raised against the authority of Buddhaghosha and Mahinda; but I do not think that scholars calling these parables the parables of Mahinda, if not of Buddha himself, and referring their date to the third century B.C., would expose themselves at present to any formidable criticism.

If we read the pages of the 'Mahâvansa' without prejudice, and make allowance for the exaggerations and superstitions of Oriental writers, we see clearly that the literary work of Buddhaghosha presupposes the existence, in some shape or other, not only of the canonical books, but also of their Singhalese commentary. The Buddhistic canon had been settled in seve

1 'Chips from a German Workshop,' 2nd ed., vol. i. p. 197.

b

ral councils, whether two or three, we need not here inquire. It had received its final form at the council held under Asoka in the year 246 B.C. We are further told in the 'Mahâvansa' that Mahinda, the son of Asoka, who had become a priest, learnt the whole of the Buddhist canon in three years (p. 37); and that at the end of the third council he was dispatched to Ceylon, in order to establish there the religion of Buddha (p. 71). The king of Ceylon, Devânâmpriya Tishya, was converted, and Buddhism soon became the dominant religion of the island. Next follows a statement which will naturally stagger those who are not acquainted with the power of memory if under strict discipline for literary purposes, but which exceeds by no means the limits of what is possible in times when the whole sacred literature of a people is preserved and lives by oral tradition only. The Pitakatraya, as well as the Arthakathâ, having been collected and settled at the third council in 246 B.C., were brought to Ceylon by Mahinda, who promulgated them orally; the 'Pitakatraya' in Pâli, and the 'Arthakathâ' in Singhalese,3 together with additional Arthakathâ of

1 The question of these councils and of their bearing on Indian chronology has been discussed by me in my 'History of Ancient Sanskrit Literature,' p. 262 seq., 2nd ed.

2 Cf. Bigandet, 1. c. p. 387.

3 Singhalese, being the language of the island, would naturally be adopted by Mahinda and his fellow-missionaries for communication with the natives. If he abstained from translating the canon also into Singhalese, this may have been on account of its more sacred character. At a later time, however, the canon, too, was translated into Singhalese, and, as late as the time of Buddhadâsa, who died 368 A.D., we read of a priest, profoundly versed in the doctrines, w ho translated the Sûtras, one of the three divi

his own. It does not follow that Mahinda knew the whole of that enormous literature by heart, for, as he was supported by a number of priests, they may well have divided the different sections among them. The same applies to their disciples. But that to the Hindu mind there was nothing exceptional or incredible in such a statement, we see clearly from what is said by Mahânâma at a later period of his history. When he comes to the reign of Vattagâmani,1 88-76 B.C., he states: "The profoundly wise priests had heretofore orally perpetuated the Pâli Pitakatraya and its Arthakatha (commentaries). At this period these priests, foreseeing the perdition of the people (from the perversions of the true doctrines) assembled; and in order that the religion might endure for ages, recorded the same in books."2

Later than this date, even those who doubt the sions of the Pitakatraya, into the Sihala language. (Mahâv. p. 247.) A note is added, stating that several portions of the other two divisions also of the Pitakatraya have been translated into the Singhalese language, and that these alone are consulted by the priests who are unacquainted with Pâli. On the other hand, it is stated that the Singhalese text of the Arthakathâ exists no longer (see Spence Hardy, 'Legends,' p. xxv., and p. 69). He states that the text and commentary of the Buddhist canon are believed to contain 29,368,000 letters. (Ibid. p. 66.)

1 See Bigandet, 1. c. p. 388.

"After the Nir

2 See also Spence Hardy, 'Legends,' p. 192. vâna of Buddha, for the space of 450 years, the text and commentaries, and all the works of the Tathâgata, were preserved and transmitted by wise priests, orally, mukha-pâthena. But having seen the evils attendant upon this mode of transmission, five hundred and fifty arhats, of great authority, in the cave called Aloka (Alu) in the province of Malaya, in Lankâ, under the guardianship of the chief of that province, caused the (sacred) books to be written." (Extract from the 'Sâra-sangraha.')

powers of oral tradition have no right to place the final constitution of the Buddhistic canon and its commentaries in Ceylon, nor is there any reason to doubt that such as these texts existed in Ceylon in the first century B.C., they existed in the fifth century after Christ, when the commentaries were translated into Pâli by Buddhaghosha, and that afterwards they remained unchanged in the MSS. preserved by the learned priests of that island. It is easy to shrug one's shoulders, and shake one's head, and to disbelieve everything that can be disbelieved. Of course we cannot bring witnesses back from the grave, still less from the Nirvâna, into which, we trust, many of these ancient worthies have entered. But if we are asked to believe that all this was invented in order to give to the Buddhistic canon a fictitious air of antiquity, the achievement would, indeed, be one of consummate skill. When Asoka first met Nigrodha, who was to convert him to the new faith, we read (p. 25), that having refreshed the saint with food and beverage which had been prepared for himself, he interrogated the sâmanera on the doctrines propounded by Buddha. It is then said that the sâmanera explained to him the Apramâda-varga. Now this Apramâda-varga is the title of the second chapter of the Dhammapada. Its mention here need not prove that the Dhammapada existed previous to the Council of Asoka, 246 B.C., but only that Mahânâma believed that it existed before that time. But if we are to suppose that all this was put in on purpose, would it not be too deep-laid a scheme for the compiler of the Mahâvansa ?1

And for what object could all this cunning have 1 In the account given by Bishop Bigandet (p. 377) of the first