is it pure fiction. It is good evidence of opinion as held at the time when it was written. And from the fact that such an opinion was then held we can argue back, according to the circumstances of each case, to what was probably the opinion. held at some earlier date. No hard words are needed and we may be unfeignedly grateful to these old students and writers for having preserved as much as we can gather from their imperfect records.1

It may be asked, perhaps, why we do not try to save the intellectual effort necessary to balance probabilities in later accounts that cannot be entirely trusted, by confining ourselves exclusively to the contemporary documents, the inscriptions? The answer is that such a method would be absurd; it would not even save trouble. The inscriptions. are scanty. The text of all of them together would barely occupy a score of these pages. They give only a limited view of the set of circumstances they deal with. Royal proclamations, and official statements, are not usually regarded as telling the truth, the whole truth, and nothing but the truth. To put it mildly, there is an economy of candour in these documents, intensely interesting though they are. And they are enigmatic. It is not possible to understand them without the light thrown upon them by the later accounts. It would only add to their difficulty to reject, for instance, the identification of the Piyadassi of the inscriptions

1 See now on these Chronicles Professor Geiger's important researches in his Dipavansa und Mahāvamsa. Erlangen, 1902.

with the Asoka of the literature, or the fact of his relationship to Chandragupta, or of his capital having been at Paṭaliputta, or any other of the numerous side-lights to be drawn from the Chronicles. As M. Senart says:

"I believe that the Chronicles have, in certain details, under the name of Asoka, preserved of our Piyadasi recollections sufficiently exact, not only to allow a substantial agreement (une concordance sensible) to appear, but even to contribute usefully to the intelligence of obscure passages in our monuments."'

Besides numerous passages scattered through other books (which have not yet been collected) we have four connected narratives dealing with Asoka. These are:

(1). The Asoka Avadāna, in Buddhist Sanskrit, preserved in Nepal.

(2). The Dīpavamsa, in Pāli, preserved in Burma. (3). Buddaghosa's account in his commentary on the Vinaya.

(4). The Mahāvamsa, in Pali, preserved in Ceylon. Of these the first was composed in the Ganges valley. The author and date are unknown; but it is probably as late as the third century of our era. It forms one of a collection of legends called the Divyāvadāna. The exact force of this title is somewhat ambiguous. Avadāna means a story, but as it is used exclusively of the life-story of a person distinguished in the religion, the collection corresponds to the Vita Sanctorum of the Christian Church. We 1 Inscription de Piyadasi, 2. 231.

know so little, as yet, of the literature in Buddhist Sanskrit that we cannot form any clear idea of the method by which the tradition it has preserved was handed down.

It is otherwise with the other three. We know that there were two great monasteries at Anurādhapura in Ceylon, the Great Minster and the North. Minster. There the canonical books were handed down, in Pali; and commentaries upon them, in Sinhalese, interspersed with mnemonic verses in Pali. In the third century of our era some one collected such of these Pali verses as referred to the history of Ceylon, piecing them together by other verses to make a consecutive narrative. He called his poem, thus constructed, the Dipa-vamsa, the Island Chronicle. The old verses were atrocious Pali, and the new ones added are not much better. Then, as the old ones were taken, not from one commentary only, but from several, we get the same episode repeated in different verses. Added to this the work was supplanted in Ceylon by the much better-written book called the Maha-vamsa, or Great Chronicle; and was completely lost there. The present text, which is corrupt, has been restored, in the excellent edition by Professor Oldenberg, from MSS., all of which are derived from a single copy that had been preserved in Burma.

Shortly after the Island Chronicle was composed, the celebrated Buddhaghosa, a brahmin from Behar, came over to Ceylon, and rewrote in Pali the old Sinhalese commentaries. His work supplanted the latter, which are now lost, and is the only evidence

He

we have of the nature of the ancient tradition. quotes, from the old Sinhalese commentary, a number of the mnemonic verses also contained in the Island Chronicle, and gives us, in Pali, the substance of the Sinhalese prose with which they had originally been accompanied.

A generation afterwards Mahānāma wrote his great work, the Maha-vamsa. He was no historian, and had, besides the material used by his two predecessors, only popular legends to work on. But he was a literary artist, and his book is really an epic poem of remarkable merit, with the national idol, Dushṭa Gāmini, the conqueror of the invading hosts of the Tamils, as its hero. What he says of other kings, and of Asoka amongst them, is only by way of introduction, or of epilogue, to the main story.

I have compared historically the various versions of one episode in these and other narratives (that of Asoka and the Buddha relics),' and have shown how interesting are the results to be derived from that method. To retell such an episode in one's own words may be a successful literary effort, but it would be of no historical value. It would give us merely a new version, and a version that had not been believed anywhere, at any time, in India. By the historical method, a few facts of importance may yet be gathered from amidst the poetical reveries of these later authors.

So, for instance, the tradition-Indian of course in origin, but preserved in Nepal-states that Asoka's mother was the daughter of a brahmin living in 1 7. R. A. S., 1901, pp. 397–410.