^{page 1 note 1} See, for instance, Prinsep's Essays, vol. 2, Useful Tables, p. 165.Google Scholar This detail could probably have been settled, if we knew the time in A.D. 1822 which Crawfurd had in view when he said (as quoted by Turnour, , Mahāwanso, introd., p. 49Google Scholar) “the year 1822 was 2364 of the era in question”, and the time in A.D. 1830 which Upham had in view when he observed (as quoted ibid., p. 6) that the era had lasted 2372 years in A.D. 1830, and the time in A.D. 1837 which Turnour himself had in view when he told us (ibid., p. 28) that 2380 years had then passed away. However, other means of determining the point can, no doubt, be found, if it is thought worth while. But it may be added that Turnour, while placing the death in B.C. 543 and arranging all his dates B.C. on that basis, arranged his dates A.D. as if the death occurred in B.C. 544.

A modern date, apparently in A.D. 1884 or 1885, with full details for calculation, is to be found in an inscription set up at Bōdh-Gayā by a pilgrim from Ceylon: see Cunningham, , Mahābōdhi, p. 19,Google Scholar note 2. But it does not help us, since the details, as published, are not correct for either of those two years: either they were incorrectly recorded, or else they have not been correctly deciphered.

^{page 2 note 1} I use the name Mahāvaṁsa in the sense to which, I think, Geiger restricts it; namely, of that portion of the whole chronicle which ends at a point in chapter 37 after which we have the words Mahāvaṁsō niṭṭhitō, “the Mahāvaṁsa is finished.” In Turnour's text, those words come after verse 48: but there seems to be a lacuna of two verses at verse 12, which would raise the number of that verse to 50, in accordance with the numbering indicated by Geiger: see his Dīpavaṃsa und Mahāvaṃsa, p. 19;Google Scholar abstract translation in Indian Antiquary, 1906, p. 157.Google Scholar The commentary stops at the same point.

Turnour has told us (Mahāwanso, introd., p. 2Google Scholar) that the remainder of the chronicle is usually known as the Suluvaṁsa. This is the modern form (optionally SuỊuvaṁsa) of the term ChūỊavaṁsa (optionally Chūlavaṁsa) which occurs in chapter 99, verse 78, where Wijesinha's translation presents “the Lesser Dynasty”. It would be convenient if this name were generally adopted, so as to avoid having to speak of “the continuation of the Mahāvaṁsa”, “the later Mahāvaṁsa”, and so on.

It has been held open to question whether Mahānāma wrote the Suluvaṁsa as far as his own time; that is, to the end, or somewhere near the end, of chapter 38. But Geiger has shown reasons (loc. cit.) for believing that the remainder of chapter 37 must have been written after A.D. 1219. And the passage in chapter 99, verses 78–83, seems to indicate that the second instalment of the entire chronicle ran from the end of the Mahāvaṁsa to the end of the reign of a king Parakkamabāhu whose initial date has been placed by Turnour at some point after A.D. 1319, and by Wijesinha in A.D. 1295. The same passage tells us that the next instalment was written under the orders of Kitti-Sśri-Rājasīha, who began to reign in A.D. 1747 or 1748, and that it extended to his time.

Geiger has arrived (op. cit., pp. 37, 159, respectively) at the conclusion that the commentary on the Mahāvaṁsa was written between A.D. 1000 and 1250. I venture to suggest that it was a result of the restoration of Buddhism under Parakkamabāhu I, A.D. 1153–86.

^{page 3 note 1} See the Special Note A, p. 28 below.

^{page 3 note 2} Mahāwanso, introd., pp. 48, 50.Google Scholar

^{page 4 note 1} He seems to have announced this first in the Journal of the Asiatic Society of Bengal, vol. 23 (1854), p. 704:Google Scholar by an oversight, he used there the expression “Asoka's conversion”, instead of “inauguration ”, i.e. anointment.

^{page 4 note 2} See his Inscriptions of Asoka, Corpus Inscriptionum Indicarum, vol. 1 (1877), preface, pp. 3–9.Google Scholar

^{page 4 note 3} See his History of Ancient Sanskrit Literature (1859), p. 298Google Scholar f.; or, for a more recent statement of the same view, his Dhammapada, Sacred Books of the Bast, vol. 10, second edition (1898), introd., pp. 43–7.Google Scholar For the expression “best working hypothesis”, see Indian Antiquary, vol. 13 (1884), p. 149,Google Scholar where he added: “I still hold to it [the date B.C. 477], though not with such unreasoning pertinacity as to consider any modification of it impossible.”

^{page 5 note 1} The words in the text (Turnour, chapter 38, verse 59; p. 257, last line) are:—Datvā sahassaṁ dīpětuṁ Dīpavaṁsaṁ samādisi. Turnour translated:—”And that he [king Dhātusēna] might also promulgate the contents of the Dīpavaṁsa, distributing a thousand pieces, he caused it to be read aloud thoroughly”, and took the term Dīpavaṁsa as denoting here the Mahāvaṁsa itself. And Wijesinha repeated that rendering. But the meaning is:—”He bestowed a thousand (pieces of gold), and gave orders to write a dīpikā on the Dīpavaṁsa.” And that is the real reason why the arrangement of the Mahāvaṁsa follows so closely the arrangement of the Dīpavaṁsa.

That the pōrāṇā, ‘the ancients, the men of former times’, of the second line of the Mahāvaṁsa, are the persons who wrote up the Dīpavaṁsa from time to time, has already been suggested by Oldenberg (Dīpavaṁsa, , introd., p. 9Google Scholar).

^{page 6 note 1} See this Journal, 1906. 984–6.Google Scholar

^{page 6 note 2} By the expression “ordinary Indian lunar year”, I mean, of course, the Chaitrādi year, the year commencing with Chaitra śukla 1 (the first day of the bright fortnight of Chaitra), as contrasted with the Kārttikādi year commencing with Kārttika śukla 1.

It may help to elucidate matters if I give the names of the Hindū lunar months, with their ordinary equivalents for the period with which we are concerned: they are as follows:

It must be borne in mind, however, that the initial day, Chaitra śukla 1, is constantly receding in the solar year by one step of about eleven days or by two such steps, and is then advanced by one step of about nineteen days as the result of the intercalation of a month. The equivalents, therefore, are only approximate: and Chaitra śukla I might often fall in March, so that the month Chaitra would then not include any part of February, and the incidence of the other months would be pushed forward accordingly.

^{page 7 note 1} See the Vinayapiṭaka, ed. Oldenberg, , vol. 3, p. 283.Google Scholar

^{page 8 note 1} So, also, the Chullavarga of the Vinayapiṭaka, while placing the First Council in the Rains (11. 1. 3–5), and plainly in the first Rains after the death of Buddha, does not say anything as to the number of months which had intervened.

^{page 8 note 2} In the course of a brief discussion of my paper when it was read, Professor Jacobi reminded us of the conventional tendency of the Jain tradition to assign identical times to the leading events in the lives of the Tīrthaṁkaras. Thus, it was on various days when the moon was in Uttara-Phalgunī that Mahāvīra (1) descended from heaven and entered the womb of Dēvānandā, (2) was transferred to the womb of Triśalā, (3) was born, (4) went forth into the houseless life, and (5) attained kēvalajñāna: but he died when the moon was in Svāti: see, e.g., Sacred Books of the East, vol. 22, pp. 217–8, 264–5.Google Scholar

^{page 8 note 3} Manual of Indian Buddhism, p. 101.Google Scholar

^{page 8 note 4} Op. cit., p. 16, n. 2.

^{page 9 note 1} The Academy, 1 March, 1884,Google Scholar and Indian Antiquary, vol. 13 (1884), p. 148 ff.Google Scholar For some remarks by MrTakakusu, , see this Journal, 1896, p. 436 f.; 1897, p. 113.Google Scholar

^{page 9 note 2} Mr. Takakusu has explained this as the full-moon day of Āśvina, which is the seventh month in the Chaitrādi year. The pavāraṇā-ceremony would be performed either on that day or on the full-moon day of Kārttika, according as the vassa fell in the particular year. On the subject of the vassa, see p. 18 below.

^{page 10 note 1} Or, if we should look for some extraneous origin, it is not impossible that the belief in the full-moon day of Vaiśākha had its source in the fact (see p. 11 below) that that day was the day of the second anointment of Dēvānaṁpiya-Tissa, by whom Buddhism was established as the state religion of Ceylon. For another possible origin of the belief, see a suggestion mentioned on p. 16 below.

^{page 11 note 1} See the Special Note C, p. 32 below.

^{page 11 note 2} The Dīpavaṁsa says (11. 14) that it was in the season Hēmanta, in the second month, when the moon was in the Ashāḍhā nakshatra. The Mahāvaṁsa explains this by saying (Turnour, p. 70,Google Scholar last line) that it was:— Maggasira-māsassa ādi-chand-ōdayē dinē, “on the day of the month Mārgaśira Which has the initial rising of the moon.”

^{page 12 note 1} This expression may be thought somewhat self-contradictory. But it is frequently used in dealing with Indian dates; and it seems to convey best what is wanted here. It is, in fact, in accordance with the Indian custom as shown in inscriptions which present such terms as vartamāna, ‘being current’, in connection with figures for expired years.

^{page 12 note 2} Thus: let us assume B.C. 483 as the year in which Buddha died, on the full-moon day of Vaiśākha—(any other year would of course suit equally well for the purposes of this illustration; but we use the above-mentioned year because we shall take it again, definitely, farther on): then we have the following position:

Aśōka was anointed king of Northern India early in the bright fortnight of Jyaishṭha, in all probability on the day śukla 5, B.C. 265,–218 years after the death of Buddha.

^{page 14 note 1} See Julien, , Mémoires, vol. 1, pp. 334–5;Google ScholarBeal, , Si-yu-ki, vol. 2, p. 33;Google ScholarWatters, , On Yuan Chwang, vol. 2, p. 28:Google Scholar see also, Julien, , Vie, p. 131;Google ScholarBeal, , Life, p. 98.Google Scholar

^{page 14 note 2} See, e.g., Beal, , Si-yu-ki, vol. 1, p. 71.Google Scholar

^{page 14 note 3} This Journal, 1894. 533;Google ScholarEpigraphia Indica, vol. 9, p. 141.Google Scholar

^{page 14 note 4} Epi. Ind., vol. 1, p. 212.Google Scholar

^{page 14 note 5} Epi. Ind., vol. 8, p. 181;Google ScholarAnnual Report of the Archæological Survey of India for 1904–1905, p. 68.Google Scholar

^{page 14 note 6} See DrVogel's, note in this Journal, 1908. 971.Google Scholar

^{page 15 note 1} See the table at p. 27 below.

^{page 15 note 2} See note 1 on p. 12 above.

^{page 15 note 3} See this Journal, 1906. 657.Google Scholar

^{page 16 note 1} This Journal, 1906. 659,Google Scholar note.

^{page 17 note 1} See the text, ed. Childers, in this Journal, 1875. 72 ff.;Google Scholar translation by Davids, in Sacred Books of the East, vol. 11, p. 34 ff.Google Scholar

^{page 18 note 1} See, e.g., the Vinayapiṭaka, Mahāvagga, 3. 2. 2; 3. 3. 2.

^{page 20 note 1} For being able to present this result and some other equivalents of Indian dates mentioned in this paper, I am greatly indebted to Professor Jacobi. Neither do his Tables, published in the Epigraphia Indica, vol. 1, p. 403 ff.,Google Scholar and vol. 2, p. 487 ff., nor do those given by Messrs. Sewell and Dikshit in their Indian Calendar, reach back to the times with which we are concerned. But, in response to a request made by me, he very kindly arranged the bases which have enabled me to make the necessary calculations according to the lines of the First Ārya-Siddhānta, the Āryabhaṭīya of Āryabhaṭa (born A.D. 476), who was one of the oldest, if not actually the oldest, of the scientific Hindŭ astronomers. Without a larger knowledge of the exact manner in which the Hindŭs regulated their earliest calendar, we cannot, perhaps, arrive at once, by these means, at a final settlement of all the chronological items: but we may pave the way towards it; and the results exhibited on this occasion are correct for all present purposes.

^{page 22 note 1} We must exhibit the examination of this detail on some other occasion.

^{page 23 note 1} It must always be borne in mind that the years are lunar years of the luni-solar system, sometimes about eleven days shorter, sometimes about nineteen days longer, than a solar year.

^{page 23 note 2} It is always a moot-point whether such a statement means something in excess of 24 years but not enough so to be counted as 25, or something less than 24 years but not sufficiently so to be counted as only 23. I take it that the 24, 28, and 37 years assigned to Chandragupta, Bindusāra, and Aśōka, mean in each case something slightly in excess of the given number.

On the other hand, any such statement as 218 years after the death of Buddha means certainly some time in the 219th year current, after the completion of 218 years.

^{page 23 note 3} There is a story in Mysore (see, e.g., Inscriptions at Śravaṇa-BeỊgoỊa, Epigraphia Carnatica, vol. 2, introd., p. 3 ff.),Google Scholar which has been accepted quite seriously, that Chandragupta abdicated, became a Jain monk, went to Southern India with the Śruta-Kēvalin Bhadrabāhu, and died at Śravaṇa-BeỊgoỊa. The story, however, is only presented in a compendium of Jain history, called Rājāvalīkathe, which was composed in the last century. And, when we examine it, we find (see Ind. Ant., 21, 1892. 156 ff.)Google Scholar that it really indicates, not Chandragupta the grandfather of Aśōka, but an otherwise unknown Chandragupta, son of Aśōka's alleged son Kunāla; that he abdicated in favour of an otherwise unknown son named Siṁhasēna; and that the Bhadrabāhu who figures in it is, not the Śruta-Kēvalin of that name, but quite a different person, the pontiff Bhadrabāhu II. The story is probably of quite modern invention. If there is anything early about it, it rests upon certain inscriptions which do assign a disciple named Chandragupta to Bhadrabāhu II, and upon a reminiscence of the abdication of Aśōka, which must have become known in Mysore through the publication of the Last Edict there. From any point of view, it has not the slightest historical value as affecting Chandragupta the grandfather of Aśōka.

^{page 24 note 1} There are the variants and . It is easy to see, from coins of the Kanishka series, how the λλ arose from a loosely formed cursive M.

^{page 24 note 2} The word naptṛi is not as explicit as pautra, ‘son's son,’ and may denote either a son's son or a daughter's son.

^{page 25 note 1} In the introduction to his Samantapāsādikā: see the Vinayapiṭaka, ed. Oldenberg, vol. 3, pp. 321, 299.Google Scholar

^{page 25 note 2} With this second appellation, taking officially (as is evident from the Greek statements) the place of the real personal name, compare Dēvānāṁpriya and Priyadarśin in the case of Aśōka: see this Journal, 1908. 482–5Google Scholar.

^{page 26 note 1} See the Special Note A, p. 28 below.

^{page 26 note 2} See the Special Note B, p. 31 below.

^{page 29 note 1} Professor Norman has indicated (this Journal, 1908. 14Google Scholar) that the correct reading is sāṭṭhārasaṁ, instead of the aṭṭhārasaṁ given by Turnour. I amend Turnour's vassa-sataṁ dvayaṁ into vassa-sata-dvayaṁ, and his ēkarajja-mahāyasō into ēkarajjaṁ mahāyasō, in accordance with the Colombo text edited by Sumangala and Batuwantudawa.

^{page 29 note 2} I will comment, on some other occasion, on this statement, based on the Dīpavaṁsa, 6. 22, which, admissible enough from a Purāṇic point of view, reads somewhat strangely in what purports to be a sober chronicle of facts: also on the verse in the Dīpavaṁsa, 6. 24, which has been supposed to say that Aśōka was anointed at 20 years of age, or when Mahēndra was 20 years old. Meanwhile, I may say that, originally, verse 22 said that Aśōka slew his brothers one by one, and verse 24 told us that he proposed to anoint Mahēndra as Yuvarāja when the latter was 20 years old, but Mahēndra preferred the religious life.

^{page 30 note 1} Kings have been ‘anointed’ (not ‘consecrated’ or ‘inaugurated’) from time immemorial to the present day. And the term applies to the Hindū process, as much as to any other, in view of the fact that the ingredients of the liquid which was poured over the king, from a jar or a conch-shell as the case might be, included ghee (sarpis, ghṛita, ājya): see, e.g., Bṛihat-Saṁhitā, 48/47. 50, below his translation of which Kern has remarked, apropos of the use of snāna, ‘bathing, washing’, in the title of the chapter, pushyasnāna:—”In so much does the ceremony more resemble an anointment than a washing.”

^{page 31 note 1} Text (Delhi-Siwālik version) in Ind. Ant., vol. 18, p. 74;Google ScholarEpi. Ind., vol. 2, p. 257.Google Scholar

^{page 31 note 2} Texts in Archœol. Surv. Southern India, vol. 1, p. 127;Google ScholarInd. Ant., vol. 19, p. 84.Google Scholar

^{page 31 note 3} Texts, loc. oit. in each case, pp. 129, 98.Google Scholar

^{page 32 note 1} Bombay text, 2 (Ayōdhyōkāṇḍa), § 4, verses 21, 22: and compare verse 2, and § 7, verse 11. According to the Bengal recension, the ultimate anointment of Rāma also was under Pushya: Gorresio, 6 (Yuddhakāṇḍa), § 112, verse 70.

^{page 33 note 1} See, for instance, Aitarēya-Brāhmaṇa, book 8, chapters 2–4; Śatapatha-Brāhmaṇa, book 5, adhyāyas 2–4; also Bṛihat-Saṁhitā, chapter 48/47. On the whole topic see the article abhishēka in Goldstücker's Sanskṛit Dictionary.

^{page 33 note 2} Against the distinct evidence of the inscriptions: see this Journal, 1908. 496.Google Scholar

^{page 34 note 1} For the use of the conch-shell in anointments, see, e.g., the Mahābhārata, 12 (Śāntiparvan), Calcutta text, § 40, verse 1457:—“Then Dāśārha (Kṛishṇa) rose, and, taking a filled conch, anointed (abhyashiñchat) the lord of the earth, Yudhishṭhira, son of Kuntī.” I follow the Kumbhakōṇam text, § 39, verse 15, in taking pūritam, ‘filled’, instead of pūjitam, ‘worshipped’.

^{page 34 note 2} Buddhaghōsha also quotes this postscript in his Samantapāsādikā (Vinayapiṭaka, ed. Oldenberg, 3. 323Google Scholar): he represents Dēvānaṁpiya-Tissa as remembering the admonition of Aśōka after his meeting with Mahēndra. The expression ēkamābhisittō, which we have just after that, cannot be properly amended into ēkamās-ābhisittō, at any rate in the sense of “one-month-anointed”, since the second anointment was five and a half months after the first: it seems to stand for something meaning “(was) at once anointed”.