^{page 983 note 1} See his Manual of Indian Buddhism, p. 76.Google Scholar

^{page 983 note 2} See, e.g., Vinayapiṭaka, ed. Oldenberg, Mahāvagga, 5. 13. 1: and compare Mahāparinibbānasutta, ed. Childers, this Journal, 1876. 250, line 7; Suttanipāta, ed. Fausböll, 15, line 14; 103, line 2; and verses 274, 1003.

^{page 983 note 3} Suttanipāta, verse 454/455: I quote verbatim Fausböll's translation in SBE, 10. 73.

^{page 984 note 1} Milindapañha; translation by Davids, SBE, 36. 374.

^{page 984 note 2} See his epitaph at Śravaṇa-Belgola, , edited by me, Epi. Ind., 5. 180.Google Scholar

^{page 984 note 3} In respect of these figures, there has now been revived a proposal which had almost been forgotten and might well have been left unrecalled. MrSmith, Vincent, in his Asoka, 1st ed., pp. 139, 141,Google Scholar wisely, though doubtfully, followed the opinion that they denote the number of years elapsed since the death of Buddha. In his recently published 2nd edition, however, he has followed other leads, and, abandoning that interpretation, has suggested on p. 150 that they may mean “256(?) departures from staging-places (or possibly, days spent abroad)”, but on p. 152 that the whole clause may mean “the precept quoted above was preached by?(me) on tour 256 [?times]”. And to the latter interpretation he has attached a footnote, in the course of which he observes that “the Hindus would consider 256 = 16² = 32 x 8 = 64 x 4 to be a ‘perfect number’.” He has not, however, understood what a “perfect number” is. “Perfect numbers [I quote the definition from the first book that I have at hand, Gow's, History of Greek Mathematics (1884), p. 70:Google Scholar but we need not look about for any better authority] are those which are equal to the sum of all their possible factors (e.g. 28 = 1 + 2 + 4 + 7 + 14).” But all the possible factors of 256 are 1, 2, 4, 8, 16, 32, 64, and 128: the sum of them falls short by one; it only amounts to 255: and any Hindū, who had the idea of “perfect numbers” at all, would certainly have detected that.

We need say no more (see this Journal, 1908, 819 ff.Google Scholar) about the idea that the figures 256 can here denote “departures from staging-places” or “days spent abroad”. The other idea, that they may mean “256 times”, has simply the following genesis. In the Mysore versions the Last Edict is followed by another, which begins, in the Brahmagiri text, sē hēvaṁ Dēvāṇaṁpiye; the word sē standing immediately after the figures. Mr. Rice offered us, in 1892, a text and translation, in a report or pamphlet entitled “Edicts of Aśōka in Mysore”. After the word sē he inserted a mark of punctuation, though the original has none. He Sanskṛitized the word as śaḥ = śas, a suffix which gives, e.g., ēkaśas, ‘one by one, singly’, śataśas, ‘in a hundred ways, by or in hundreds’, and bahuśas, ‘in many ways’. He confused the śas with kṛitvas, with which we have, e.g., ēkakṛitvas, ‘once’, śatakṛitvas, ‘a hundred times’, and bahukṛitvas, ‘many times’. And so he arrived at the rendering:—“This exhortation has been delivered by the vyūtha (or ? society) 256 times.” He abandoned that rendering in his later treatment, given in Epi. Cam., vol. 11, Chitaldroog (1903), translations, p. 93, No. 21. But the influence of it has survived.Google Scholar

^{page 985 note 1} See the text, ed. Childers, , this Journal, 1876. 249;Google Scholar and compare the translation by Davids, , SBE, 11. 108.Google Scholar In the third line of the verse the words are vassāni paññāsa samādhikāni. The last of them is taken as samā or sama + adhikāni, ‘more by one year:’ but it is possible that it stands by metrical license for samadhikāni, ‘having something more’; i.e. “for fifty years with somewhat more”. The translation refers us to the Jātaka, ed. Fausböll, , 2. 383Google Scholar, where the same expression is used, but in a different connexion. There we have the various reading samadhikāni, to be scanned with the a in the second syllable treated as ā: and the commentary seems on the whole to explain the words as meaning atirēka-paññāsa-vassāni, “fifty years and somewhat more”. With samadhika in that sense, compare sātilēka = sātirēka in line 1 of the Rūpnāth text, p. 1013 below: in the Sahasrām text the word is sādhika: the Brahmagiri text uses adhika and sātirēka.

^{page 986 note 1} Bṛihaj-Jātaka, Bombay text, 1882, chap. 7, verse 5. We may note that he gives 32 years for horses, 25 for mules or donkeys and camels, 24 for bulls and buffalos, 12 for dogs, and 16 for goats and such animals.

^{page 987 note 1} Prabandhachintāmaṇi; translation by Tawney, , 150.Google Scholar

^{page 987 note 2} The seventh pillar-edict supports this to the extent at any rate that he reigned on into the twenty-eighth year after his anointment. The Vāyu, Matsya, and Brahmāṇṭa Purāṇas assign to him a reign of thirty-six years.

^{page 987 note 3} See this Journal, 1908. 486–98, and compare 822.Google Scholar

^{page 988 note 1} This Journal, 1908. 811.Google Scholar

^{page 988 note 2} We have to rely principally on the Sahasrām, Rūpnāth, and Brahmagiri texts. The other three, at Bairāt, Siddāpura, and Jaṭṭiṅga-Rāmēśvara, are so much damaged that, to avoid introducing too many details, it is better not to quote them, except when they may be needed to supplement, or may be found to differ from, our principal guides.

For the latest published full critical treatments of the record, see Professor Bühler's articles in Ind. Ant., 22 (1893). 299Google Scholar, for the Sahasrām, Rūpnāth, and Bairāt texts, and Epi. Ind., 3. 138, for the Mysore texts.Google Scholar

The Sahasrām text is available, so far, only from the lithograph given by SirCunningham, A. in his Inscriptions of Asoka, Corpus Inscriptionum Indicarum, vol. 1, plate 14Google Scholar, reproduced with ProfessorBühler's, article in Ind. Ant., 6, 155Google Scholar, and from a facsimile, published with his article in Ind. Ant., 22. 299Google Scholar, of an ink-impression supplied by me. Sir A. Cunningham's lithograph was worked up by hand, and is consequently not a facsimile. But it was based on, or revised according to, a photograph of the original: see op. cit., 21. And it has a special value because the original suffered some appreciable damage in between the times when the photograph was taken and my ink - impression was made.Google Scholar

For the Rūpnāth text we have at Ind. Ant., 6. 156Google Scholar, a facsimile of a good ink-impression supplied by SirCunningham, A., and at Ind. Ant., 22. 299, a facsimile of another good ink-impression supplied by me.Google Scholar

For the Mysore texts we have at Epi. Ind., 3. 138, 140Google Scholar, facsimiles of good ink - impressions supplied by Professor Hultzsch. Mr. Rice's lithographs in Epi. Carn., 11, Chitaldroog, at pp. 162, 164, 168,Google Scholar of the texts in roman characters, were worked up by hand, and are not facsimiles of the originals. See a remark on them made by me in this Journal, 1908. 815, note 2. Regarding two particular details in the Brahmagiri lithograph, see p. 999 below, note, and p. 1012.Google Scholar

^{page 989 note 1} For his translations see Ind. Ant., 6. (1877). 156Google Scholar; Epi. Ind., 3. 140.Google Scholar

^{page 990 note 1} For his translations see Inscriptions de Piyadasi, 2 (1886). 195Google Scholar; Ind. Ant., 20 (1891). 165Google Scholar; Journal Asiatique, 1892, 1. 487.Google Scholar

² The Dīpavaṁsa does not present this particular statement: it only says (3. 57, 58) that Buddha was thirty-five years old when he began to preach “his eternal truth”, and that Bimbisāra was then thirty. But the period follows naturally from that statement combined with what we have in the Mahāparinibbānasutta: see p. 985 above.

^{page 991 note 1} Text, ed. Fausböll, , verses 405–24Google Scholar: translation by the same scholar in SBE, 10. 66–8Google Scholar. I have cited some passages from the Sutta in this Journal, 1904. 24, 25.Google Scholar

^{page 991 note 2} Pischel, , Grammatik der Prākrit-Sprachen, § 132.Google Scholar

^{page 991 note 3} Pischel, , § 366aGoogle Scholar; Childers, , Pali Dictionary, under parakkamo.Google Scholar

^{page 992 note 1} Mahāparinibbānasutta (1875), 52, line 26.

^{page 992 note 2} The word Bhagavat, as applied to Buddha, has been usually rendered by “the Blessed One”. I prefer to abandon that expression.

^{page 992 note 3} Id. (1876), 236, line 20.

^{page 992 note 4} Suttanipāta, 92, line 1.

^{page 992 note 5} Id., 184, verse 1010.

^{page 992 note 6} For Bühler's translation, see Epi. Ind., 2. 468.Google Scholar

^{page 992 note 7} See ibid., 469.

^{page 992 note 8} Ibid., 455, 459, and plates: Archœol. Surv. South. Ind., 1. 118, 120, and plates.Google Scholar

^{page 992 note 9} Ibid., 455, 459, and plates.

^{page 993 note 1} A verb prākram itself (pra + ā + kram) does not seem to occur.

^{page 993 note 2} Pischel, , § 148.Google Scholar

^{page 993 note 3} Childers, , under aggaṁ.Google Scholar

^{page 993 note 4} Gray, , Indo-Iranian Phonology, § 12.Google Scholar

^{page 994 note 1} Ancient Coins and Measures of Ceylon (1877), p. 58.Google Scholar

^{page 994 note 2} See my remarks in this Journal, 1908. 816 f.Google Scholar

^{page 995 note 1} See Epi. Ind., 3. 138, and plate.Google Scholar

^{page 996 note 1} Mahāparinibbānasutta, this Journal, 1875. 49, line 10.Google Scholar

^{page 996 note 2} Ind. Ant., 19. 125, and plate.Google Scholar

^{page 997 note 1} Archœol. Surv. South, Ind., 1. 125, line 1; 127, line 1.Google Scholar

^{page 997 note 2} There is no valid reason for thinking that Aśōka's sovereignty included Mysore. On the contrary, this preamble if distinct evidence that that was not the case. We know that he possessed Kāṭhiāwāḍ, and conquered the Kaliṁga countries. But it is improbable that, except in Kaliṇga, his dominions extended anywhere to the south of the Narbadā; unless, perhaps, they included the strip of country between the Western Ghauts and the sea, from the mouth of the Narbadā down to near Bombay. The value of the existence of a remnant of the eighth rock-edict at Sopārā in the Ṫhāṇa District has still to be weighed.

^{page 997 note 3} Compare my remarks in this Journal, 1908. 497.Google Scholar

^{page 997 note 4} Mr. Rice, indeed, would find in the word Isila the origin of the siddā of Siddāpura: see Epi. Carn., 11, Chitaldroog (1903), introd., p. 3,Google Scholar and his Mysore and Coorg from the Inscriptions (1909), p. 11, and preface, p. 11. We need not make any comments on this remarkable proposition.Google Scholar

^{page 998 note 1} On the Jaṭṭiṁga-Rāmēśvara hill there are also other records, two of which—Western Chālukya inscriptions of A.D. 1064 and 1072—have been edited by me in Epi. Ind., 4. 212, 214.Google Scholar

^{page 998 note 2} In opposition to my identification we are now told (Smith, V., Asoka, 2nd edition, 94,Google Scholar note) that there is a “Songîr” in Khāndēsh, “with an old fort,” and “a Songaṛh, the early Gaikwâr capital, with ‘vast ruins,’ in Baroda”: the said ruins dating from quite late times. So there are. And we could indicate various other places with the same or similar names: for instance, there is Suvarṇdurg, also with a fort, in the Ratnāgiri District, much nearer to Mysore. But what does it all matter? We have a Suvarṇagiri itself in the heart of Aśōka's dominions. Why should we look about for any other place with even the same name, much less a different one?: except, of course, from the point of view, which has produced confusion in so many directions in our inquiries into the ancient history of India, of seeking every opportunity to leave the track of common sense in order to find any explanations rather than natural ones.

^{page 998 note 3} Girivraja was the city inside the hills; Rājagṛiha was just outside them, on the north: see this Journal, 1907. 360–2.Google Scholar

^{page 999 note 1} The lithographs given by MrRice, in Epi. Carn., 11, Chifcaldroog, plates at pp. 162, 164,Google Scholar of the texts in roman characters, are mostly in agreement with rny reading: but they are not facsimiles; and after the khō they give us saṁvachhareṁ, the termination of which is strongly suggestive of modern Marāṭhī. His reading in the Brahmagiri text is saṁvachharaṁm, which is not in accordance with his lithograph, and does not even match the sātirēke which he has given correctly.

^{page 1000 note 1} In actual illustration of the two possibilities we may cite chha-vassa from the Dīpavaṁsa, 7. 22, and chhab-bassa from the Mahāvaṁsa, 2. 27. For the Mahāvaṁsa I am of course using Professor Geiger's invaluable text.

^{page 1000 note 2} Here we may quote chha-māsa from Dīpavaṁsa, 11. 14, and chammāsa from Mahavaṁsa, 13. 5.

^{page 1000 note 3} It might perhaps be said that we could translate “a certain period of six years”. But the use of ēka in that manner seems to be confined to somewhat late literature: see Whitney, , Sanskrit Grammar, § 482, cGoogle Scholar; Speijer, , Sanskrit Syntax, § 281.Google Scholar

^{page 1000 note 4} I quote from the latest published readings, those by Senart, M. in Epi. Ind., 8. 59 ff.Google Scholar

^{page 1001 note 1} Archœol. Surv. West. Ind., 4. 113, No. 21, and plate.Google Scholar

^{page 1001 note 2} Epi. Ind., 2. 242, and plate.Google Scholar

^{page 1001 note 3} Compare, e.g., my remarks in this Journal, 1908. 496, and 819, note.Google Scholar

^{page 1002 note 1} See this Journal, 1907. 521, note 2.Google Scholar

^{page 1002 note 2} Speijer, , Sanskrit Syntax, § 87.Google Scholar

^{page 1002 note 3} Epi. Ind., 8. 168, line 9 f., and plate.Google Scholar

^{page 1002 note 4} The practice of punctuating the texts of ancient unpunctuated inscriptions cannot be too strongly deprecated. There is, indeed, no objection to doing it when we have to separate metrical passages from prose: there can hardly be any questions of doubt in such a case as that. But the matter is very different as regards the prose passages: to treat them in the same manner, by breaking up a text into clauses which are not marked in the original, prejudges the case at once: it lays before us a text shaped as the individual translator has understood it, which may be very different from the manner in which it was intended to be understood; and it creates an additional and gratuitous obstacle to a reconsideration of the meaning. I do not claim to have never myself committed the fault which I deprecate: but I do not think I have ever carried the practice to the length to which it has been taken in Dr. Vogel's otherwise excellent transcription of the Sārnāth text, and in some of the previously given texts of the Last Edict.

^{page 1003 note 1} Vanaparva: Calcutta text, verse 8277; Kumbakonam text, § 84, verse 2.

^{page 1003 note 2} The discrepancies in the recensions of this passage are very marked. The same feature, however, both in words and in phrases, as well as in actual forms, runs more or less through all the records of Aśōka, and suggests that perhaps in not a single case have we any of his proclamations as they were actually dictated by him, or even framed in his secretariat. It would seem as if copies framed there were sent out to other offices, and fresh drafts were then made to suit local peculiarities.

^{page 1003 note 3} It may be noted that an essential word, kaṭe or sāvāpite, is wanting in this version.

^{page 1005 note 1} This explanation of satā, probable enough in itself, is now abandoned. As regards another view, of recent date, the words sata, satā, could no doubt stand for satta, sattā, of satta as = sattra, ‘a board-and-lodging house, a hospice’: but the proposal to interpret them here in that manner need not be seriously noticed again. They might also stand for sattā, ‘existence’; or for satta, sattā, of satta as = śakta, ‘able, strong’; or, again, for saṁa, saṁtā, of saṁta as = śrānta, ‘weary’: but these possibilities, too, may be at once put aside.

^{page 1006 note 1} Turnour, , Mahāvanso, 250, line 12; 255, line 3.Google Scholar

^{page 1006 note 2} The term dharmapada, dhammapada, seems to be a difficult one to translate; chiefly as regards the second component. Max Müller favoured “path of virtue” or “path of the law” (SBE, 10, 2nd ed., preface, 10), and translated appamādo amatapadaṁ, in Dhammapada, verse 21, by “earnestness is the path of immortality”. Childers, in his Pali Dictionary, 269b, rendered the same expression by “diligence is the way of Nirvāṇa”.

^{page 1007 note 1} I quote verbatim Müller's, Max translation in SBE, 10 (2nd ed.), 87.Google Scholar

^{page 1007 note 2} Mahāparinibbānasutta; this Journal, 1876. 252Google Scholar. ProfessorDavids, Rhys' translation, SBE, 11. 118,Google Scholar is highly appreciative; but it is metrical and expanded. He has rendered santi by Nirvāṇa; whence we may conclude that it is so explained in the commentary, which I have not been able to look up: at any rate, that is obviously its meaning. The meaning to be given here to assāsa-passāso, lit. ‘inhalation and expiration’, is indicated by his rendering, “gasping struggle”.

^{page 1007 note 3} Epi. Ind., 2. 456, and plate.Google Scholar

^{page 1007 note 4} It seems hardly necessary to treat this word as an irregular spelling of niyyāsu. The ñ implies an original ny: Childers gives nīyā, as well as niyyā, as = niyā, ‘to go forth’: and from nīyā we might surely have nyayāsu, which would become ñayāsu.

^{page 1007 note 5} Read ētārisāni.

^{page 1008 note 1} Without any inclusion of duplicates, the Aśōka records give us fifteen other such passages, in which the participle sat might have been used but was not used.

^{page 1008 note 2} Loc. cit. The Mansehra text is supposed to have saṁtaṁ: but we might, I expect, easily read saṁta or saṁto instead of that. The Kālsī text is presented as having saṁtaṁ nikhamiṭhā (or °thā) saṁbōdhi: but it seems possible that it may have sate, for saṁte (see loc. cit., 457, note 8); and no special difficulty is raised by the point that the word after that presents one syllable too many for the metre.Google Scholar

The Jaugada text does not help in this matter; the words after lājā dasa being hopelessly illegible. The Dhauli text apparently omits its equivalent (saṁte or sate) for saṁto, and gives:—sē Dēvānaṁpiye Piyadasi lājā dasa-vas-ābhisite nikhami saṁbōdhi[ṁ y]ēn = ētā, etc.: see Archœol. Surv. South. India, 1. 119,Google Scholar and plate 65. If the word really is absent from that text, the omission would not be of much avail against the point that the word stands in four other independent texts. But the Aśōka edicts present many cases in which the writers at first omitted syllables, and then inserted them on revision, sometimes in full-size, sometimes in miniature. The immediate repetition in °site sa(ṁ)te might easily lead to an omission here, which might quite possibly be overlooked even on a revision. And the essential word kaṭe or sāvāpite was omitted in a certain clause in the Sahasrām text of the Last Edict (see p. 1003 above, and note 3), and apparently was not inserted on revision. Before basing any argument on an apparent omission here in the Dhauli text of edict 8, we should have to examine the surroundings closely, and ascertain whether the word was, or was not, supplied, either in miniature above its place, or in the margin or some other position.

Can this be the explanation of the mysterious detached word, read as sēto, but equally capable of being read as sate or saṁte, which stands at a short space after the end of Dhauli edict 6?: see op. cit., plate 64. This word is in different characters from those of the Dhauli texts; and, in fact, in apparently much later characters, of “an alphabet closely resembling that of the Guptas” (ibid., 115). And it has been supposed (ibid., 124, note) that it represents svēto = śvētaḥ, “the white one”, with reference to the dream of Māyā, in which she saw a white elephant descend from the heavens and enter her side; and that it was added in connexion with the elephant sculptured in relief over the panel containing the Dhauli edicts 1 to 6: see Inscriptions of Asoka, plate 29. But Professor Bühler observed (loc. cit., 115) that the Dhauli dialect would require svēte (not sēto) as = śvētaḥ, and that “it is difficult to imagine how, in later times, anybody could have an interest in making such an addition.” In this detail, and in some other peculiarities in the Dhauli texts, he was inclined (ibid.) to find support for a view that “the alphabet of the Aśoka inscriptions was not the only one known and used in his times.”

^{page 1009 note 1} This Journal, 1908. 490.Google Scholar

^{page 1009 note 2} See loc. cit. (preceding note), 487.

^{page 1009 note 3} See Dīpavaṁsa, 7. 22; where chha-vassamhi means “at the sixth year”; i.e., in the sixth year completed.

^{page 1010 note 1} I notice in one passage, as an exception, antaradhāyissati, ‘will vanish away, disappear from sight’: text, loc. cit. (1876), 240, 241Google Scholar; trans., SBE, 11. 89. Perhaps a few others might be found: the Sutta is an artistic work.Google Scholar

^{page 1010 note 2} e.g., text, loc. cit., 236–8, 241, 245.Google Scholar

^{page 1010 note 3} Text, , loc. cit., 250Google Scholar. I quote verbatim the translation by ProfessorDavids, Rhys, SBE, 11. 112.Google Scholar

^{page 1010 note 4} Atītasatthukaṁ pāvachanaṁ; lit. “the word has a teacher who is dead.”

^{page 1011 note 1} Mam = achchayēna; text, line 18. So again in lines 20, 24–5, 26.

^{page 1011 note 2} Bhagavato pana mayaṁ bhante achchayēna, etc.: text, 241, line 8; translation, loc. cit., 90.Google Scholar

^{page 1012 note 1} See, e.g., Vinayapiṭaka, , ed. Oldenberg, Mahāvagga, 1. 7. 10Google Scholar, which passage narrates the admission of the first Tēvāchika - Upāsaka: and compare the definition of an Upāsaka, given by Buddha to Mahānāma the Sakka, as presented in the Aṅnguttara-Nikāya, , part 4, p. 220Google Scholar. Originally there were only two refuges, Buddha and the Dhamma: see Vinayapiṭaka, Mahāvagga, 1. 4. 5,Google Scholar which passage narrates the admission of the first Dvēvāchika-Upāsakas. The two passages in the Mahāvagga indicate that, while Buddha was alive, it was customary to add the words: “let the Bhagavat receive me as an Upāsaka who has taken refuge from this day forth as long as life endures.”

^{page 1012 note 2} See, e.g., ibid., 5. 13. 1, and Suttanipāta, verse 376/375.

^{page 1013 note 1} From the facsimiles in Ind. Ant., 6. 156; 22, 299. Square brackets are used to mark syllables or parts of them which in the original are illegible, seriously damaged, or imperfectly formed. Ordinary brackets mark doubtful readings.Google Scholar

^{page 1013 note 2} As observed by Professor Hultzsch, this is an abbreviation of vasāni.

^{page 1013 note 3} We must, perhaps, correct this into hakā, in accordance with Professor Bühler, who held that “the stroke, intended for the curve of the first consonant, has been attached by mistake to the top.”

^{page 1013 note 4} The correct word appears to be ārādhave or ālādhave: see, for instance, rock-edict 6; svagaṁ ārādhayaṁtu, Girnār, line 12; svagaṁ ālādhayaṁtu, Jaugada, line 6. If we compare the lā in silāṭhabhe at the beginning of line 5, we may almost take it that the writer intended to give lā, not rō, here.

^{page 1013 note 5} Professor Bühler read pakar. va: M. Senart read pakarā va. I follow Professor Hultzsch's proposal to restore pakame va, which, in spite of the apparent ra or rū, seems to be required by the Sahasrām and Siddāpura texts.

^{page 1013 note 6} This cannot be a 2nd person plural, even of the imperative; the termination would be tha, not ta: moreover, aṭhe is a nominative, not accusative. The word may possibly stand for lēkhāpētaṁ, as the rare 3rd person sing, of the imperative, ātmanēpada. But it is more probable that the original has °tu, 3rd person sing. of the imperative, parasmaipada, used as the passive.

^{page 1014 note 1} Regarding these marks of punctuation, see p. 1004 above.

^{page 1014 note 2} Words supplied to complete the sense in translating are given in italics, in brackets.

^{page 1014 note 3} Lit. “by a lowly person somewhat or somehow applying himself.”

^{page 1014 note 4} The reference seems plainly to be to rock-edict 10, on the subject of application with a view to welfare in the other world: in the Girnār text, line 4 (EI, 2. 459, and plate), we have:—Dukaraṁ tu khō ētaṁ chhudakēna va janēna usaṭēna va añata agēna parākamēna savaṁ parichajitpā; “but that, indeed, is difficult to be done by either a lowly person or an exalted one, otherwise than by extreme application, laying aside every (other aim).”Google Scholar

^{page 1015 note 1} It does not seem necessary to take pavatisu as equivalent to parvatēshu, ‘on hills or mountains’. In the old literature we have parvati, parvatī, in the sense of ‘rock, stone’.

^{page 1015 note 2} I take the cha at the beginning of this clause as a mere expletive. With this clause in answer to the preceding one, compare what we have at the end of the seventh pillar-edict: “wherever there are stone pillars or stone surfaces (for writing on), there this writing about dhamma should be placed (lit., made); by which means this (matter) may endure for a long time.”

^{page 1015 note 3} The translation of the Sahasrām version of this postscript will be:—“And this address [was composed or delivered] by him (Dēvānaṁpiya) who has passed away two (hundred and) fifty-six (years) since the Tranquil One passed away: (in figures) 256.”

The translation of the Brahmagiri version will be:—“And this address was delivered by him (Dēvānaṁpiya) who has passed away (in the year) 256.”

^{page 1015 note 4} Dying speeches are not altogether unknown. We may note the dying speech of Duṭṭhagāmaṇī, king of Ceylon, presented in instalments in the Mahāvaṁsa, , 32. 16–62.Google Scholar