^{page 2 note 1} See Davids, Rhys, Introduction to Milinda I, p. xiGoogle Scholar, “Sacred Books of the East,” vol. xxxv; compare the same, pp. xxii, xxiii, “Milinda must have reigned for a considerable time in the latter half of the second century b.c., probably from about 140 to about 115, or even 110 b.c.”

^{page 2 note 2} The Chinese may represent a different original, aa MM. S. Lévi and E. Specht think, or a portion of a text of Pāli recension mixed with a comment or notes made in Siam or somewhere else. These points will become clear when the promised translation of the above two scholars has been laid before us.

^{page 2 note 3} See the Proceedings of the Ninth Oriental Congress, 1892, vol. i, pp. 520–529Google Scholar.

^{page 2 note 4} That a reference is in the commentary has been pointed out by Burnouf, in his “Introduction,” etc., p. 570Google Scholar.

^{page 2 note 5} See Milinda, part i, p. xxvi, and part ii, p. xvii.

^{page 2 note 6} Note to his translation of Tāranātha, p. 298.

^{page 3 note 1} Bunyu Nanjio's Catalogue of the Chinese Translation of Buddhist Books, No. 1358.

^{page 3 note 2} Milinda, part ii, p. xii (S.B.E. xxxvi).

^{page 4 note 1} Very likely they are the same: see below.

^{page 4 note 2} See Davids, Rhys, Introduction, Milinda II, p. xiGoogle Scholar. He thinks that, as there is nothing about this curious introduction in either of M. Specht's papers, it seems possible that there are really three Chinese books on the “Milinda.”

^{page 4 note 3} The importance of this identification has been emphasized by ProfessorDavids, Rhys, Milinda II, p. xii, noteGoogle Scholar.

^{page 5 note 1} Three vols. in all—

Vol. i consists of 23 leaves, the first word being the last .

The division of the collection is marked at the edge, . Though the India Office text is an old Japanese edition (1681), yet it is practically a reproduction of the Chinese edition of the Ming dynasty (1600).

^{page 5 note 2} This is the editor's remark.

^{page 5 note 3} Mark that this book does not begin with “evaṃ mayā ṣrutaṃ,” though it pretends to be a sūtra.

^{page 5 note 4} Anāthapiṇḍika in Pāli, another name of Sudatta.

^{page 5 note 5} Kolaka, “black pepper”? Or do the three characters stand for a name something like Karañja, pongamia glabra.

^{page 6 note 1} These may be “advanced in age,” “middle-aged,” and “young.”

^{page 7 note 1} Caṅkrama here, of course, cannot be a “covered walk.” I-tsing, a Chinese traveller in India (a.d. 671–695), mentions “Caṅkrama” of the Nālanda monastery in his “Record of Buddhist Practices in India and the Islands of the Southern Sea,” a translation of which will soon be published by the Clarendon Press in the Anecdota Oxoniensia series. For “Caṅkrama,” see also Mahāvagga v, 1, 14, note 2; “Sacred Books of the East,” vol. xvii.

^{page 7 note 2} lit. “Actionless Nirvāṇa Path.”

^{page 7 note 3} Kālavana?

^{page 8 note 1} This is very curious; may be for “lodged in mid-air,” as an adjective to the “human life.”

^{page 8 note 2} “Lan” is never used for “lin” in transcribing a Sanskṛit word, but used almost exclusively for “lan” or “ran.” This shows that the original had Mi-lan or Me-lan as its first part, not Mi-lin or Me-lin. St. Julien's “Méthode pour Transcrire les Noms Sanscrits en Chinoise” also does not give any instance of “lan” being used for any Skt. syllable, but “laṃ,” “lan,” or “ran” (including “la” or “ra” ): see p. 135.

^{page 9 note 1} For Rohaṇa, see Davids, Rhys, Milinda I, p. 13seq.Google Scholar

^{page 9 note 2} The four grades are—(1) Sotāpanna, (2) Sakatāgāmin, (3) Anāgāmin, (4) Arhat. See Childers, s.v. nibbānaṃ, p. 268, and Müller's, Max “Diamond Cutter,” IX (S.B.E. vol. xlix, part ii), p. 120, note 2Google Scholar.

^{page 10 note 1} See Milinda, part i, p. 30.

^{page 10 note 2} Ibid., p. 31.

^{page 10 note 3} Heretic doctrines are said to be ninety-six or ninety-five. In the Brahma-jāla sūtra sixty-two views are enumerated, as can be seen in Davids, Rhys' Milinda II, pp. xxiii–xxvGoogle Scholar. Of its two Chinese translations, the one made in a.d. 222–280 (Nanjio's Catal. 554) is called “Brahma-jāla sūtra on the 62 views”; but the other, of a.d. 406 (Nanjio's Catal. 1087), is called simply Brahma-jāla sūtra. The latter belongs to the Mahāyāna and the former to the Hīnayāna.

^{page 11 note 1} See Davids, Rhys' Introduction, Milinda II, p. xiiGoogle Scholar.

^{page 13 note 1} Bodleian Library, Japanese, 65^DD.

^{page 13 note 2} Nanjio's Catal., p. xxvii (1). All the contents of this Catalogue have been given by Nanjio in his Catalogue, pp. xiii–xvii.

^{page 13 note 3} Nanjio's Catal., p. xxvii (8).

^{page 13 note 4} There seems to have been two translations, and the compiler of the catalogue may have taken them to be one and the same translation (the difference not being very great).

^{page 13 note 5} Not in Nanjio's Catalogue. Bodl. 65^DD, vol. xiii, p. 47^b.

^{page 14 note 1} Nanjio's Catal., p. xxvii (11).

^{page 14 note 2} Ibid. (12).

^{page 14 note 3} , Jap. p. 91^a (Tâ-tsân mn-lu), and , Jap. p. 104^a (Tâ-pu-ning-shih Tâ-tsân mu-lu).

^{page 14 note 4} All these are included in above 7 in my table.

^{page 15 note 1} See Milinda, part ii, p. xii.

^{page 15 note 2} See the same as above.

^{page 15 note 1} For the thirteenth edition of Buddhist Books, see Nanjio's Catalogue, p. xxviii.

^{page 16 note 1} .

^{page 16 note 2} .

^{page 16 note 3} Pāli Miscellany, part i, p. 55.

^{page 17 note 1} Milinda, part i, pp. xviii, xix.

^{page 18 note 1} Chan-mi-li-wang-chun by name, according to the Chinese text of Milinda.

^{page 18 note 2} Āyupāla by name, according to the Pāli text, and Ya-ho-la (Āyupāla) in the Chinese text.

^{page 18 note 3} Compare Davids, Rhys, Milinda, part i, p. 32Google Scholar (S.B.E. vol. xxxv).

^{page 19 note 1} Perhaps Nāgasena did not like to submit himself to the king.

^{page 21 note 1} Compare Rhys Davids, part ii, v, 7, 24 (S.B.E. xxxvi, p. 112), where there is no such answer as this given.

page 21 note * This is a translation of the Introductory part of Milinda.