^{page 646 note 1} I have altered the transliteration in accordance with that used in this Journal, and have given capitals in the case of the words taken by M. Senart as proper names.

^{page 649 note 1} There are some minor points where I differ from M. Senart's reading, but they do not affect the sense.

^{page 650 note 1} See, e.g., Jāt. I, 278:Google Scholar bodhisatto nāgabalo thāmasampanno nadiyā orimatīrato uppatitvā—dīpakassa orato nadīmajjhe eko piṭṭhipāsāṇo atthi — tasmim nipatati.

^{page 650 note 2} Ep. Ind., vol. 4, p. 55.Google Scholar

^{page 650 note 3} Mr. Thomas is inclined to look upon these phrases as derived from Persian models; see Ep. Ind., vol. 9, p. 139.Google Scholar

^{page 651 note 1} Ep. Ind., vol. 9, p. 140.Google Scholar

^{page 651 note 2} In hia translation of the word Mr. Thomas is guided by etymological reasons. He traces hora to the Iranian ahura. But even if this etymology should be correct, it is hardly necessary to say that it is always unsafe to assign a certain meaning to a word on etymological grounds alone.

^{page 651 note 3} Ind. Ant., vol. 37 (1908), p. 64 and plate.Google Scholar

^{page 652 note 1} See Pischel, , Grammatik der Prahrit-Sprachen, § 401.Google Scholar

^{page 652 note 2} On the other hand, also, no counter-instance is known to me. The dialect of the Shāhbāzgarhī and Mansehra inscriptions, where tm is represented by t (i.e. tt) and tm respectively (see Edict XII), of course, cannot prove anything in this respect, as the Aśoka edicts are more than two hundred years older than the present inscription.

^{page 652 note 3} Ep. Ind., vol. 9, p. 147.Google Scholar

^{page 653 note 1} See Ep. Ind., vol. 9, p. 146, pl. 4,Google Scholar Table of Aksharas. I may mention that also Professor Hoernle transcribed the character as sta.

^{page 654 note 1} Pāli mid Sanskrit, p. 112.Google Scholar

^{page 654 note 2} Gött. Nachr., 1906, p. 145;Google ScholarEp. Ind., vol. 9, p. 192.Google Scholar

^{page 655 note 1} Ind. Ant., vol. 19, p. 59.Google Scholar

^{page 655 note 2} Journ. Bo. Br. Roy. As. Soc., vol. 22, p. 152 f.Google Scholar

^{page 655 note 3} Franke, , loc. cit., p. 111.Google Scholar

^{page 655 note 4} Journ. As., sér. 9, vol. 12, p. 206.Google Scholar

^{page 656 note 1} M. Senart's reading ṣeho is a mistake.

^{page 657 note 1} Thus Sk. iva is generally represented by va, but by ba in A¹ 6; A² 4; B 28; C^vo 14. Medial p frequently becomes v, and accordingly the enclitical api appears as vi in C^vo 2; 37; but in A³ 10; C^ro 7; 9; C^vo 21; 32; 33, we find bi. The combination rv has become v in nivana B 35, nivinati A³ 1–3, but b in babalca C^vo 31. Original b is replaced by v in avalaśa A³ 15, and the form supraudhu A⁴ 4–9 goes back to *supravudhu = Sk. suprabuddham.

^{page 659 note 1} See my remarks, Arch. Surv. Ind. Annual Report, 1903–1904, p. 290.Google Scholar

^{page 660 note 1} The exact meaning of kuśalamūla in this phrase appears from several Buddhist inscriptions at Mathurā, where, instead of etena Icuāalamūlena we find anena (or imena) deyadharmaparityāgena, “through this liberality in religious gifts;” see Ind. Ant., vol. 33 (1904), p. 154 f.Google Scholar

^{page 661 note 1} JRAS., vol. 20 (1863), p. 255 ff.Google Scholar and plate. The passage was read also by Senart, M., Journ. As., sér. 8, vol. 15 (1890), p. 121,Google Scholar but I differ from him in several points.

^{page 661 note 2} These three words are doubtful.

^{page 661 note 3} Mahiya corresponds to Sk. mahyam, used in the sense of a genitive.

^{page 661 note 4} Similar phrases are found in the rest of the inscription, but the context is partly obscure.

^{page 661 note 5} Cunningham, , Mahūbodhi, pl. xxv.Google Scholar

^{page 661 note 6} Ind. Ant., vol. 33 (1904), p. 155.Google Scholar

^{page 661 note 7} Ep. Ind., vol. 1, p. 390,Google Scholar No. 18; cf. Senart, , loc. cit., p. 9.Google Scholar

^{page 661 note 8} Ep. Ind., vol. 1, p. 240.Google Scholar The words agrebhāvapratyaṃśatāyāstu are a parenthetical phrase. Bühler separated the words °pratyaṃᑛatāyās tu. I prefer to take them as °pratyaṃśatāya astu, °pratyaṃśatāya being the Prakrit form for either °pratyaṃśatāya or °pratyaṃśatāyai. That agrebhāva corresponds to agrabhāya in the Wardak inscription, has been pointed out already by Senart, M., loc. cit., p. 10.Google Scholar

^{page 662 note 1} The MSS., however, have sāsad, and the metre of the line is wrong.

^{page 663 note 1} Perhaps of the same type as in taeṇa or taena in 1. 7, but I do not venture to decide this from the photolithograph alone.

^{page 664 note 1} Ep. Ind., vol. 4, p. 54 ff.Google Scholar

^{page 664 note 2} For the elision of the t, Bühler compares the elision of k in saṁvatsaraye and aṭhasatatimae, which is not the same. There would be another epigraphical example for the elision of a t if Mr. Banerji were right in reading kae (=Sk. kṛtam) in the Muchai inscription, Ind. Ant., vol. 37 (1908), p. 64.Google Scholar But according to an impression and a photograph before me the true reading is undoubtedly kue, which stands for *kuve= Sk. kūpah; compare the Paja inscription, Ibid., p. 65, where, by the way, we have to read Anaṁdaputrena Saṁgamitrena hue karite, not katite, as Mr. Banerji thinks, karite corresponding to Sk. kāritaḥ. Also, the words before and after kue I do not read as sahayatena and vasisugena, but as sahayarana and vaṣiśugana. However, the meaning of these words would require a fuller treatment than can be given in a note. A third example would be saśpae in the Mathura lion-capital inscription A, ii, if this should really correspond to Sk. śaśvate, but it is hardly necessary to say that the explanation of the word is quite uncertain.

^{page 665 note 1} The question, of course, cannot be decided without inspecting the plate itself.

^{page 665 note 2} Arch. Surv. Rep., vol. 5, pl. 16;Google ScholarJRAS., vol. 20, pl. 10.Google Scholar

^{page 666 note 1} This line is properly to be inserted after 1. 7.

^{page 666 note 2} This line is properly to be inserted after saṁ 10 4 4 in 1. 1.