^{page 366 note 1} Sometimes written on the right of the character, when by reason of the presence of vowel symbols there is no room for it on the top. A similar reason appears to account for (but there is apparently one case of ).

^{page 367 note 1} The forms in the first column are transliterated to the best of my ability, having regard to the shapes and probable values of the letters. Those in the second column are taken from one of the other versions, usually the Burmese, as spelt therein and transliterated (except that I here write w instead of v) in the ordinary way, without any reference to the modern peculiarities of Burmese or Talaing pronunciation. I leave the visarga and anusvāra symbols (and their variants and combinations) in their original shapes for the present.

^{page 367 note 2} The line-references throughout are to A unless otherwise stated.

^{page 367 note 3} The engraver forgot the Sū and put it in afterwards below the line, adding a small cross to mark its proper place.

^{page 367 note 4} A mistake for Vrahmad yo: v. infra.

^{page 368 note 1} The o has a hook on the top turning to the right, the force of which I do not know: it looks like o + au, an improbable combination; perhaps it is meant to denote ō (or å?).

^{page 368 note 2} I distinguish it conventionally by a dot underneath, as I have previously done in the case of the Talaing equivalent.

^{page 369 note 1} These and some of the other groups in parentheses are of course alternatives to some of the double letters similarly grouped in the first part of the Table. The object in displaying them thus is to draw attention to the peculiar phonetic character of the language pointed out infra.

^{page 370 note 1} I am not quite sure that the letter I have rendered d may not (in some cases, at any rate) be ṭ.

^{page 370 note 2} Or, alternatively, that the language of our text derived its alphabet not direct from India, but through a language in which this phenomenon had occurred. In Talaing similar changes have taken place; but whether they can be dated as far back as some centuries before the date of this inscription is a question that needs further inquiry. In any case the form of this alphabet is much more archaic than the contemporary Talaing.

^{page 371 note 1} Aspirated consonants do not count as two but as one. This may serve to explain the apparent exception hm in Vrahma-; or it may have sounded Vrahama- (of. the variant spelling mhaṭhe for mahaṭhe). It is noticeable that even ṁ and ṅ are unrepresented in the language of this text. This goes to show that its anusvara does not stand for either of these.

^{page 371 note 2} And even sr in Sri; but this may be merely a conventional way of writing. There are some subscript letters in the text where a vowel must be supplied.

^{page 372 note 1} B rightly has Ḅūdha here.

^{page 372 note 2} Or initial i?

^{page 372 note 3} Perhaps, however, the hook is merely a prolongation of the left part of the m, and the vowel symbol stands for i.

^{page 372 note 4} Hereafter, when giving a word in the Roman character and calling it simply Burmese I imply that it occurs in the Burmese text of the Myazedi record.

^{page 374 note 1} The ch is a new letter, conjecturally identified by its characteristic shape, which is not unlike the older Indian forms and the contemporary form in the parallel Burmese and Pāli versions.

^{page 375 note 1} Probably primarily a noun: cf. tra ḅå sagha, infra.

^{page 376 note 1} Very possibly “slave” is here used as in Burmese to replace a personal pronoun of the 1st person, so that the combination tra ḅå would mean “my lord”. It must have some such meaning in any case, however arrived at. Note that the syntax is analogous to that of possessive pronouns (which precede) but contrary to that of descriptive words (which follow the principal, noun).

^{page 376 note 2} The gh is a new letter conjecturally identified by its shape.

^{page 377 note 1} It is possible that the subscript anusvāra in this and other such words is the tonal mark of the first, indeterminate, syllable. But as I have at present no means of deciding the point I transcribe them just as I find them.

^{page 377 note 2} Unfortunately it appears that different expressions for “twenty” are used in these two places. Consequently I have not been able to identify either of them with certainty, and cannot be quite sure of the equivalents for 1,000, 6, and 100 which should occur in 1.1.

^{page 378 note 1} I cannot explain kya , which may be a verbal affix helping out the sense of se. The letter k is a new one, conjecturally identified as such by its resemblance with old forms in various Indian alphabets.

^{page 379 note 1} This tra may represent Sanskrit dravya, “goods.”

^{page 379 note 2} On the other hand there is the curious position of the word ti°(apparently = “in”) between the demonstrative yå and the noun pri . Cf. its use in 11. 23, 24. 26.

^{page 381 note 1} The recent death of this explorer (by drowning in the rapids of the Me Khong) is a great loss to Indo-Chinese exploration and research.

^{page 381 note 2} The latter is illustrated in JRAS., Jan., 1911, Plate VIII. 1 (facing p. 150).

^{page 381 note 3} See also his Report of the Superintendent, Archæological Survey, Burma, 1910, pars. 44–5. He there quotes some remarks made by the late Professor Biihler on our text. But that eminent epigraphist only glanced at the inscription and did not attempt to study it, as he at once recognized that the language was not Sanskritic. His remarks are therefore not very helpful. As to the votive tablet, see ibid., par. 38.

^{page 383 note 1} Many letters are still doubtful, and there are also in several places marks above or between the lines to which I cannot at present attach any definite meaning.

^{page 383 note 2} This and the three or four preceding words are somewhat of a puzzle. They must represent “1620”. I am not at all sure that what I have transcribed e is not the old numeral figure for “20”. The readings cū and jha are also very doubtful. Perhaps the latter is really hña or jhña. The former might possibly be thū.

^{page 383 note 3} B omits ‖‖.

^{page 383 note 4} I am by no means sure of the final vowel.

^{page 383 note 5} B reads Triḻogavadasagadevi.

^{page 383 note 6} B inserts ‖‖ ḅå.

^{page 383 note 7} B reads maya .

^{page 384 note 1} B wrongly reads ḅi.

^{page 384 note 2} Very doubtful reading, perhaps dū.

^{page 384 note 3} B has ‖‖ o ‖‖ for ‖‖.

^{page 384 note 4} A appears to read to, but the mark under the line may be accidental. B has to, only.

^{page 384 note 5} Perhaps to be read tkha.

^{page 384 note 6} B reads Būdha.

^{page 384 note 7} B omits ḅå.

^{page 384 note 8} B omits o.

^{page 384 note 9} B perhaps reads ri°

^{page 384 note 10} A appears to read pha, a doubtful letter; B has pa.

^{page 384 note 11} B omits this pra.

^{page 384 note 12} A appears to have chohere. I have followed the B reading.

^{page 384 note 13} B reads mhaṭhe.

^{page 384 note 14} B reads Mūgalubūdi°saṭhe.

^{page 384 note 15} B reads Saumedhaḅadi°.

^{page 384 note 16} B reads rightly Vrahmadyo. (The -e- is doubtful in both copies.)

^{page 385 note 1} B has ‖‖. There is a blur in A.

^{page 385 note 2} B reads staḅana.

^{page 385 note 3} For Būdha: B reads Ḅūdha.

^{page 385 note 4} B rightly reads ḅi . A has ḅi only, no trace of any .

^{page 385 note 5} B reads ma ro instead of tog (which is not quite distinct in A: it might be ro).

^{page 385 note 6} B reads Samanaḽō .

^{page 385 note 7} B has ‖‖ instead of ‖.

^{page 385 note 8} B reads Ji°vū(or, possibly, Jivu).

^{page 385 note 9} B has ‖ instead of ‖‖.

^{page 385 note 10} B reads u sa.

^{page 385 note 11} B inserts ḅå.

^{page 385 note 12} B has ‖‖.

^{page 385 note 13} B reads yå.

^{page 385 note 14} Or perhaps pi or pau, though I hardly think so. It has one of the unexplained marks over it.

^{page 385 note 15} Conceivably these two words may be read kůdū.

^{page 385 note 16} B inserts ‖.

^{page 385 note 17} B inserts ‖‖ or ‖.

^{page 385 note 18} B reads Ḅūdha.

^{page 386 note 1} Very doubtful transcription; perhaps hñi should be read.

^{page 386 note 2} B has some more punctuation marks to indicate the end of the text.

^{page 386 note 3} If ḻa is the right transcription here, then ḍain 1. 9 is probably to be read ḻa.