________________
130
EPIGRAPHIA INDICA.
[VOL. XII.
Dl commemorates some near relations of Säriyavikrams, whether his consort be among them or not. The string of words (with honorifios) after his name, for which there is no parallel in the other records, seems to indicate that much. Besides Suriyavikrama's own urn is C, and no man requires more than one coffin. It is tempting to interpret pho as meaning "grandchild", on the strength of 1. 24 of the Myazedi inscription, but this last has pli, not pho, which is a doubtful reading anyhow.
I pass on to the next phrase. In the Myazedi inscription sni8 means "year", and I am confident that it has the same sense in our orn-inscriptions. A priori it is reasonable to assume that it would be closely associated with numerals, and I note that that is the case here. In BI it is followed by hrd, which in the Myazedi inscription represented "eight". In Dl it is followed by hau8, which we might perhaps read ho8 and which in any case reminds one of the word ho8 that stood for "three" in the Myazedi record. Later on in A there is a word tå, which in the Myazedi record meant "one." The inference is that the corresponding unknown words are also numerals. Among them there is one which constantly appears in the samo relation to the other words, thongh these change. The constant is sū, and assuming "Pya" to use a decimal system, we must conclude that sū means "ten". For reasons that will presently appear, it cannot be " a hundred " nor is it likely to be "twenty."
At this point a digression becomes necessary. M. George Cedès has published a very kind appreciation of my paper on the "Pyu" text of the Myazedi inscription and drawn my attention to the fact that the symbols in II. 1-2 thereof which I had read cũ jha 6 are not " Pyu” words of number as I had supposed but the conventional symbols employed in some ancient Indian inscriptions to represent 1000, 600, and 20, respectively. I accept these identifications the more readily as I had myself felt (and suggested in a note that my o might after all possibly be the old numeral symbol for 20. I can now confirm M. Codes' view, as I have compared the original rubbings (which are much larger and also clearer than the plate published with my paper in the Journal) with Bühler's Indische Palæographie (Pl. IX) and find that the symbols, including that for 600, correspond. But with regard to krå I am not 80 sure. M. Codes would also make of it & conventional symbol. But the symbol is hra not hrd. And what has a tonal mark to do with a numeral figure ? Sooondly, hrá is used in l. 7 of the Myazedi text in connexion with an entirely different form of 20, which I conjecturally transliterated shu but now propose to identify with the tpu (or npü ?) of A and Bl. This I take to be a genuine "Pyu" word for "twenty", not an Indian numerical symbol. Thirdly, hrá is apparently used in Bl as a multiplier of si, ten. Therefore I still think that I may have been right in taking hrá to be a "Pyu" word and a relative of the Burmese Thach, of gennine Tibeto-Burman descent,
To return to the other numerals in our four inscriptions. There is no internal evidence as to the values of the unidentified ones not yet mentioned. The following table is therefore to be considered as based largely on conjecture tempered by a general comparison with the forms of numerals in other Tibeto-Burman languages :I tå
5 na, pîna 9 tko 2 hni 6 tra 10 60, (sau)
hans, (hos) 7 kni 4 p!å 8 brå
tpă
1 Bulletin de l'Enole Française d'Extrême-Orient, 1911, pp. 435 f. ? Journal of the Royal Asiatic Society, 1911, p. 383.
Compare as a nandy reference Journal of the Royal Asiatic Society, 1918, PP. 916 #.
