________________
10
EPIGRAPHIA INDICA.
(VOL. XI.
confasion that there was added the projecting tail to the right, which produced the form illustrated in C., D., and E. above.
Professor Bühler's latest instances of the earlier type are as follows. As noted in the preceding paragraph, his plate VIII, 3, II, is an i of A.D. 804, and is really a transitional form. His plate VIII, 3, III, is from the record on the Kadaba plates from Mysore, bearing a date in A.D. 812 or 813 : the vowel is found fourteen times in this record : the instance figured seems to be from iva, line 15, No. 23 (side ii a, line 1), Ind. Ant., vol. 12, p. 14, plate ; Epi. Carn., vol. 12 (Tumkur), Gb. 61, plate. His remaining two illustrations are from the eastern parts of Southern India. The later of them, plate VIII, 3, V, is entered as if it was taken from the copperplate record which gives the date of the coronation of Amma II in A.D. 945, Ind. Ant., vol. 7, p. 15, plates: but that record does not include any initial i ; and the illustration seems to have been supplied from ih-aijatē, line 40, No. 7 (side iii b, last line), in the 'Påganavaram' plates or Diggubapra grant of the period A.D. 934-45, Ind. Ant., vol. 13, p. 214, plate. On the other side, his earliest instance of the later type is plate VIII, 3, VI: it is from the eastern part of Southern India, from the copperplate record which gives the date of the anointment of Rājarāja I in A.D. 1022, and may be of any date from that year up to about A.D. 1083; and it appears to be the i of iti, line 11, No. 5, Ind. Ant., vol. 14, p. 50, plate. There is thus a gap in the history of the Kanarese and Telugu initial short i, of roughly a century, from A.D. 934-45 to 1022-63, to be extended, in fact, as regards the Kanarese country, to even two centuries, in respect of which we have, so far, no information. We have now to see how this gap can be bridged over. It was chiefly due, as far as Professor Bühler was concerned, to a lack of materials in the shape of published facsimiles. The same want still exists to almost the same extent. Bat we can now cite various facsimiles which have been published since his time and I can supplement them by ink-impressions which necessarily were not available to him.
We will consider first such materiala as are available from the Kanarese districts of the Bombay Presidency; taking the matter up from the earliest instance, known to me, after Professor Bühler's latest instance of the earlier type. Here I use only records which are specifically dated ; and, when I cannot refer to a published plate, I cite my details from inkimpressions which were prepared under my direction when I was in the districts in question.
From these parts, we still have the earlier type of the initial short i running through the records of the time of the Rashtrakūta king Amoghavarsha I. We have it in the Nilgund inscription, dated in A.D. 866, from the Gadag taluka, Dharwār; ante, vol. 6, p. 102, plate; ir iti, line 6, No. 17: it occurs also in idan= towards the end of line 28, to which part of the record, however, the plate does not extend. We have it again in the Sirur inscription of the same date, from the Nawalgund täluka, Dharwăr; ante, vol. 7, p. 206, plate; in iti, line 4, No. 6, and ilnürvuorum, line 20, No. 13: the illustration A. on p. 7 above is from the latter instance. Again in an unpublished inscription, dated in A.D. 872 or 874, at Chinchli in the Gadag täluks; in indapayyam and intui, line 4. Again in an unpublished inscription, dated in A.D. 874, at Ron, the head-quarters of the Ron taluka in the same district; in idans, line 10: also in the same word in line 4 of a second record, not dated, below that one. And again in the Nidagundi inscription of A.D. 874-75, or within a year on either side, from the Bankäpar tāluka, Dharwär; ante, vol. 7, p. 213, plate; in idam, line 16, No. 13, in idan-, line 17, No. 12, and in i (for 1) kallar, line 19, No. 8: also in i (for) fānamar in line 25 in the supplementary record at the cop of the stone, not shown in the plate. And for the next reign, that of Kộishọa II, we
I have abstained from quoting a few records which are of doubtful authenticity: and it has of course been useless to quote records which refer themselves to (for instance) the reign of Kanparadeva or Kandaravallabha, withont anything to indicate which king Kpiebna is intended. But I have not found in any of them anything that conficta in any way whatsoever with the facts and resolta which I bring forward.