________________
§ 32, B.]
INDIAN PALEOGRAPHY.
36, III, XVII, and plate VII, 36, II, IV). Finally, the greatly cursive I (plate VIII, 3, XVI) appears to be the result of a peculiar combination of three curves, which replaced the ancient dots. But an I of this kind has hitherto not been traced.
This analysis of the Tamil alphabet of the 7th century makes it probable that it is derived from a northern alphabet of the 4th or 5th century, which in the course of time was strongly influenced by the Grantha, used in the same districts for writing Sanskrit.
75
The next oldest specimen of the Tamil script, which is found in the Kaśākūḍi plate of about A. D. 740 (not represented in plate VIII), shows no essential change except in the adoption of the later Tamil ma.
But the inscriptions of the 10th, 11th and later centuries (plate VIII, cols. XVII-XX) offer a new variety, which is more strongly modified through the influence of the Grantha. The ta, pa and va have now the peculiar Grantha forms. Besides, in the 11th century begins the development of the little strokes, hanging down on the left of the tops of ka, na, ca, ta and na. In the 15th century (plate VIII, cols. XIX, XX) these pendants are fully formed, and ka shows a loop on the left. It is worthy of note that in the later Tamil inscriptions the use of the Virama (Pulli) first becomes rarer and finally ceases, while in the quite modern writing the Virama is again marked by a dot.
B. The Vatteluttu.
Among the Vatteluttu inscriptions, the Sasanas of Bhaskara-Ravivarman in favour of the Jews (pl. VIII, cols. XXI, XXII) and of the Syrians of Kocin, as well as the Tirunelli copperplates of the same king, have been published with facsimiles. Trusting to rather weak arguments, BURNELL ascribes the first-named two documents to the 8th century. But the Grantha letters occurring in the Sasana of the Jews belong to the third and latest variety of that alphabet, and the Nagari sa or si (probably for áriḥ) at the end of the document, to which HULTZSCH has called attention," resembles the northern forms of the 10th and 11th centuries (compare plate V, 39, 47, VIII; 48, X).
From a paleographical point of view, the Vatteluttu may be described as a cursive script, which bears the same relation to the Tamil as the modern northern alphabets of the clerks and merchants to their originals, e. g., the Modi of the Marathas to the Balbodh and the Täkari of the Dogras to the Sarada. With the exception of the I, probably borrowed from the Grantha, all its letters are made with a single stroke from the left to the right, and are mostly inclined towards the left. Several among them, such as the na (plate VIII, 15, XXI) with the curve and hook on the left, the va with the open top and the hook on the left (plate VIII, 38, XXI, XXII; compare-cols. XVII-XX) and the round ra (plate VIII, 45, 46, XXI, XXII; compare 47, XVII-XX), show the characteristics of the second variety of the Tamil of the 11th and later centuries. And with the usage of the later Tamil inscriptions agrees the constant omission of the Virama. Some other characters, such as the round ta (plate VIII, 20-23, XXI, XXII; compare col. XVI), the ma with the curve on the right (plate VIII, 34, XXI, XXII; compare col. XVI), and the ya with the loop on the left (plate VIII, 35, XXI, XXII; compare col. XVI), seem to go back to the forms of the earlier Tamil. And three, the rounded U (plate VIII, 5, XXI), the pointed E (plate VIII, 8, XXI) and the na with a single notch (plate VIII, 26, XXI, XXII), possibly show characteristics dating from a still earlier period.
1 SII. 2, plates 14, 15.
2 Compare the facsimiles, of 10th and 11th centuries, at EI. 3, 284; SIL. 2, plates 2-4; of the 15th century, at SII. 2, plate 5; uncertain, at SIT. 2, plate 8; IA. 6, 142; alphabet, B.ESIP. plates 18, 19.
Compare VENKATYA, EI. 3, 278 ff.
Madras Journ. Lit. Soe, 13, 2, 1; IA. 8, 388; B.ESIP. pl. 82a; EL. 3, 72; alphabet, IA. 1, 229; B.ESIP.
pl. 17.
5 IA. 20, 292,
7 EI. 3, 67.
IA. 1, 229; B.ESIP. 49; disputed by HULTZSCH, IA, 20, 289. • Compare above, § 25, note 8.
