________________
264
EPIGRAPHIA INDICA.
(VOL. XVIII.
The first word was read san ... by Cunningham, and sa hvatsarasa by Dowson. E. Thomas demurred to Dowson's reading, but did not suggest any alternative. Messrs. Senart and Boyer read sambadhae and explained this word as meaning "du comput continu ", " in the continuous reckoning ". There cannot, however, be any doubt that Dr. Thomas was right in reading sarbatéaraē. The traces of the letters visible in the impressions and the plates make this reading absolutely certain. Moreover, the use of té for Sanskrit ts in this very word seems to be quite regular in Kharðshthi inscriptions. Thus we find sanvatsaraye in the Taxila copper-plate of the year 78, in the Mount Banj inscription of the year 102 and in the Paja record of the year 111. Similarly sanvatsarë is the usual form in the Kharðshthi documents from Eastern Turkistan.3 In the Suë Vibär plate we have samvatsarē, but the upper part of the compound is misshaped, wherefore Professor Franke roads sa nuachare and Mr. Majumdar sa nuachchhare. The Hidda epigraph of the year 28 and the Ara inscription of the year 41, finally, have sambatfaraē.
The writing té is no doubt meant to render the current pronunciation of the compound ts. The usual Prakrit representative of ancient ts is chohh, which only differs from tf in being aspirated. It is probable that the form tá is due to the linguistic tendency of the north-western dialect which the Indo-Skythians adopted.
The same compound occurs in the Kharðshthi manuscript, of the Dhammapada, where M. Senart reads it vis, in ahitsai, A* 8; bhamētsu, B 34; bhätsidi Cvo 3 and matsana Cxviii vo 2. The two last words correspond to Sanskrit bhētsyati and matsyanām, respectively, where ts has a similar origin as in samvatsara. In ahitsai and bhamatsu & t seems to have been inserted between a nasal and with a consequent change of sto s. There is nowhere any trace of the aspiration which comes in in the common Prakrit form.
There are indications which seem to show that the absence of aspiration in similar compounds is an old feature of north-western vernaculars. For we find pacha for pascha in the Shahbāzgarhi version of the Aboka inscriptions. On the other hand, pātcho, afterwards, in Ancient Khotani might lead us to the conclusion that we are faced with a phonetic tendency in the home-tongue of the Indo-Skythians.
The change of v to b in sambatéarae is perhaps to be explained as a result of the influence of the Indo-Skythian substratum. For not only do we find b for v after an anusvāra in the Kharo.hthi manuscript of the Dhammapada, where api and tvu become vi and va, respectively, after vowels, but bi and ba, respectively, after an anusvāra. But in ancient Khotani b frequently corresponds to Sanskrit , not only in loanwords such as bina, Sanskrit vinā, but also in indigenous words such as bissi, Sanskrit visva. The actual sound was perhaps a bilabial spirant, English w
The word following after sarnbatsuraē is certainly tisatimae. M. Senart read tisatamaē, but the s-stroke of the second ta is too distinct to be acoidental. The form fatimua also occurs in the Pāja inscription, where we read ēkada[sa"]éatimayē, and, as remarked by M. Boyer, in the Skärah Dherl image inscription, where I follow Dr. Fleet in rending ēkupachadusatimae.
Then follow the numerical symbols 1, 100, and 1 1 1, and Vesakhasa masasa divase, as has always been recognized.
L. 3 ff. were not read by Dowson, and Cunningham only attempted to make out some few words. M. Senart was the first one to give a transliteration and interpretation of the greater portion of these lines, and M. Boyer gave a continuous reading of the whole. .
1J.R. A. 8., New Series, Vol. IX, 1877, p. 10, footnote. 'J.R. 4. 8., 1913, p. 636*. . Cf. Konow, Acta Orientalia, II, p. 114. • Pali und Sanskrit, pp. 96 f. • Sir Antosh Mookerjee Silver Jubilee Volumes, Vol. IIT, PL. 1, PP. 489 # . Cf. Konow, Festschrift Windisch, p. 91.