________________
SANCHI STOPA INSCRITIONS.
I. 79, Arapána for Arápáná (ablative), I. 89 Jitamitaye for Jitámitáye, I. 101, Vedisakaya for Vedisakáya, I. 117, Dhamasenaya for Dhamasenaya,
II. 9, and Nagapalitaya for Nágapálitaya, II. 10." (2) The occasional omission of short i, e.g., in Dhamarakhatáye for Dhamarakhitáye,
I. 76, and bhich hunaya for bhichhuniyá, I. 119 (3) The almost invariable omission of medial anusváras, e. g. in Udubaraghara
and its derivatives (see Appendix, List V), in Nadinagara and its derivatives (see ibidem), Tubavana (ibidem), Upidadata (see Appendix, List III), in the proper names beginning with Dhamao (see Appendix, Lists I-V), and in
atenási for anterási I. 12, 66, 88, 112. It seems to me that the omission of the long á is due everywhere to carelessness, and has for its ultimate cause the custom, common even in late times among clerks and tradesmen, of omitting the vowel-marks altogether; for, the real existence of forms like jamata, jataka and jitamita, in the spoken language, has little probability. As regards the omission of i-strokes and of the Anusvåras, it is possible to conceive also of other explanations. The syllables in which the iis omitted stand, one and all, in thesi, i.e. immediately before or after the syllable which bears the stress-accent. A vowel which is placed in such a position is very indistinct, and it is not impossible that na and kha were sometimes actually pronounced instead of ni and khi in words like bhichhúnavá and Dhamarakhatáye. A good many vowel-changes in the literary Prakrits are due to this principle, and a still greater number in the modern vernaculars of India. It must also be noted that in Asoka's inscriptions aya appears occasionally instead of iya," and for the same reason.
With respect to the omission of the Anusvåra it must be borne in mind that the spelling dhamao instead of dhammao is perfectly admissible, because in these inscriptions a single consonant does duty for a double one, and hence dhama may be read dhamma, as the word is invariably spelt in Pali. Further, Upidadata and Tuhavana may stand for Upidadata and Túbavana, because long { and d are marked very rarely in these, as in other old inscriptions. But forms like Upidadatta and Tábavana are again admissible in Pali and other ancient Prakrits, where t and a may be put for in and un as in siha, pisati and so forth. And it must be noted that the form fda for Indra occurs repeatedly, e.g. in L. 1 of the large Nânåghat inscription, Archeological Rep. West Ind., Vol. V, p. 60. A form like Nadinagara may be a negligent spelling for Nádinagara, which actually occurs in our inscription I. 7; for the Pali and the other Prakrits show in a number of cases á for an, e.g., in datha for danshtrá, sandása for samdamba and so forth. It is no matter of surprise that the dialect of the inscrip. tions should show these changes more frequently than the literary language of the Buddhist canon. These explanations of the omission of the Anusvåra seem to me more probable than the assumption that it is simply in every case due to negligence.
Turning to the contents of the inscriptions, the latter offer, in spite of their extreme brevity, a good many points of interest. Among the two hundred and eighty-five in
17 Analogous cases are extremely common in the Bharhut inscriptions, and it will suffice to quote one sentence, No. 155, yan brdhmano anayesi jatakan, which, as Dr. Holtzsch has pointed out, is a quotation from Jatakas, vol. I. p. 293, verne 61 (Fausböll), where the text has yan brahmano avddesi.
* See M. Separt's collection of such forts in the Indian Antiquary, vol. XVII, p. 304; and my additions in der Zeitschrift der Deutschen Morgenländischen Gesellschaft, vol. XLV, p. 164, to which instances chelaya for Sanskrit chai tye, Pali chetiya, bas to be added from the Bharbut inscription No. 11.
x 2
