________________
952
EPIGRAPHIA INDIOA.
[VOL. XXI.
nantic r is denoted by a loop in sarua, I. 5, while the incomplete akshara in sa[r]va, 1.4, is more like the old form, with a curved cross-bar instead of the loop.
The anusvāra is noted in the usual way in santvat saraye, Chandrabhi, 1.1; Nardivadhanena, 1. 3 , Jivanardina, 1. 4, but omitted in Idroe, I. 4. A redundant anusvära is found in Dhranimaisa, 1. 2, while l. 4 has Dhramae.
The language is the North-Western Prakrit which we know from other Kharoshthi inscriptions and from the Dutreuil de Rhins manuscript. The phonetical system is broadly the same. The vowel ri is represented by ra in grahavati, Skr. grihapati, 11. 2, 3, but by a in gahathubami, Skr. grhastúpe, 1. 2. Intervocalic k has disappeared in uasia, Skr, upāsika, I. 1; shuushaehi, Skrenusha kābhyām, 1. 4, and pratiae, Skr. prāpti kāyai, 1. 5, but is represented by y in samvatsaraye, Skr. samvatsarake, I. 1. Intervocalic g appears as k, i.e., probably & voiced guttural fricative, in-nikamo, Skr. -nigamam, 1. 5. Instead of -ch- we find y in ayariena, 1. 4, and, similarly, also twice ya for the enclitic cha, 1.4, but cha, 1. 3. If Saïta represents Skr. Sachitta this ya has regularly disappeared before i. Intervocalic ja becomes ya in puyae, Skr. pújāyai; puyaita, Skr. pujayitvā, 1. 5, but remains in the name Raja, l. 4, which I cannot explain. In ajasa, 1.1, - - denotes the voiced 8, i.e., 2. Intervocalict is usually preserved, but was probably pronounced as d'; cf. grahavati, Skr. grihapati, II. 2, 3; dhita, Skr. duhitā, 1. 2; dhituna, Skr. duhitrā, 1. 3 ; hotu, Skr. bhavatu, l. 5, but bhraduna, Skr. bhrātrā, I. 3. In the unaccented prefix prati it has been dropped, evidently in consequence of the absence of stress, in praiskuveli, Skr. pratishthapayati, 1. 2; cf. prethavetiye in the Taxila gold plate; prehavide in the Jamälgashi inscription. Intervocálic d disappears, probably after having become a fricative, in sarvastiraana, Skr. sarudstivādânām, 1. 4. Intervocalic p regularly appears as o, which is dropped after u; cf. uasia, Skr. wpåsika, 1. 1 ; grahavati, Skr. grihapati, 11. 2, 3, but becomes b as in some other Khardanthi records in thuba, Skr. stöpa. I. 2. As in the silver scroll, the dental n has throughout been replaced by n. The compound ry becomes ria in ayariena, Skr. acharyena, l. 4, and y in bhaya, Skr. bhāryā, 1. 2. The same double treatment is also found in other Kharöshthi records. Similarly we have privana for Skr. nirvana, 1. 5, but sarva, 1I. 4, 5, as in the silver scroll. The transposition of r in Dhrama, Skr. Dharma, 11. 2, 4, is also known from other sources, but has not been met with in other Kharoahthi inscriptions. New is also the assimilation of a(n) to the ensuing sh in shoushacki, Slar. mushabibhām, 1. 4.
Of inflexional forms I shall only mention the peculiar instrumentals fem. dhituna, Skr. duhitrā, 1. 3, and shņushachi, Skr. snusha kābhyām, 1. 4. They seem to be coined after the pattern of masculine nouns, but we have no right to characterize them as simple mistakes. They were evidently used in the dialect, and Pischel's valuation of the stray Prakrit examples of fem. instr. plur. in ehil should be modified.
After the date, with which I shall deal below, the inscription goes on to record that the female worshipper (upāsi kā) Chandrabhi (Skr. Chandrābhi), the daughter of the householder (grihapati) Dhramma (Skr. Dharma), the wife of Bhadravala (Skr. Bhadrapāla) puts up relica in the house- stūpa' (grihastupa) at Chhadašila.
The term gahathuba (Skr. grihastupa) is new, but evidently means a stupa standing in a griha, 1.0., & roofed building. For we have already seen that our stupa was situated within a chapel that had been roofed over,
Chhadašila, on the other hand, must be the name of the district, or of an old village or town in the immediate vicinity of the monastery to which our stūpa belonged. It cannot well be a synonym of Takshasilā, though the last part of the two names, silā, is evidently the same. Por, in the first place, the new site does not seein to belong to ancient Takshasila. Chhadasil.
* Graminatik der Polbril-Sprachen, $376,
