________________
No. 7.]
THE HATHIGUMPHA INSCRIPTION OF KHARAVELA
(1. 17). The writing seems to have been done by three different hands, and it is possibly on this account that the forms of the characters sometimes differ. The first part of the inscription was written on a carefully dressed surface but after the 4th and 5th lines the surface was undressed; and towards the end of all the lines after the 5th there are numerous chisel-marks which make the decipherment extremely difficult.
73
The language of the inscription is Prakrit. The use of ra, the affix o at the end of the nominatives of masculine stems ending in a, and the absence of palatal a show that it is not Magadhi or any eastern dialect of it. Throughout the inscription the dental has practically replaced all other sibilants. The doubling of consonants, even when necessary, has been omitted. The cerebralisation of dentals is also to be seen, as a paisanthapanam (1. 3). There is also an attempt to approach classical Sanskrit in certain cas es, e.g., Mähämegha vahanena (1. 1) and vamdapayati (1. 12), etc. Throughout the record the liqu ifaction of consonants is absent. This would indicate that the record was composed by a man from Western India who wrote in a literary dialect. It is quite possible that the record was composed by a Jaina monk from Gujarat or the Maharashtra who might have been brought into Kalinga by King Kharavela for the purpose mentioned in 1. 16. The language of the record is a very near appros.ch to the canonical Pali. But, as in Jaina Sauraseni, tha in this dialect becomes dha, cf. padhame (1. 3), radha (1. 4), vitadha (1. 5), Goradhagiri- (1. 7), Madhuram (1. 8), Bharaalhavasa- (1. 10), Utaräpadha- (1. 11), but not in all cases, cf. Pithumda- (1. 11). The extraordinary form chavuthe (1. 5) is a graphic form of chaithe, but Prakrit cha" becomes cho in Choyathi (1. 16) which. is a later Prakrit form of chatuḥshashti. So also Sanskrit tvaritam becomes tur iyam (1. 16) instead of turiam. Perhaps turitam was the intermediate form but the change of ta into ya instead of a is exceptional.1
The characters of the inscription show great var iety. Among the vowels, the initial forms of a, a, i, e, o are to be found in it. The medial a its denoted by a perfectly horizontal line in many cases as well as by a slanting stroke, as in pathaj payati (1. 4) and -samuthäpitähi (1. 15). The medial form of the long i shows two diverging slanting strokes, in -sarira- (1. 2) and two converging ourved strokes in kiḍitā (1. 2). Among consonants two different forms of kha are noticeable, Khāravele (1. 1) and lekha (1. 2) in which the triangle at the bottom is either present or absent. Two different forms of ga are present, the angular Maurya form, as in nagarim (1. 5), and the broad. backed Kushāna form, as in Kalimg-a° (1. 1). Three forms of gha are clear: (1) the Maurya form with a circular bottom, (2) the transitional form in which the only side of the curve has turned into a right angle, ef. -oghatitam (1. 6), and (3) the complete second century form in which, though the length of the left-hand side vertical is not reduced, the lower part of the letter consists of two right angles instead of a curve. Two different forms of ta are to be found: the Maurya, as in hita-, -ti-vasa- (1. 6) and the later, as in "pateye (1. 6). In the case of da also we find two forms: the Maurya, as in pamdarasa- (1. 2) and the later, as in dati (1. 9). Three different forms can be distinguished in the case of pa and ma(1) the Maurya form, as in pathapayati and pachkima(1. 4); (2) the transitional form, as in patina (1. 1), -käripanahi (1. 5), matuka- (1. 7), Madhuram (1. 8), etc.; (3) the later form with angles fully developed in the place of curves as in kapa-rukhe (1. 9), apayato (1. 8), -samaja (1. 5), and Satamam (1. 7). It should be noted that the Kushapa form of ma is altogether wanting in this inscription. So also in the case of ha we find the Maurys form in -sahasehi and haya- (1. 4), the transitional form in Mahārāja and Mähämeghavāhane (1. 1) and the angular form in -vavahara- (1. 2). The general duct of the writing shows that the Hathigumpha inscription was inscribed at a time when the length of the verticals had not begun.. to decrease and the curves had just begun to become angularised. Therefore, this inscription
[866 f. hk. 11 & 12 on page 80 below.-Ed.]