________________
No. 7.]
THE HATHIGUMPHA INSCRIPTION OF KHARAVELA.
inentions a people called Avami or Atuarni near the Krishna. It is quite possible that the city of Pithumda mentioned by Ptolemy as Pitundra (Ind. Ant., Vol. LV, p. 145), was founded by these people (J. B. 0. R. 8., Vol. XIV, pp. 15 f.).
34. Pithunda-According to Ptolemy, & city in the upper part of the Coromandal coast: This city has perhaps to be taken as the capital of the Ava of Avarni.
35. Tramira or Tamira is the equivalent of Tamila just as Aira represents Aiļa. It has been pointed out that Tamil is the origin of Dravida and of Dramila. Tramila would thus be a perfectly correct substitute for Dramla.
36. Māgadhat.-We have examined the tock very carefully. The cast of the inscription in the Patna Museum shows that the word cannot be Muriyd as proposed by Dr. F. W. Thomas. The chisel-cuts of the letters are still clear though the decay of the rock surface has made inpressions of this part of the inscription smudged.
37. Bahasatimitan.-The sa in the word is of the later 2nd century B.C. type in which & clear right-angle has been formed in the right limb of the letter and therefore it looks more like pa than sa. The reading is absolutely certain.
38. Kālimga-Jina perhaps means Sitalanātha who was born at Bhadalapute, which is the same as Bhadrapura' or Bhadrachalam in the Godăvari district of the Madras Presidency.
39. padīhārehi stands for pratihāraih.-The da denotes its difference from parihara in 1. 9.
40. Anga- Magadha.—The mention of these two provinces indicates that the campaign of the 12th year was speciaby directed against South Bihar which now includes the ancient provinces of Adga (Bhāgalpur and Monghyr districts) and Magadha (Patna, Gayá såd part of Hazaribagh districts).
41. sata-visikanan.-The reading is not doubtful though it was tead slightly differently before. The reference seems to be do architects (baris mikas).
42. hathi-niväst-parisaram. This reading of the second word of the phrase is more reliable than the previous one (hathi-nāvana-paripuram). nivasa=nitrāsa.. The teferet ce seems to be to elephant preserves or some original form of theda arrangement, invented by Khāravela. The elephants of Kalinga were very famous in ancient India. The strerigth of Khårávela probably lay in elephants. He took the fort of Pataliputra with the help of elephants.
43. Parda-raja. For the form of rani in l. 15. The Sanskritized version of the name of the great Dravidian dlatt is Pandya and is detived from Párdu by a speciál rule of early grate marians. In the inscriptions of Asoka also the vowel in the first syllable is long.
44. Pa-khi nd-samsitehi.-The reading seems to be tolerably certain on the rock, though 16 looks doubtful in the impressions. Pa-khind=prdkskind, sarsito =sanssitio.
45. Yāpa-flavahelvi. The first part of the compound cannot be satisfactorily explained by ug though the reading is absofttely certai.
46. vasá-sitāni.-The medial vowels except that on the first syllable are somewhat uncertain. It may also be taken to be -satāni. It was previously read as vasäsitāni.
47. Uvisagd-Keranele previously road to Urdsd-Nhäradeldt. The go and te ate to tinct.
48. samghayanan.-The reading is certain except of the vowel value on gh: Certain improvements have been effected, e.g., filaninań which makes the purport of this term fairly oertain. 1 Ind. Anb., Vol. XLIII, p. 64.
*J. R. A: 8., 1922, p. 84. Inid. Ant., Vol. HI, p. 136.
.Of. Kamandakin-nitisara, XV, 7. [See Muni Punyarijaya, Anekanta' (Delhi), Vol. I, p. 142, where he cites Jours torte on yapa' ("religious life')-K. P.J.}
