________________
DECEMBER, 1925)
ANCIENT TOWNS AND CITIES IN GUJARAT AND KATHIAWAD
56. Harshapura. There is only one inscriptional reference to Harshapura ; it occurs in the Kapadwanj plates of Akalavarsha Subhatuiga and his feudatory Mahasamanta Prachchhands 201 Therein we read :-अस्तु वः मविविदितं यथा श्रीखेटकहर्षपुरकासद्रह एतम्र्धाष्टमयं...मया श्रीहर्षपुर्धाष्टमश FIFT:
a raffarna... From this it is evident that (i) the towns Khetaka, Harshapura and Kasadraha were situated not far from one another and that (ii) the Kapadwanj or Karpata vånijya sub-division formed part of Harshapura district.
Dr. Bhagwanlal Indraji thinks that this Harshapura may be Harsol in the Präntei Taluka of Ahmadabad district. This seems to be the case, for the distance of Harsol from Karpatavanijya or Kapadwanja being only 20 miles, it is possible to regard the latter as a sub-division of the former district. Nor is Harsol too far from Khetaka and Kasadraha ; for Kaira and Kasandra are only 50 and 40 miles respectively from Harsol. These distances are not considerable for we must remember that Harshapura was the headquarter of a district which was distinct from the Kheţaka district. Distance between the headquarters of two contiguous districts may well be 50 miles.
Nor are there any philological difficulties in the way of the proposed identification. We have already quoted 202 instances of a disappearing 'pa' changing its preceding 'a' or 'A' to o': so Harshapura, first became Harshor and then Harsol, l' being as usual substituted for 'r'not only for simplification but also for dissimilation, one 'r' having already occurred.
That Harsol though now a village was once a town can be seen by some fragmentary ruing near it. There is a tower to the east of the town bearing an Arabic inscription of 1599. This shows that Harsha pura continued to be a place of some consequence to the end of the sixteenth century.
57. Hastakavapra. Hastaks vapra is the same as Hathab in the Bhavnagar territory. There are no philological difficulties in the identification. 'Ka' was optional as early as the sixth century; for two grants of King Dharasena I separated by the distance of only two years from each other spell it differently, once retaining and once omitting the,' ka.208 The liquida ' was dropped and the preceding 'a' lengthened by way of compensation. There being two consecutive conjunct consonants, the first sta' was simplified by changing it to 'tha'; and finally the resultant Hathapra changed into Hathab, the final conjunct being simplified, and 'p obanging into 'b' owing to the stress which the syllable originally bore. Locally the name is still pronounced as Hath&p.
The plates however supply additional evidence for the identification of Hastakavapra with Hathab; for the villages Kukkuța and Maheshwara-Dasenaka mentioned in them as being situated in the Hastakavapra Ahåra are modern Kakad and Mahadevapur, respec. tively, both being within a few miles of Hathab.
Hastakavapra then was at Hathåb, neither at Talaja as Yule thinks nor at Gopinaths as Lassen opines. It is difficult to see how Hastakavapra can develop into Talaja, as neither 'la' norja' occur in the original word. Gopinatha on the other hand has no philological connection with Hastakavapra and Lassen does not state if, when, now, and why the old name Hastakavapra was changed into the modern one of Gopinatha. 301 Ep. Ind., I, p. 55 ft.
203 Vide back under Vatapadrapura. 308 Cf. 91 at a tu...dated 207 G.E.; and 16
TE 9 78: dated 209 G.E.
