________________
540 POLITICAL HISTORY OF ANCIENT INDIA
with Yendipalli in Vizagapatam or Endapilli in Ellore Taluk. Kanchi is Conjeeveram near Madras. Avamukta cannot be satisfactorily identified. But the name of its king Nilaraja reminds us of Nilapalli, "an old seaport near Yanam" in the Godavari district. Vengi has been identified with Vegi or Pedda-Vegi, 7 miles north of Ellore between the Krishna and the Godavari. Its king Hastivarman was identified by Hultzsch with Attivarman (of the Ananda family). But the more probable view is that he belonged to the Salakayana dynasty. Palakka is probably identical with Palakkada, (or Palatkaṭa) a Pallava royal residence or seat of a viceroyalty in Guntur or Nellore in South India. Allan and G. Ramdas locate it in the Nellore district. Devarashtra is the Yellamanchili taluk of the Vizagapatam district. 5 Kusthalapura is, according to Dr. Barnett, probably Kuttalur, near Polur, in North Arcot."
The capture and liberation of the southern kings, notably of the ruler of Koṭṭūra near Mt. Mahendragiri remind us of the following lines of Kālidāsa's Raghuvaṁśam ;
Grihita-pratimuktasya sa dharma-vijayî nripah Śriyam Mahendra-nāthasya jahāra natu medinim
1 Gazetteer of the Godavari District, Vol. I, p. 213. Curiously enough, the Brahma Purana (ch. 113. 22f) mentions an Avimukta-kshetra on the bank of the Gautami, i. e., the Godavari. Cf. Avimuktesvara, Anantapur, 164 of
Rangacharya's List.
2 Attivarman was wrongly assigned to the Pallava race. Cf. IHQ, I, 2, p. 253 Ind. Ant., IX, 102. But he is actually described as born in the lineage. of the great saint Ananda (Bomb. Gaz., I. ii. 334; Kielhorn, S. Ins., 1015; IA, IX, 102; ASI, 1924-25, p. 118).
3 The name Hastivarman is actually found in a Śālaйkāyana Vamśāvali (IHQ, 1927, 429; 1933, 212; Pedavegi plates of Nandivarman II).
4 IHQ, I. 2, 686. Cf. Ep. Ind. xxiv. 140.
5 Dubreuil, AHD, p. 160; ASR, 1908-09, p. 123; 1934-35, 43, 65.
6 Cal. Rev., 1924, p. 253 n. Cf. Kutalaparṛu, MS. 179 of Rangacharya's List.
