THE PALAS OF BENGAL. 89 



defeated the troops of a king of Kanyakubja, whose name has not been discovered 

 as yet. The position of Pithi is also indicated by this reference. It seems 

 to have been a buffer state between those of Kanyakubja and Gauda. The 

 Pratihara dynasty was falling, and the kingdom of the proud Gahadavala was 

 rising on its ruins. It is quite possible that Bhimayasas of Pithi assisted Candra- 

 deva, the Gahadavala, to obtain the city of Kanyakubja and to overthrow the last 

 Gurjara-Pratihara King. The next prince in the order adopted by Sandhyakara- 

 nandi is Viraguna of the forest of Kota, who is also styled "the over-lord of the 



Southern thrones." But nothing is known about this king. 



Dr. Kielhorn's lists of Northern and South Indian Inscrip- 

 tions do not contain any record which mentions this king of the South. J ay a 

 Sirhha, the Lord of Dandabhukti, seems to have been a man of great importance. 

 The position of Dandabhukti has already been indicated. 1 It is represented at 



the present day by the District of Midnapur. So Jaya 



Jaya Simha of Danda- S imha was the march-lord of the South. The commentary 

 bhukti. . 



very appropriately mentions the defeat of the King Karna- 



kesarl of Utkala by this prince. It is more natural for the King of Orissa to fight 



with a prince, whose land lay on his border, than with one, whose possessions 



were separated from his by a belt of mountains and forests. The position of 



Devagrama in Vala-valabhi, the king of which, Vikrama 

 Vikrama-Kesari of _ Deva- Ke «< ari comes next in order is far less cer tain. The com- 

 grama in Vala-valabhi. ' ' 



mentary adds : '' c Devagrama-prativaddha-vasudha-cakravala- 



vaTavalabhi-tarahga-vahala-galahasta-prasasta hastavikramo. ' ' % The explanation 

 of this is not quite certain and nothing can be made out beyond what has already 

 been stated by Mahamahopadhyaya Hara Prasada Sastri. Vikrama Kesari was the 

 King of Devagrama and the surrounding country which was washed by the rivers of 

 Vala-valabhi. Vala-valabhl has been identified by Pandit Hara Prasada Sastri with 



Bagadi, one of the five divisions into which Bengal proper 



The position of Vila- was divided before the Muhammadan conquest, but no 

 valabhi. p ^ ' 



reliable authority whatsoever can be cited in support of it. 



The name Vala-valabhi itself was unknown in Bengal before the discovery of the 



Bhuvanesvara prasasti of Bhavadevabhatta 3 and has not been found anywhere else 



except the Ramacarita. There are hundreds of villages in Bengal bearing the name 



of Devagrama, and I do not find any reason to confine it to one of them. 



Even in the Nadiah district itself there are several Devagramas, and so the attempt 



to identify it with the materials at present at our command is premature. Daksml- 



sura is said to be the Madhusudana of another Mandara and is described as the head 



of all Forest feudatories " Samast-atavika-samanta-cakracuddmanih." + The Mandara 



mentioned here seems to be the hill of that name at present in the Bhagalpur district 



of Bihar. Mandara hill commands the surrounding hilly and 

 Laksmisura of Mandara. r , •,. , ■. ., ■> .-, ,'. .- . • 



forest country to a great distance, and it may be that its king 



1 See Ante, p 71. 2 Comm. on V. 5, Chap. II, p. 36. 



3 Ep. Ind. Vol. VI, p. 205. * Comm. on V. 5, Ch. II, p. 36. 



