R. D. BANERJI ON THE PALAS OF BENGAL. 69 



Gupta era as in that case it would be equal to 1207 a.d., which is certainly too late. 

 The Kalacuri-Cedi era has never been found to have been used in Bengal. The Saka 

 era suits best though it has but been sparsely used in the North- East, In that case 

 S. 888 = 966. a.d. falls just after the reign of Devapala, the Pratihara King of Kanauj. 

 At that time the invaders must have settled down so that the invasion itself must 

 have taken place some time earlier. Northern Bengal was in the undisputed posses- 

 sion of Narayanapala at the time of the incision of Guravamisra's record. So this 

 invasion must have taken place some time between 850-950 a.d. The irruption of 

 these Mongolian hordes must have taken place through the Himalayas, and most 

 probably they were dispossessed of their former homes in the hills by some other 

 invaders. So the Palas after Narayanapala. i.e. Rajyapala, Gopala II, and Vigraha- 

 pala II, were having a rather bad time of it with the Gurjara Empire in the West 

 and occasional Rastrakuta raids thrown in, and with Barbarian hordes advancing 

 in untold numbers through the mountain passes of the North. No wonder that 

 Magadha was annexed to the Gurjara-Pratihara Empire. At the time of the invasion 

 of Indra III, the Eastern Frontier of the Gurjara-Pratihara Empire most probably 

 extended right up to the modern Bhagirathi, and its confluence near Saugor Island. 

 North Bengal must have remained in the possession of the Mongolian kings up to 

 the end of the tenth century a.d. In the beginning of the eleventh century we find 



that the Palas have recovered possession of Northern Bengal, 

 Kamboja kings of Cauda. and frQm ^ time onwards rigM up to the end of the second 



Pala Empire, Northern Bengal continued to be in their possession. At the time of 

 the Dinajpur inscription the Palas seem to have been deprived of Gauda and conse- 

 quently the Mongolian king became Gaude§vara. The name Kamboja itself is of 

 great interest. Thus far the Kambojas or Kamvojas were known to be a northern 

 tribe who lived side by side with the Greeks in i\fghanistan and the Western 

 Punjab, as shown by the phrase " Yona-Kambojesu" in the XIII Rock Edict of 

 Asoka.' The occurrence of the name in a Bengal inscription does not mean that 

 the Kambojas, whole or part, immigrated into Bengal from the Punjab across the 

 whole of Northern India, because that would have been an impossibility in those 

 days, but shows that all Mongolians were called Kambojas, and that people with 

 Mongolian features crossed over into Bengal through the Northern Mountains and 

 as Kambojas. They may or may not have been a part of the people who became 

 known during the Maurya period as the Kambojas. 



The occupation of Gauda by a barbarian tribe, at a time when the whole of 

 Magadha was in the possession of the Gurjaras, shows that the kings of the Pala 

 dynasty between Narayanapala and Vigrahapala II and Mahipala I were kings in 

 name only. Most probably they ruled over an msignficant kingdom surrounded 

 by a large number of petty monarchies. The Tirumalai inscription of Rajendra 

 Cola I shows that the ancient Gauda and Vahga had become divided into a large 

 number of small kingdoms. The exact state and extent of the Pala dominions under 



1 Epi Ind., Vol. II, p. 465. 



