86 Journal of the Asiatic Society of Bengal. [March, 1914. 



inscriptions in the eighth or ninth century a.d." Dr. R. L. 

 Mitra was also of the same opinion. But a careful comparison 

 with some of the inscriptions of the seventh century a.d. 

 would readily show that the plates must belong to that period. 

 The characters of the Shahpur image inscription of Aditya Sena 

 dated in the 66th year of the Harsha Era (672 a.d.) l , and the 

 Apshad inscription of the same king 2 are much more developed 

 than the characters of the Asrafpur plates. Indeed, the charac- 

 ters of these plates are so akin to those of the Madhuban and 

 the Banshkera 3 plates of Harsha that these four plates must 

 belong to the same period. Further corroborations are not 

 wanting. 



We know from the Allahabad Pillar Inscription of Samudra 

 Gupta that the kingdom of Samatata, Davaka, Kamarupa, 

 Nepala, etc., were on the border of his empire. So we see that 

 there were clearly defined kingdoms in East Bengal at this 

 early period. We cease to hear any more of the kingdom of 

 Davaka, but that the kingdom of Samatata continued to have 

 an individual existence is clear from the accounts of Hiuen- 

 Tsang, who visited the kingdom during his travels. It is to be 

 regretted that he does not mention the name of the king of 

 Samatata. From his description, it appears that there were 

 thirty or more Sangharamas at Samatata with 2000 priests. 

 The maintenance of thirty Sangharamas with 2000 priests 

 seems to urge that the king belonged to the Buddhist creed. 

 This is further corroborated by the assertion that Silabhadra, 

 the teacher of Hiuen-Tsang, belonged to the Royal family of 

 Samatata; moreover, It-Sing expressly mentions that the king 

 of Samatata was a devout Buddhist. 



What was this Buddhist Royal Family of Samatata that 

 had given a principal to the University of Nalanda in the 

 person of Shilavadra ? We believe it was the Khadga family. 

 The only difficulty in the way of the identification is the 

 mention of Shilabhadra aa a Brahmin by Hiuen-Tsang. The 

 title of Khadga is avowedly a Kshatriya title. But Hiuen- 

 Tsang also calls Bhashkara Varman, king of Kamrupa, a 

 Brahmin. If Varman can be a brahmin, I do not see why a 



cannot. 



I have already pointed out the close resemblance of the 

 characters of the Asrafpur plates with those of the plates of 

 Harsha. This itself, as we have already said, is a proof that the 

 Buddhist Khadga family belongs to that period. The asser- 

 tions of It-Sing (673 to 687 a.d.) dispel the least shade of 



doubt. 



The first plate of Asrafpur states that it grants land to secure 



i Fleet's 6 * Gupta Inscription," pi. xxix. 



2 Ibid. . pi. xxviii. 



3 Epigraphia Indica, Vol. I, p. 67, and Vol. IV, p. 210. 





