500 Journal of the Asiatic Society of Bengal. [August, 1911, 
India, that is, probably not until about 620 or even 625. 
The date might be even later, because he continued fighting 
mpaig 
Ganjam as late as 643 (ibid., p. 313). His empire lasted till 
his death in 646-7.' After his decease it fell to pieces. ‘‘ After 
his death the local Rajas no doubt asserted their indepen. 
dence’’ and very little is known oie the history of 
Bengal for nearly a century (zbrd., 6 
There were independent kings riishelote in Bengal before 
Harsa extended his supremacy over it, and again after his’ 
death ; and Samacaradeva must be placed either in the first quar- 
ter of the seventh century before Harsa’s conquest, or in the 
third or fourth quarter after his empire decayed. The latter 
alternative is not probable, because of the character of the 
script in this grant. It has been pointed out that this grant 
exhibits the eastern forms of certain letters, Behiese: thoug 
ousted by the western forms generally, yet remain ned ‘in use 
in this remote locality. They might have sonkuniae till is 
first quarter of that century as shewn above (p 499), 
could hardly have persisted about half a century longer (il 
rs fourteenth year of a new king after the dissolution ot the 
mpire), because Harsa’s supremacy over Bengal would have 
facilitated the predominance of the western forms and hasten 
the disuse of the eastern forms. It may be reasonably in- 
ferred therefore, that Samacaradeva reigned in the first quartet 
of the seventh centur 
There is another consideration which supports this infer- 
ence. The king of Pundravardhana, that is Bengal, w 
kingdom was more or less eae to Harsa, belonged to A 
brahman caste, as Mr. V. Smith says (History, p p. #29)? 
this (Farid pur) district poate have appertained to Pu Panes 
Metin ana, and the termination deva in names often designated. 
ans 
Bt 
vardhana dynasty which was reigning when Harsa conquered 
Bengal. . 
The conclusions then which seem fairly established are, 
(1) that this grant was later than C which was executed in the 
year 586 (at the latest), and (2) that it was prior to Har3a 8 
not el of foxes which may be —— to about the 
ee Ree a NS GR See 
re. Vi ea tells me that this year is the correct date of Harse 8 
death, and» t 647. 
ana, an a ly om this sos ge ~ suggests that 
psa did belong to the beats aste y : 
