Vol. VILL. No. 9.| The Bakhshali Manuscript. 353 
[N.8.] 
a. 
Having indicated that Dr. Hoernle’s arguments are not 
altogether convincing, it remains to examine from our own 
standpoint the manuscript itself. 
When Dr. Hoernle edited the Bakhshali MS. the great 
pesuonity on epigraphy was Biihler, who wrote °: ‘* Lhe ol 
wn Sa 
Kiragrama (Kangra), dated a.D. 804.... 
able that the Bakhshali manuscript, found in the Yisufzai 
district, belongs to the same or even a somewhat earlier 
period.”” It has, however, since been shown that the date of 
the Baijnath Prasastis is a.p. 1204! 
This was proved by Kielhorn,’ and Dr. Vogel states that 
independently of Professor Kielhorn’s researches a close examin- 
The earliest Sarada inscriptions of Chamba, which can be ap- 
idagdha, the immediate successors of Sahilla, who may be 
placed in the tenth century.’’ * 
There are thus very clear indications that a re-examination 
of the Bakhshali manuscript is required before the conclusions 
of Dr. Hoernle or Professor Biihler can be accepted. We have 
now material for comparison that was not available formerly, 
viz., Dr. Vogel’s account, already referred to, of Sarada inscrip- 
tionsfound inChamba. Utilizing the criteria given by Dr. Vogel, 
we have fairly certain indications of the age of the manuscript. 
These criteria individually, however, are not infallible and 
there are some difficulties in determining the age peculiar to 
the Sarada script. Nevertheless, the evidence is sufficiently 
clear to enable us to arrive at a fairly accurate result. 
V. 
The Bakhshali manuscript is written on birch bark. This 
material was used in early times, but few of the extant birch- 
bark manuscripts are earlier than the fifteenth century.’ It 
1 Indian Paleography, p. 57. 
2 Indian Antiquary xx (1891), 154. There is a correlated point of 
great interest for these piates were quoted as containing some of the 
earliest examples of the modern place-value arithmetical notation. 
38 Antiquities of Chamba State, p. 43. : 
4 ib., p. 46. ; 
5 Biihler, Ind. Pal., p. 93 
