340 Translation of a Bactrian Inscription. [No. 4, 
that of the last named prince. Further, as the princes named, although 
Tartars by birth, were staunch Buddhists by profession, I supposed 
that it would not be unreasonable to assume that they had adopted, 
with his religion, the era of their patron saint and reformer, and the 
century of Huviska would therefore be the 5th and not the 4tli. 
According to the most recent calculations* Buddha died 477 years 
before Christ, and the Tartar princes Kaniska and his brothers 
Huviska and Juska lived between 42 B. C. and 18 A. C., conse- 
quently their age after the Nirvana should be 441 to 497. Col. 
Cunningham places the accession of Huviska 41 years and 9 months 
before Christ, f or 436 years after the Nirvana, which would very 
closely accord with the date of the Wardak record, if its figures be 
taken for 441. That the era of Buddha (BucldhaMla) was at one 
time current in India, as it is to this day in Ceylon, there can be no 
question, and that a zealous follower and prince should adopt it at a 
time when the religion of /Sakya was at its culminating point, was 
certainly not improbable. No doubt this mode of arguing would 
have the appearance of postponing facts to theories, but it would not, 
I thought, be deemed illegitimate in cases where the value of figures 
has to be deduced from our knowledge of the contents of the docu- 
ments in which they occur, and by a calculation of probabilities ; 
though the result must, in such cases, remain open to future correction. 
There are serious objections, however, to this reading of the date, 
and the most important of them appears to be the fact that the ciphers 
used are identically the same as those found in Sanskrit inscriptions 
of the 1st century before Christ, which necessitates our reading them 
from the left to right, and thereby removes the possibility of 
the first two figures being of the same value. Messrs. Bayley and 
Thomas take the date for 133, reading from the left to right. Col. 
Cunningham, I understand, follows the same mode of reading, 
but by assigning different values to the figures, makes them stand 
for 377. If, as no doubt is the case, all the ciphers used in Indian 
writing are nothing more than contractions of words or modifications 
of their initial letters, a priori , the argument in favour of reading 
them always from the left, whether they be found in Arian or Semitic 
records, would be strong, and yet there is nothing absolutely to 
* Miiller’8 History of Sanskrit Literature p. 2G3 et seq. 
+ Numismatic Chronicle, Yol. YI. p. 18. 
