1855.] Ancient Indian Numerals. 561 



to symbols shaped like the letters j% and ^f% and ending with a 

 form like the Bactrian ch ; but I am inclined to consider each and 

 all of these examples as either ignorantly executed or incorrectly 

 copied types of the standard s. 



The value of the cipher for five was also first contributed by the 

 copper-plates, its component letters here read clearly as «fT. Its 

 numismatic equivalent appears to exist in the unit to be seen on the 

 leaden piece, No. 29, PI. II. J. E. A. S. XII. of which the East 

 India Company's coin of Vijaya Sah seems to offer another variety.* 

 It is difficult to trace the slightest affinity between any of these 

 forms and the modern figure of the Devanagari system. The only 

 verbal number that would at all accord with the initial «tt, would be 

 •TT1, a serpent, which term however is used to represent eight : the 

 Nasik inscription further determines, that with these very dissimilar 

 types, must be classed a figure offering an absolute counterpart 

 of the ^ ru of the Sah alphabet, add to this, the same authority re- 

 quires the admission to a joint classification, of a variant of more 

 formal outline, which gives an additional forward head stroke such 

 as is used to express the letter ^8fT ;t under which aspect, the sign 

 may be compared to an English E. Dr. Stevenson exhibits, beyond 

 all these, two other varieties of the five, the one in a reversed ^ or 

 rather of a form like the Gupta #r n,J while the second in a mea- 

 sure follows the outline of a Bactrian P, but in both these instances, 

 I am more disposed to concur with that author in distrust of his 

 materials, than in fanciful comparisons with the Bactrian alphabet. 



The ancient Indian six, is given in Dr. Stevenson's plate, but with 

 the expression of a doubt regarding the identification ; the second 

 or Satrap form, which may be read as fi, is seemingly entered with 

 more confidence. I am unable to trace any numismatic figure that 

 will satisfactorily accord with the latter. 



The seven, under the evidence of the Nasik Inscriptions, may be 

 indicated in the unit letter on coin 14, PL I. J. E. A. S. XII. which 

 is elsewhere of frequent occurrence. The correct outline of the 



* J. R. A. S. XII. 54. 



t J. R. A. S. XII. p. 40, and fig. 23, PI. I. 



% J. R. A. S. XII. PI. II. fig. 41. 



4 e 2 



