556 . Ancient Indian Numerals. [No. 7. 



perpendicular strokes between the two words, to fix definitively the 

 written number."* Major Cunningham claims to have discovered — 

 on the authority of " a stone slab, which gives, in regular order, the 

 nine numerals! — that the numerical signs of this language were 

 expressed by the initial letters of " their Pushtu names written in 

 Ariano-Pali." How soon after Asoka's time this system was intro- 

 duced is not stated, but Major Cunningham considers that he has 

 obtained direct evidence of its local use from 144 B. C. to 31 B. C. 

 and he proceeds to add, with reference to these literal figures, " the 

 first four are given in two distinct forms # * and the two forms show 

 in the clearest manner how the straight horizontal strokes of Asoka's, 

 and even later, days, gradually became the 1, 2, 3 of India, from 

 whence they were transmitted through the Arabs to Enrope."{ 



I now arrive at the most important elucidation that this subject 

 has received since Jas. Prinsep's original discovery, in the " Observa- 

 tions on the dates found in the cave inscriptions at Nasik" by the 



occurred in the Cuneiform Inscriptions, where the low numbers were often defined 

 by little more than rude combinations of the equivalent number of simple strokes, 

 while the decimals and hundreds were far less crudely rendered. Rawlinson, 

 J. R. A. S., No. X. p. 172; Hincks, idem, XVIII. 423. 



* J. R. A. S. XII. 42. 



t J. R. A. S. XII. 225. 



% J. A. S. B. 1854, p. 703. I must confess that I regard this theory with some 

 suspicion ; in the first instance it implies, in effect, the use of a second language in 

 the body of an inscription, the bulk of which is expressed in another tongue ; it is 

 admitted that even the limited number of the unit numerals will not stand the 

 Pushtu test, that the initial for four must still be taken from the chaturo as it 

 occurs ia Asoka's Kapurdigiri inscription; and that the signs for 1, 2, 3 must be 

 traced to other sources. 



Next, we have to concede that this Arian character which was soon to be super- 

 seded by the more exactly expressive Indian Pali, was enabled either at the time of 

 its own eclipse, or at some subsequent period — to associate its literal numerals with 

 the southern system of writing, and having thus early entered into the alphabets 

 of the Sanscrit and certain Indian dialects, that these figures and the perfect sys- 

 tem of notation that they represented, remained uncoramunicated to India at large, 

 till at least the end of the 4th century A. D. (Vallabhi Grant J. A. S. B. VII. 966- 

 and PI. XX. ibid), if not till a far more modern epoch. 



Major Cunningham does not appear to have been sufficiently impressed with the 

 real importance of that portion of his discovery, which determines that so early as 



