558 Ancient Indian Numerals. [No. 7. 



obvious, and are only perplexing in the multiplicity of forms, some 

 of their exponents are seen to take ; the larger sums on the other 

 hand, are expressed by a crude and uncertain method, under which 

 the amount has to be read backwards in the current line of writing ; 

 thus, the generic symbol for thousands is ordinarily entered first, that 

 for hundreds, second, while the specific decimal or unit cipher, which 

 determines the value of the whole, is placed last in the order fol- 

 lowed by the rest of the inscription. At times again, the mark for 

 hundreds is indifferently inserted before or after the figure which 

 indicates the total.* If, by any possibility, further argument were 

 required to that end — this double system of arranging the ciphers 

 would alone establish, that they were incapable of having their 

 value enhanced or diminished by change of place. 



Dr. Stevenson's point of departure, like my own on a previous 

 occasion, was from Jas. Prinsep's investigations of April, 1838, he 

 does not seem to have seen my paper of 1848, and therefore ex- 

 presses no opinion either for or against my position, but continues 

 to follow Prinsep in reading "1 as three, in preference to three hun- 

 dred; at the same time that he admits, that the three horizontal 

 lines fully suffice to express the lower number, for which indeed he 

 has a second variant, and while his own materials contribute se- 

 parate and independent signs for 10, 20, 30, and one hundred : the 

 latter being specifically distinguished from the ordinary generic sign 

 for hundred. 



The next item I have to advert to, is the idea advanced that the 

 Satrap numerals owe their forms to the Bactrian alphabet.f This 



perfect a copy of them as can be obtained in the present state of the rocks. As 

 the facsimiles are the property of Government and executed by another gentleman 

 [Lt. P. F. Brett], I have done nothing more than, to the best of my ability, see 

 that the lithographer executed his task faithfully." Bombay Journal, 1853, p. 57. 

 And again p. 50, Dr. S. observes " it is difficult for me at present to say whether 

 the frequent omissions of the point for ^and other anomalies, belong to the origi- 

 nal, or are the faults of the facsimile." 



* Nasik Inscription, No. 2, plate 7. 



t Dr. Stevenson remarks " In the Satrap inscriptions, the numerals used to ex- 

 press the different sums of money there mentioned are peculiar. At first I could 

 determine nothing about their origin, but on a careful examination I found a strik- 



