1896.] G. A. Grierson — "Report on investigation at Bodh-Gayd. 57 



by-form of / \ ga, which should have appeared on No. 3, and which 

 Cunningham says he saw there under the form ( ] with rounded (not 



angular) corners at the top. According to Cunniugham this base (No. 

 5) should have borne the Brahma letter rja, which, though very rare, 



occurs under the form . 



As this point is of some importance for the purposes of palreography, 

 I give in figure 2 a reproduction of Cunningham's representation of the 

 rubbings of the letters taken by him, and in figure 3 a photograph 

 (actual sizes) of the rubbings taken by me. 



The ga with the round top represented by General Cunningham, 

 occurs, according to Dr. Biihler, once in the Delhi Siwalik column (Ed. 

 VII), and in all inscriptions of the second century B.C. Its form 

 ( together with that of cha, to be noted later on) is of importance as 

 proving the date of the masons' marks. But did it ever exist here ? 

 I cannot believe that so acute an observer as General Cnnningham 

 should not have noticed the perfectly plain mark on No. 5, and it ap- 

 pears to me to be most probable that the round topped ga which he 

 gives is (whatever letter it represents) merely a faulty impression of 

 the square topped letter on No. 5, and that, led by his reading it as ga, 

 he has, by a slip of memory, located it on No. 3, where ga ought to 

 have existed. I may add that No. 3 is now so badly damaged that I 

 do not think it likely that any letter could have been visible on it so 

 comparatively short a time ago as General Cunningham's last visit. 



As to what this angular topped letter on No. 5 is, I do not like to 



give any decided opinion. It is certainly not 5'« ( ), as hitherto known, 

 which should appear on that base (if it is arranged in alphabetical 



order). It might possibly represent hha i I with the right 



hand top appendage worn away, but this is the 23rd consonant of the 

 Indian alphabet, and would be altogether out of place. It may also be 

 a later development of a round topped ga, but I know no other similar 

 case in the ancient A9oka alphabet, though it occurs in later alphabets, 

 especially in the old Pali alphabet of the Burmese. In this case the 

 mark (if the bases are in alphabetical order) should be on the third, 

 and not on the fifth base, and either the bases were not originally in 

 alphabetical order, or their order was changed at some time subsequent 

 to their erection. In the first alternative, the whole theory of the ma- 

 sons' marks representing an alphabetical order falls to the ground. In 



