1840.] from Bactrian and Indo-Scythian coins. 347 



Here we therefore indeed observe an inconsistency in the 

 characters, which depends not simply upon the state of the coins, 

 while from the addition of variously situated points with both 

 letters, we may presume, that the indistinctness was to be amend- 

 ed by additional diacritical symbols. This idea as to the points, 

 cannot be established but by carefully examining the coins 

 themselves. The point at 1 serves perhaps to distinguish it 

 from *1, r. 



Be it as it may, it is evident from our review, that some coins 

 obviously distinguish the second symbol from the third. As 

 now i by the omission of the cross line below, and by a greater 

 curvature in the middle, may easily be altered in the figure £, 

 nothing prevents us from supposing, that the second syllable 

 may still be I or na. 



The penultimate symbol, being proved so incontrovertibly 

 as n, £, I do not hesitate to declare it the above adopted dh I. 

 In the word dhdmiko we observed also instances, showing that t 

 has a straight form 5, and this approaches so much to n, that we 

 can hence likewise account for the confounding of both. 



For r we never meet with a peculiar symbol, and the name 

 must therefore be read Minado. The adoption of t as dh, will 

 lastly be confirmed by the fact, that in the same word it is 

 commonly written 1, and seldom t. The *l upon the coin, 

 R. R. I. No. 10, appears to be indeed confounded in the revers- 

 ed way, viz. "I for i. But I if it be a dental sound, dhdmiko 

 decides for the adoption of dh. 



In Minadho r is absorbed, according to rules of Pracrit, for 

 instance, kandras becomes kando. As t for d appears to be the 

 rule in Menandros, we cannot consider t to be substituted by 

 mere chance for *"l ; but it must be founded upon the nature of 

 the language. We may here recall to mind the rule of the 

 Zend, according to which the Sanscrit putra is changed into 

 puthra. The omission of the r may have had an effect, similar 

 with the Persian sur, (si, or three) which presupposes a previous 

 form thi, the th of which still remained, after r had dis- 

 appeared out of the more ancient form thri in Zend for the 

 Sanscrit tri. 



We have in Minadho for Menandros an evident instance, 



2y 



