642 Revision of the Bactrian Alphabet, [July, 



u ma X. This letter admits of no doubt whatever ; but in the Menander 

 form, u>, I now recognise the inflection me, corresponding with the Greek 

 name more closely. — Mi is written f ; md, V or V ; and y may be mu. 

 The second or what may be called the printed form of m has a consi- 

 derable affinity in form with the old Sanskrit y or H, whence it may be 

 almost as readily derived as the Burmese form of Pali, Q. 



A ya. This letter is unchanged : it invariably replaces z, and y, and 

 sometimes j where the latter would be expressed by the Sanskrit ^? or 

 Jj. It may perchance have been modified from the letter, for in some 

 examples it is turned up on the sides thus, ou ; the inflected form A yi 

 is of common occurrence : fy yu, less common. 



v/->, h, S, ra t It is necessary to preserve these three representatives of r ; 

 I incline to think that the prolongation below may be the mdtra or 

 the long d inflection, rd; for the first form is used in Ermaiou where 

 there is no intervening vowel. It is only distinguishable from d by 

 the foot-mark of the latter, which seems to be often omitted notwith- 

 standing: its inflections are fi, 1, ^, ri, re, ru. 



<i, la. Further acquaintance has taught me that this is the only 

 representative of a in Greek names : the instances wherein the I before 

 appeared to be replaced by 1 have been disproved by duplicate coins. 

 The inflected form f+i, li, has numerous examples among our new ac- 

 quisitions. H le, also occurs in inscriptions. 



1 va, and ^ vi, rest on strong but not undisputable authority, as will 

 be seen below. 



ru, 1, ha, has been removed from its former position as / on ample 

 grounds ; and the value now assigned has I think equally strong support 

 — though as far as Greek names are concerned it rests solely on the 

 initial syllable of Heliocles, X, he. There is, again a similarity worthy 

 of remark between ru inverted, and the old Sanskrit ha, [p, fa. 



*P, sa. To this letter I gave the sound of o on the former occasion 

 because I found it the general termination of nominatives masculine 

 in Zend and Pali — replacing the Sanskrit visarga, ah or as. Since 

 then I have found the same letter (affected with the vowel i) in two 

 Greek names as the equivalent of si, p, and I am too happy on other 

 considerations to adopt this as its constant value ; whether the dental s 

 of the Sanskrit will best represent it remains to be seen, but the nearest 

 approximation in form oecurs in the Hebrew d s : there are certainly 

 two other characters, *T, or T, and 71, having the force of* or sh. The 

 former I should presume to be the Sanskrit sha H. from its likeness 

 to the old form fl\. The latter, T1,may be a variation of A for which it is 

 sometimes used, but rather by change of the Greek z to 2, than as 

 being the same letter, for elsewhere it takes the place of the Greek 2 



